JP2001273728A - Magnetic disk device, disk access method thereof, and disk access program recording medium for magnetic disk device - Google Patents

Abstract

(57) [Summary] [PROBLEMS] Even if an uncorrectable error occurs during normal use, a request for large-capacity continuous data such as AV data without generating a rotation waiting time accompanying error processing of a host device. It is an object of the present invention to provide a magnetic disk device, a disk access method, and a disk access program recording medium that can ensure the bandwidth to be accessed and the continuity of access. SOLUTION: When a large-capacity continuous data such as AV data written on a magnetic disk 2 is verified and an uncorrectable error is detected in a sector 4, the sector 4 is mapped to a nonvolatile storage device 5 by an error map. The data is registered, and data assurance processing such as rewriting is performed before reading access while referring to the error map, thereby eliminating errors.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a magnetic disk device known as a large-capacity storage device using a magnetic disk as a recording medium, a disk access method thereof, and a disk access program recording medium for the magnetic disk device. In particular, even if an uncorrectable error occurs during normal use, the occurrence of the error can be eliminated without generating a rotation wait time, and the bandwidth required for large-capacity continuous data and the continuity of access can be secured. Regarding what you did.

[0002]

2. Description of the Related Art In a magnetic disk drive, a medium defect, that is, a scratch on a recording medium or an incomplete portion of a magnetic layer, etc., is common, and inspection for these magnetic disk defects is performed at the time of shipment from a factory. The information of the defective part detected at the time of factory shipment is written in the error free area of the magnetic disk, and the host device reads the information and issues a replacement instruction such as a replacement track or a replacement sector.
These parts are not used. On the other hand, read formatting is performed after formatting to check whether the magnetic disk is properly formatted. These processes are performed to eliminate uncorrectable errors and prevent errors from occurring at the time of read access.

[0003]

However, the above-mentioned 2
The two processes are not performed at the time of write access or read access during normal use. That is, there is no information on the defect of the recording medium that occurred during normal use, and no verification has been performed on the written data. Therefore, an uncorrectable error that has occurred in data written during normal use cannot be known until an error occurs during read access. Further, when an error occurs, the fact is notified to a higher-level device.

FIG. 2 shows a disk access method of a conventional magnetic disk drive. In FIG. 2, 4A is a GOOD SECTOR which is a sector having no medium failure or has a correctable error, and 4B is a BADSECTOR having an uncorrectable error.

[0005] During read access to certain data,
Sector B with uncorrectable error on magnetic disk medium
If there is an ADSECTOR 4B and an error occurs, the magnetic disk device notifies the host device of the error.
During that time, the access is interrupted, the host device performs an error process, and then sends a read restart command. During that time, a rotation wait time occurs in the magnetic disk device.

By the way, recently, A, such as television broadcasting,
In many cases, a magnetic disk device is used as a recording medium for V data.
When handling large-capacity continuous data such as AV data, it is necessary to ensure the bandwidth required by the data and the continuity of access. However, when an uncorrectable error occurs, a rotation waiting time occurs due to the error processing of the host device, and it is difficult to secure them. Therefore, the higher-level device secures the bandwidth required by the data and the continuity of the access,
Since error processing must be performed, the data management and error processing are complicated.

The present invention has been made in order to solve the above-mentioned problem, and detects an error before the occurrence of an uncorrectable error, and performs processing such as verification performed in a special case such as shipping or formatting. A magnetic disk device that can be performed even during normal use, a disk access method thereof,
And a disk access program recording medium for a magnetic disk drive.

Further, the present invention eliminates the rotation waiting time associated with error processing of a host device, and secures a band required for large-capacity continuous data such as AV data and continuity of access, and a disk access method therefor. , And a disk access program recording medium for a magnetic disk device.

[0009]

A magnetic disk drive according to the first aspect of the present invention uses a magnetic disk as a recording medium and accesses the magnetic disk to read and write various data. Write access means for performing data write access to the magnetic disk; read access means for performing data read access to the magnetic disk;
Verifying means for verifying the data on the magnetic disk; non-volatile storage means for registering an error map of the magnetic disk; and storing the position and result of the error on the magnetic disk detected by the verification in the non-volatile memory. Error map registration means for registering in the error map of the storage means, and when transferring data between the magnetic disk device and a higher-level device connected to the outside thereof, write access of the data to the magnetic disk is performed. After the completion, the data is verified on the magnetic disk, and if an uncorrectable error is detected,
Sequence control means for controlling the write access means, the verification means, the read access means and the error map registration means so as to register the position and the result of the error on the magnetic disk in an error map of the nonvolatile storage means; It is characterized by having.

According to a second aspect of the present invention, there is provided a disk access method for a magnetic disk drive, wherein a magnetic disk is used as a recording medium and the magnetic disk is accessed to read and write various data. In the disk access method, when transferring data to or from a higher-level device connected to the outside, after verifying the data write access to the magnetic disk, verifying the data to the magnetic disk And when an uncorrectable error is detected, the position and the result of the error on the magnetic disk are registered in an error map of a nonvolatile storage unit.

A control program recording medium for a magnetic disk device according to the invention of claim 3 of the present application uses a magnetic disk as a recording medium and accesses the magnetic disk to read and write various data. In a recording medium on which a data control program is recorded, when transferring data to and from a higher-level device connected to the outside, after terminating the data write access to the magnetic disk, Verifying the data, and when an uncorrectable error is detected, registering the position and the result of the error on the magnetic disk in an error map of a nonvolatile storage unit;
It is characterized by the following.

A magnetic disk device, a disk access method thereof, and a disk access program recording medium for a magnetic disk device according to claims 1, 2 and 3 of the present invention are provided.
As described above, a magnetic disk is used as a recording medium, and in a magnetic disk device that accesses the magnetic disk and reads / writes data of various types of information, a fixed bandwidth is set between the magnetic disk device and a host device connected to the outside. When transferring large-capacity continuous data such as AV data that requires
After a write command is received from a host device and write access to large-capacity continuous data such as AV data is completed, verification is continuously performed on large-capacity continuous data such as AV data. When a possible error is detected, the physical position of the erroneous sector on the magnetic disk and the error level are registered in an error map of the nonvolatile storage means. Thereafter, by referring to the error map, processing such as rewriting is performed on the sector having the uncorrectable error. In this state, if a read command is received from the host device,
The control circuit performs read access, but when performing read access to a sector having an uncorrectable error, the error processing is already performed for the sector having the uncorrectable error. So no error occurs.

According to this structure, the magnetic disk device,
The disk access method and the disk access program recording medium for a magnetic disk device are provided with an error map registered in a nonvolatile storage device before receiving a read command from a higher-level device and performing read access to large-capacity continuous data. And performs error processing on the erroneous sector. Therefore, it is possible to eliminate the occurrence of an error at the time of read access as in the prior art, and it is not necessary for the higher-level device to receive the notification of the occurrence of the error and perform processing for the error at the time of read access. Therefore, the rotation waiting time associated with the error processing of the host device is eliminated, and the bandwidth required for large-capacity continuous data such as AV data and the continuity of access are ensured.

According to a fourth aspect of the present invention, there is provided the magnetic disk drive according to the first aspect, wherein the sequence control means uses the verifying means to execute the error correction in addition to the uncorrectable error. Correctable E
Even when a CC error is detected, the nonvolatile memory means and the error map registration means are controlled to register the position and the result of the error on the magnetic disk in an error map of the nonvolatile memory means, and After the verification is completed, the upper-level device is notified that an ECC error has occurred.

According to a fifth aspect of the present invention, there is provided a disk access method for a magnetic disk drive according to the second aspect of the present invention.
Even when a correctable ECC error other than the uncorrectable error is detected by the verification, the position and the result of the error on the magnetic disk are registered in an error map of a non-volatile storage unit. After the verification, notifies the host device that an ECC error has occurred.
It is characterized by the following.

A magnetic disk drive control program recording medium according to a sixth aspect of the present invention is the magnetic disk drive control program recording medium according to the third aspect, wherein the uncorrectable error is obtained by the verification. In addition, when a correctable ECC error is detected,
Registering the position and the result of the error on the magnetic disk in an error map of a nonvolatile storage unit, and, after the completion of the verification, notifying a host device that an ECC error has occurred;
It is characterized by the following.

According to a fourth aspect of the present invention, there is provided a magnetic disk drive, a disk access method therefor, and a disk access program recording medium for a magnetic disk drive.
As described above, a magnetic disk is used as a recording medium, and in a magnetic disk device that accesses the magnetic disk and reads / writes data of various types of information, a certain bandwidth is set between the magnetic disk device and an externally connected higher-level device. When transferring large-capacity continuous data such as requested AV data, in normal use, a write command is received from a higher-level device, and write access to large-capacity continuous data such as AV data is terminated. The verification is performed on large-capacity continuous data such as AV data, and when a verifyable ECC error is detected in addition to an uncorrectable error as a result of the verification, the magnetic field of a sector having a correctable ECC error is detected. The physical position on the disk and the level of the error are registered in an error map of the nonvolatile storage means. After the verification is completed, the magnetic disk device notifies the host device that there is a correctable ECC error. The determination of the process for the correctable ECC error is made by the host device.

With this configuration, the host device is notified from the magnetic disk that there is a correctable ECC error, so that the host device can always judge the process of the correctable ECC error. E which is correctable at the moment
Even with a CC error, there is a possibility that an uncorrectable error may occur during repeated read access, but since the host device can determine the processing by itself, the error becomes an uncorrectable error. Before, the correctable ECC error is corrected by the judgment of the host device,
Processing such as writing on a magnetic disk can be performed. Therefore, the ECC error does not become an uncorrectable error, an error does not occur at the time of read access, the rotation waiting time associated with the error processing of the host device in the conventional method is eliminated, and the AV data and the like are eliminated. The bandwidth required for large-capacity continuous data and the continuity of access are ensured.

According to a seventh aspect of the present invention, in the magnetic disk drive according to the first aspect, the sequence control means stores the error map in the error map by the higher-level device before performing the verification. E to register
When a level of a CC error is set and the sequence control means detects that the ECC error detected by the verifying means has reached the set level, the sequence control means stores the position and the result of the error on the magnetic disk in the nonvolatile memory. Controlling the non-volatile storage means, the error map registration means, and the write access means so as to register the data in the error map of the non-volatile storage means and automatically perform processing such as rewriting on the magnetic disk. It is a feature.

The disk access method for a magnetic disk drive according to the invention of claim 8 of the present application is the disk access method for a magnetic disk drive according to claim 2,
Before performing the verification, the level of the ECC error to be registered in the error map is set by the host device, and when the ECC error detected by the verification has reached the set level, the position of the error on the magnetic disk and The result is registered in an error map of a nonvolatile storage means, and processing such as rewriting is automatically performed on the magnetic disk.

According to a ninth aspect of the present invention, in the control program recording medium for a magnetic disk device according to the third aspect of the present invention, the control program recording medium for the magnetic disk device is provided by a higher-level device before performing the verification. When an ECC error level to be registered in the error map is set and the ECC error detected by the above verification has reached the set level, the position and the result of the error on the magnetic disk are stored in an error map of the nonvolatile storage means. Register with
A process, such as rewriting, is automatically performed on the magnetic disk.

According to a seventh aspect of the present invention, there is provided a magnetic disk drive, a disk access method therefor, and a disk access program recording medium for a magnetic disk drive.
As described above, a magnetic disk is used as a recording medium, and in a magnetic disk device that accesses the magnetic disk and reads / writes data of various types of information, a certain bandwidth is set between the magnetic disk device and an externally connected higher-level device. When performing transfer of large-capacity continuous data such as requested AV data, a write command is received from a host device, write access to large-capacity continuous data such as AV data is completed, and then verification of large-capacity continuous data is performed. Before doing
The higher-level device sets an ECC error level to be registered in the error map of the nonvolatile storage unit. As a result of the verification, an ECC error is detected. If the error level has reached a set level, the physical position of the erroneous sector on the magnetic disk and the error level are stored in a nonvolatile storage means. And the magnetic disk device automatically performs processing such as rewriting on data having an ECC error.

With this configuration, an ECC error can be detected by verifying large-capacity continuous data such as AV data immediately after the end of the write access to large-capacity continuous data such as AV data. Since the magnetic disk device can refer to the error map of the non-volatile storage unit in which the detected physical position of the sector having the ECC error on the magnetic disk and the error level are registered,
Data assurance processing such as rewriting of data having an ECC error can be automatically performed. Therefore, when there is an uncorrectable ECC error at the time of read access, an error does not occur, and the rotation waiting time accompanying the error processing of the host device in the conventional method does not occur. Therefore, the bandwidth required by large-capacity continuous data such as AV data and the continuity of access are ensured.

According to a tenth aspect of the present invention, in the magnetic disk drive according to the first aspect, the sequence control means performs the read operation such that the verify operation is performed in units of one track. Controlling access means and verify means.

In the disk access method for a magnetic disk drive according to the present invention, in the disk access method for a magnetic disk drive according to the second aspect, the operation is performed in units of one track in the verification. It is characterized by the following.

A disk access control program recording medium for a magnetic disk device according to the invention of claim 12 of the present application is the disk access control program recording medium for a magnetic disk device according to claim 3, wherein Is performed on a track-by-track basis.

As described above, the magnetic disk device, the disk access method thereof, and the disk access program recording medium for the magnetic disk device according to the tenth, eleventh, and twelfth aspects of the present invention use a magnetic disk as a recording medium. In a magnetic disk device which accesses the magnetic disk and reads / writes data of various types of information, transfer of large-capacity continuous data such as AV data requesting a certain band to / from an externally connected upper-level device. In this case, during normal use, after a write command is received from a higher-level device and write access to large-capacity continuous data such as AV data is completed, the operation is continuously performed for large-capacity continuous data such as AV data. For verifying, for one track on the magnetic disk, A
Immediately after the writing of large-capacity continuous data such as V data is completed, the above-described verify is performed on the same track, and then the head of the magnetic disk drive SEEKs to another track on the magnetic disk, and Immediately after write access to large-capacity continuous data such as AV data is completed, verify is performed on the same track, and so on. The verification is performed on the same track in units of one track.

With this configuration, in normal use, the magnetic disk device performs one-track verification on the continuous large-capacity data such as the AV data after the write access to the large-capacity continuous data such as the AV data is completed. Since the write operation is performed in units, the SEEK time is minimized rather than verifying in units of a plurality of tracks after receiving a write command from a higher-level device and writing all the large-capacity continuous data such as the AV data to be written. And the most efficient verification is realized.

In the magnetic disk drive according to the invention of claim 13, the sequence control means automatically performs the verify operation without the intervention of a host device. Controlling the read access means and the verify means,
It is characterized by the following.

A disk access method for a magnetic disk drive according to a fourteenth aspect of the present invention is the disk access method for a magnetic disk drive according to the second aspect, wherein the verifying operation is performed without the intervention of a host device. It is performed automatically.

A magnetic disk drive control program recording medium according to claim 15 of the present application is the magnetic disk drive control program recording medium according to claim 3, wherein the operation of the control program recording medium is performed by the higher-level device in the verification. It is performed automatically without intervention.

As described above, the magnetic disk device, the disk access method thereof, and the disk access program recording medium for the magnetic disk device according to the thirteenth, fourteenth, and fifteenth aspects of the present invention use a magnetic disk as a recording medium. In a magnetic disk device which accesses the magnetic disk and reads / writes data of various types of information, transfer of large-capacity continuous data such as AV data requesting a certain band to / from an externally connected upper-level device. In this case, in normal use, after receiving a write command from a higher-level device and completing write access to large-capacity continuous data such as AV data, the magnetic disk device performs continuous verification of large-capacity continuous data such as AV data. AV data to be written upon receiving a write command from the host device Of after writing large sequential data, without interaction with the host device performs verification to large continuous data of the like AV data automatically written. The magnetic disk device ends the verification before the next write command comes from the host device. When the next write command is received from the host device, the magnetic disk is automatically verified by the next write command.

With this configuration, in normal use, the magnetic disk device receives a write command from the host device, terminates write access to large-capacity continuous data such as AV data, and then continuously writes the AV data. Verification performed on large-capacity continuous data such as data is automatically performed by a magnetic disk during a period between a write command and a write command for a magnetic disk from a higher-level device. It is possible to effectively use a period between the write commands.

According to a magnetic disk drive of the present invention, in the magnetic disk drive of the first aspect, the sequence control means detects an uncorrectable error by the verification means. In this case, the position and the result of the error on the magnetic disk and the result are registered in an error map of the non-volatile storage means, and the read map is referred to at the time of the read access by the read access means, and the erroneous sector is skipped for the read access. The nonvolatile storage means and the read access means are controlled so as to perform the following.

According to a disk access method of a magnetic disk drive according to a seventeenth aspect of the present invention, in the disk access method of the magnetic disk drive of the second aspect, an uncorrectable error is detected in the verification. In this case, the position and the result of the error on the magnetic disk and the result are registered in an error map of the non-volatile storage unit, and the read access is performed by referring to the error map at the time of read access and skipping the sector having the error. It is characterized by the following.

The control program recording medium for a magnetic disk device according to the invention of claim 18 of the present application is the control program recording medium for a magnetic disk device according to claim 3, wherein an uncorrectable error is detected in the verification. If detected, the position and the result of the error on the magnetic disk and the result are registered in an error map of a non-volatile storage unit, and the read access is performed by referring to the error map at the time of read access and skipping the sector having the error. It is characterized by the following.

As described above, the magnetic disk device, the disk access method, and the disk access program recording medium for the magnetic disk device according to the sixteenth, seventeenth, and eighteenth aspects of the present invention use a magnetic disk as a recording medium. In a magnetic disk device which accesses the magnetic disk and reads / writes data of various types of information, transfer of large-capacity continuous data such as AV data requesting a certain band to / from an externally connected upper-level device. In such a case, during normal use, the magnetic disk device receives a write command from a host device, terminates write access to large-capacity continuous data such as the AV data, and then continuously performs verification on the large-capacity continuous data. . When an uncorrectable error is detected by the verification, the physical position of the sector having the uncorrectable error on the magnetic disk and the error level are registered in an error map of the nonvolatile storage unit. afterwards,
When a read command is received from a host device and read access is made to large-capacity continuous data such as the AV data, the position of the erroneous sector is determined in advance by referring to the error map of the nonvolatile storage means. You can know. Therefore, as a read access to the sector having the uncorrectable error, it is possible to perform a skip that simply passes through without reading anything from the sector. The skip is performed, and normal read access is performed from a normal sector next to the sector having the uncorrectable error.

With this configuration, when a read command is received from a higher-level device and an uncorrectable error occurs in large-capacity continuous data such as the AV data in a range where the magnetic disk performs read access, a conventional method is used. Then, the upper device is notified of the occurrence of the error, and the read access operation is interrupted during the error processing of the upper device. Physical location of
By referring to the error map of the non-volatile storage means in which the level is registered, the position of the sector having the uncorrectable error can be known in advance. Since it is possible to skip without reading anything and to perform a skip operation to start reading again from the next normal sector, it is possible to maintain the continuity of the access at the time of the read access.

According to a magnetic disk drive of the present invention, in the magnetic disk drive according to the first aspect, the sequence control means detects an uncorrectable error by the verifying means. In this case, the position and the result of the error with respect to the data on the magnetic disk are registered in the error map of the non-volatile storage means, and rewriting is performed from the sector immediately after the sector in which the error occurred using the data in the buffer. Thus, the nonvolatile memory means, the error map registration means and the write access means are controlled.

In the disk access method for a magnetic disk drive according to the present invention, an uncorrectable error is detected in the verification. In this case, the position and the result of the error on the magnetic disk are registered in the error map of the nonvolatile storage means, and immediately after that, using the data in the buffer, rewriting is performed from the sector next to the sector having the error. It is characterized by the following.

A control program recording medium for a magnetic disk device according to the invention of claim 21 of the present application is the control program recording medium for a magnetic disk device according to claim 3, wherein an uncorrectable error is detected in the verification. If detected, the position of the error on the magnetic disk and the result are registered in the error map of the non-volatile storage means, and immediately after that, rewriting is performed from the next sector to the sector having the error using the buffer data. Perform.

As described above, the magnetic disk device, the disk access method thereof, and the disk access program recording medium for the magnetic disk device according to the nineteenth, twentieth and twenty-first aspects of the present invention use a magnetic disk as a recording medium. In a magnetic disk device which accesses the magnetic disk and reads / writes data of various types of information, transfer of large-capacity continuous data such as AV data requesting a certain band to / from an externally connected upper-level device. When performing the normal operation, the magnetic disk receives a write command from a host device, completes the write access to the large-capacity continuous data such as the AV data, and immediately thereafter executes the large-capacity continuous data such as the AV data. If an uncorrectable error is detected by verification of the A unrecoverable error and physical location and error level on the sector of the magnetic disk is registered in the error map of the nonvolatile memory means. The magnetic disk device shifts to a rewrite operation from the sector immediately after detecting the uncorrectable error. At this time, since the buffer holds the data to be written, the rewriting is performed using the data in the buffer. The buffer is a buffer having a capacity for storing data for several tracks.

With this configuration, if a write command is received from a host device and an uncorrectable error occurs in large-capacity continuous data such as AV data in the range written by the magnetic disk device, the error is written. Since detection is performed by verification performed on large-capacity continuous data such as the AV data after access is completed, the physical position and level on the magnetic disk can be registered in an error map of the nonvolatile storage means. At the time of read access, the sector having the uncorrectable error is skipped by referring to the error map of the non-volatile storage means, and rewriting is performed from the next normal sector, whereby the large-capacity continuous data such as the AV data is read. Continuity of access can be ensured.

A magnetic disk drive according to a twenty-second aspect of the present invention is the magnetic disk drive according to the first aspect, in which data exchanged between the magnetic disk and the host device is temporarily stored. The sequence control means, when an uncorrectable error is detected by the verification means, registers the position of the error on the magnetic disk and the result in an error map of the nonvolatile storage means, Refer to the error map at the time of rewriting, write the data of the erroneous sector to the replacement track of the magnetic disk using the data of the buffer, at the time of read access, pre-read the data of the replacement track to the buffer, The non-volatile sector is read so that the data of the erroneous sector is read directly from the buffer. Sex storing means, for controlling the error map registration means, write access means reads the access unit and the buffer, it is characterized in.

According to a disk access method of a magnetic disk drive according to the invention of claim 23 of the present application, in the disk access method of the magnetic disk drive of claim 2, an uncorrectable error is detected in the verification. In this case, the position and the result of the error on the magnetic disk are registered in the error map of the non-volatile storage means, and the error map is referred to at the time of rewriting, and the data of the erroneous sector is replaced using the data of the buffer. At the time of writing and reading access to the track, the data of the replacement track is pre-read into the buffer, and the data of the erroneous sector is read directly from the buffer.

The control program recording medium for a magnetic disk device according to the invention of claim 24 of the present application is the control program recording medium for a magnetic disk device according to claim 3, wherein an uncorrectable error is detected in the verification. If detected, the position on the error magnetic disk and the result are registered in the error map of the non-volatile storage means, the error map is referred to at the time of rewriting, and the data of the sector having the error is read using the data of the buffer. At the time of writing and reading access to the replacement track, the data of the replacement track is pre-read into the buffer at the time of access, and the data of the sector having the error is directly read from the buffer.

As described above, the magnetic disk device, the disk access method thereof, and the disk access program recording medium for the magnetic disk device according to the present invention use a magnetic disk as a recording medium. In a magnetic disk device which accesses the magnetic disk and reads / writes data of various types of information, transfer of large-capacity continuous data such as AV data requesting a certain band to / from an externally connected upper-level device. In this case, during normal use, the magnetic disk receives a write command from a higher-level device, completes write access to large-capacity continuous data such as the AV data, and performs verification on the large-capacity continuous data immediately after that. If an uncorrectable error is detected, the error The level of the sector physical location and error on the magnetic disk registering the error map of the nonvolatile memory means. After that, the magnetic disk device starts rewriting, and the data to be rewritten uses the data held in the buffer. For rewriting, the position of the sector having the uncorrectable error can be known in advance by referring to the error map of the non-volatile storage means. Therefore, the data to be rewritten to that sector is replaced with the replacement data in another track called an error replacement track. It is written as.
At the time of read access, the replacement data is previously read into the buffer, data is not read from the sector having the uncorrectable error, and the replacement data is read from the buffer at a timing at which the data should be read. Read directly.

According to this configuration, the A command written on the magnetic disk in response to the write command
If there is an uncorrectable error in the large-capacity continuous data such as V data, the write access to the large-capacity continuous data such as the AV data is terminated, and then the large-capacity large data such as the AV data is continuously performed. Since the continuous data is detected by the verification, the physical position of the uncorrectable errored sector on the magnetic disk and the error level are registered in an error map of the non-volatile storage means. Since the rewriting performed by the disk device refers to the error map of the nonvolatile storage means,
Since the position of the sector having the uncorrectable error is known in advance, the data to be written in the sector can be written as replacement data in the error replacement track.
At the time of read access, the replacement data is read out to the buffer first, and the data of the sector having the uncorrectable error is read from the buffer that has been read ahead at the timing when the data is originally read.
The required bandwidth and continuity of access are ensured without losing large-capacity continuous data such as AV data.

According to a magnetic disk drive of the present invention, in the magnetic disk drive of the present invention, data exchanged between the magnetic disk and the host device is temporarily stored. A buffer, wherein the sequence control means transfers the original data and its copy data at the time of data write access when transferring data to / from a higher-level device externally connected to the magnetic disk device. When an uncorrectable error is detected by writing to another track on the disk and verifying the original data, the position and the result of the error on the magnetic disk are registered in an error map of a nonvolatile storage means, and the error is registered. With reference to the map, the copy data of the data of the erroneous sector is pre-read into the buffer and read. The write access means, the non-volatile storage means, the error map registration means, the read access means, and the buffer are controlled so that the data of the erroneous sector is read directly from the buffer at the time of write access. .

According to a disk access method for a magnetic disk device according to claim 26 of the present application, the disk access method for a magnetic disk device according to claim 2 allows data to be exchanged with a higher-level device connected externally. When performing the transfer, write the original data and its copy data to another track at the time of write access of the data,
If an uncorrectable error is detected by verifying the original data, the position of the error on the magnetic disk and the result are registered in an error map of a non-volatile storage unit. The data of the erroneous sector is pre-read into the buffer using data, and the data of the erroneous sector is read directly from the buffer during read access.

The control program recording medium for a magnetic disk device according to claim 27 of the present application is the control program recording medium for magnetic disk device according to claim 3, wherein When transferring data at the time of writing access to the data, the original data and its copy data are written to another track, and the original data is verified,
When an uncorrectable error is detected, the position and the result of the error on the magnetic disk are registered in an error map of a non-volatile storage unit, and the error map is referred to when rewriting, and the error is The data of a certain sector is pre-read into a buffer, and the data of the erroneous sector is directly read from the buffer at the time of read access.

As described above, the magnetic disk device, the disk access method thereof, and the control program recording medium for the magnetic disk device according to the twenty-fifth, twenty-sixth and twenty-seventh aspects of the present invention use a magnetic disk as a recording medium. In a magnetic disk device that accesses the magnetic disk and reads / writes data of various types of information, large-capacity continuous data such as AV data requesting a certain band is transferred between the magnetic disk device and an externally connected host device. In this case, during normal use, the magnetic disk receives a write command from a higher-level device, performs write access to large-capacity continuous data such as the AV data, and also writes the copy data to another track. When an uncorrectable error is detected by verifying the original data of the large-capacity continuous data performed immediately after the end of the write access, a physical position on the magnetic disk of the sector having the uncorrectable error, The error level and the error level are registered in the error map of the nonvolatile storage unit. Before the magnetic disk receives a read command from a host device and performs read access to the original data of the large-capacity continuous data such as the AV data, the magnetic disk refers to the error map of the non-volatile means and performs the correction. The copy data corresponding to the data of the sector having a possible error is pre-read in the buffer. At the time of read access, referring to the error map of the non-volatile storage means, data is not read from the sector having the uncorrectable error, and the read-ahead copy in the buffer is performed at the timing when the data should be read. Read data directly.

According to this configuration, upon receiving the write command from the host device, the A
If there is an uncorrectable error in the original data of the large-capacity continuous data such as V data, the write access to the original data and its copy data is terminated, and then the verification is continuously performed on the original data. Is detected. The physical location of the detected uncorrectable errored sector on the magnetic disk and the error level are registered in the error map of the nonvolatile storage means. The copy data corresponding to the data of a possible erroneous sector can be read out first into a buffer. At the time of read access, referring to the error map of the non-volatile storage means, data is not read from the sector having the uncorrectable error, and the data in the buffer is read at the timing at which the data should be read. Without losing data, it is possible to secure the bandwidth required by the large-capacity continuous data such as the AV data and the continuity of access.

[0054]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS (Embodiment 1) The following corresponds to the invention described in claims 1, 2 and 3 of the present invention.
Embodiment 1 will be described with reference to FIGS. 1, 3, and 12. FIG. In FIG. 1, reference numeral 1 denotes a magnetic disk device, 2 denotes a magnetic disk as its recording medium, and 3 denotes a control circuit (a write access unit, a read access unit, a verify unit, an error map registration unit, a sequence unit) of a magnetic disk 2 such as an HDD controller. Control means), 4 is a sector formed on the magnetic disk 2, 5 is a non-volatile storage device (non-volatile storage means) composed of a semiconductor memory or the like,
Reference numeral 6 denotes a higher-level device such as a personal computer that issues a control command to the control circuit 3. Reference numerals 7 and 8 denote tracks on the magnetic disk 2. Reference numeral 9 temporarily stores data exchanged between the control circuit 3 and the higher-level device 6. Buffer 10 to be used in place of the track in which the error occurs,
Reference numeral 11 denotes replacement data to be replaced and recorded on the error replacement track 10, and reference numeral 12 denotes copy data corresponding to error sector data on the original data track.

Next, the operation of the magnetic disk drive according to the first embodiment will be described. In FIG. 1, a control circuit 3 includes a CPU (not shown) and a ROM (program recording medium) which are not shown, and a ROM for a magnetic disk device based on the disk access method of the magnetic disk device according to the first embodiment. The disk access program is recorded, and the CPU executes the disk access program for the magnetic disk device, whereby the magnetic disk device 1 performs the following operation.

That is, in response to an instruction from the host device 6, the control circuit 3 makes write access to large-capacity continuous data, such as AV data, to the magnetic disk 2 using the data held in the buffer 9. Immediately after the end of the write access, the control circuit 3 verifies the large-capacity continuous data such as the written AV data (see step S1 in FIG. 12). As a result, for example, if an uncorrectable error is detected in the sector 4 (see step S3), the control circuit 3
As shown in FIG. 3, the sector is designated as BADSECTOR 4
As B (see step S4), the physical position on the magnetic disk 2 and the error level are registered in an error map of the nonvolatile storage device 5 as nonvolatile storage means (see step S5).

As described above, in the magnetic disk device according to the first embodiment, the write data is verified immediately after the end of the write access, and when an uncorrectable error is detected, the error is detected. BADSE
CTOR 4B and the position on the disk and the error level are registered in the error map.
It is possible to detect uncorrectable errors and perform processing on those errors before read access during normal use, eliminating the rotational waiting time associated with error processing of higher-level devices that occurs with conventional magnetic disk devices. Becomes
The bandwidth required by large-capacity continuous data such as AV data and the continuity of access can be maintained.

(Embodiment 2) Next, Embodiment 2 corresponding to the inventions described in Claims 4, 5 and 6 of the present invention will be described with reference to FIGS. 1, 4 and 13. .
The configuration of the magnetic disk drive according to the second embodiment is the same as the configuration shown in FIG. 1 in the first embodiment. Next, the operation of the magnetic disk drive according to the second embodiment will be described. In FIG. 1, a control circuit 3 includes a CPU (not shown) and a ROM (program recording medium) not shown therein, and the ROM includes a second embodiment.
A disk access program for a magnetic disk device based on a disk access method for a magnetic disk device according to the present invention is recorded. I do.

In other words, in accordance with an instruction from the host device 6, the control circuit 3 performs write access to the magnetic disk 2 for large-capacity continuous data such as AV data, using the data held in the buffer. Immediately after the end of the write access, the control circuit 3 verifies the large-capacity continuous data such as the written AV data as shown in FIG. 4 (see step S12). .

By this verification, for example, not only an uncorrectable error but also a correctable ECC (E
When an error is detected (see step S13), the control circuit 3 determines the physical position of the sector 4 on the magnetic disk 2 for the uncorrectable error and the correctable ECC error. And the error level are registered in an error map of the nonvolatile storage device 5.

Here, the ECC error will be described. Even if an error on the magnetic disk 2 is a correctable ECC error at a certain point in time, the error becomes an uncorrectable error after repeated read access. there is a possibility.

In the second embodiment, when an uncorrectable error is detected, the BAD
The SECTOR 4B is set (see step S14), and the physical position of the sector 4 on the magnetic disk 2 and the error level are registered in an error map in the nonvolatile storage device 5 (see step S15). For a correctable ECC error such as
While keeping the OODSECTOR 4A, the error map registration of the physical position of the sector 4 on the magnetic disk 2 and the error level in the non-volatile storage device 5 is performed (see step S16), and the host device 6 Is notified that there is such a correctable ECC error (see step S17). The host device 6 receives this notification, determines what to do with the error, and issues a rewrite command before the error becomes uncorrectable.
Send to control circuit 3. The control circuit 3 receives the command and performs rewriting.

As described above, in the magnetic disk drive of the second embodiment, the uncorrectable error and the correctable EC
By notifying both the C error and the higher-level device 6 that there is such a correctable ECC error, the higher-level device determines what to do with the error and re-executes. A write command is sent to the control circuit 3 before the error cannot be corrected. Control circuit 3
Receives the instruction and rewrites it,
No error occurs at the time of read access during normal use, and it is possible to eliminate the rotation waiting time associated with the error processing of the higher-level device, which occurs in the conventional magnetic disk device,
The bandwidth required by large-capacity continuous data such as AV data and the continuity of access can be maintained.

(Embodiment 3) Next, Embodiment 3 corresponding to the invention described in claims 7, 8 and 9 of the present invention will be described with reference to FIG. 1, FIG. 5 and FIG. .
The configuration of the magnetic disk device according to the third embodiment is the same as the configuration shown in FIG. Next, the operation of the magnetic disk drive according to the third embodiment will be described. In FIG. 1, a control circuit 3 includes a CPU (not shown) and a ROM (program recording medium) which are not shown, and a ROM for a magnetic disk device based on the disk access method of the magnetic disk device according to the third embodiment. The disk access program is recorded, and the CPU executes the disk access program for the magnetic disk device, whereby the magnetic disk device 1 performs the following operation.

That is, in the third embodiment, after the write access of large-capacity continuous data such as AV data to the magnetic disk 2 is completed (see step S21 in FIG. 14), the host device 6 Before verifying the large-capacity continuous data such as the written AV data, the level of the ECC error is set in the control circuit 3 (see step S22) as shown in FIG.
Performs verification on the written data (see step S23), and when an ECC error is detected in sector 4 (see step S24), determines the level of the error and returns to the level set for itself. If it has reached (see step S25), the physical position of the sector 4 on the magnetic disk 2 and the error level are registered in an error map in the nonvolatile storage device 5 as nonvolatile storage means (step S26). reference). Thereafter, the control circuit 3 refers to the error map and automatically performs processing such as rewriting on the data of the sector 4 having an error.

As described above, in the magnetic disk device according to the third embodiment, after the write access of large-capacity continuous data such as AV data to the magnetic disk 2 is completed, the large-capacity continuous data such as the written AV data is terminated. If the ECC error level is set before verifying the data, if an ECC error is detected in sector 4 and the error level has reached the set error level,
An error map of the physical position of the sector 4 on the magnetic disk 2 and the error level is registered in the non-volatile storage device 5, and the sector 4 is automatically rewritten. Error occurrence during read access during normal use is eliminated, and rotation waiting time associated with error processing of a higher-level device, which occurs in a conventional magnetic disk drive, can be eliminated, and the bandwidth required for large-capacity continuous data such as AV data, Access continuity can be maintained.

(Embodiment 4) Next, claim 1 of the present invention will be described.
Embodiment 4 corresponding to the inventions described in 0, 11, and 12 will be described with reference to FIGS. 1, 6, and 15. FIG. The configuration of the magnetic disk device in the fourth embodiment is the same as the configuration shown in FIG. 1 in the first embodiment. Next, the operation of the magnetic disk drive according to the fourth embodiment will be described. In FIG. 1, the control circuit 3
Has a not-shown CPU and a ROM (program recording medium) therein, and the ROM stores a disk access program for the magnetic disk device based on the disk access method of the magnetic disk device according to the fourth embodiment. The CPU executes the disk access program for the magnetic disk device, whereby the magnetic disk device 1 performs the following operation.

A magnetic disk drive 1 according to the fourth embodiment
In FIG. 15, the control circuit 3 performs write access of large-capacity continuous data such as AV data to a certain track 7 on the magnetic disk 2 as shown in FIG.
In step S31), when the write access is completed, the control circuit 3 performs verification on the large-capacity continuous data of the same track 7 continuously immediately after that (see step S32). When the next write access is to the track 8, when the writing to the track 8 is completed, the verification of the large-capacity continuous data is performed on the track 8 continuously. That is,
The control circuit 3 performs verification on a track-by-track basis. In this case, the SEEK time, that is, the time required for moving the magnetic head, is only the SEEK time from track 7 to track 8 and is the minimum. Similarly, no matter how many tracks are accessed for writing, the most efficient verification can be performed.

As described above, in the magnetic disk drive of the fourth embodiment, the write access of the large-capacity continuous data is performed to one track 7 on the magnetic disk 2 and when this is completed, the continuous access is performed immediately after that. Then, verification is performed on the large-capacity continuous data in the same track 7, and when the next write access is track 8,
When the write operation is completed on the track 8, the verify operation is performed on a track-by-track basis such that the track 8 is continuously verified on the large-capacity continuous data. The time d required for the head to move to the target track is track 7
Thus, the verification can be performed with the highest efficiency regardless of the number of tracks to be accessed for writing. In the fourth embodiment, processes such as registration in the nonvolatile memory device performed after the verification are as follows.
Any of the processes of Embodiments 1 to 3 may be performed.

(Embodiment 5) Next, claim 1 of the present invention will be described.
A fifth embodiment corresponding to the inventions described in 3, 14, and 15 will be described with reference to FIGS. 1, 7, and 16. FIG. The configuration of the magnetic disk device according to the fifth embodiment is the same as the configuration shown in FIG. 1 in the first embodiment. Next, the operation of the magnetic disk drive according to the fifth embodiment will be described. In FIG. 1, the control circuit 3
Has a not-shown CPU and a ROM (program recording medium) therein. The ROM stores a disk access program for a magnetic disk device based on the disk access method of the magnetic disk device according to the fifth embodiment. The CPU executes the disk access program for the magnetic disk device, whereby the magnetic disk device 1 performs the following operation.

In the magnetic disk drive 1 according to the fifth embodiment, the control circuit 3 performs a verify operation on large-capacity continuous data such as AV data written on the magnetic disk 2 without exchanging with a higher-level device. Do it automatically.

First, as shown in FIG. 7, the magnetic disk device 1 receives a write command from the host device 6 (FIG. 1).
No. 6, step S41), and data write access is performed (see step S42). Then, before the write command from the next higher-level device 6, the control circuit 3
Verify is automatically performed on the written data (see step S43).

Upon receiving the next write command, write access is performed, and until the next write command comes, the control circuit 3 automatically verifies the written data.

As described above, in the magnetic disk drive of the fifth embodiment, the verification performed on large-capacity continuous data such as AV data written on the magnetic disk 2 can be performed without any exchange with the host device. 3 automatically performs the AV command while effectively utilizing the time between the write command issued from the host device and the write command.
Verification can be performed on large-capacity continuous data such as data. In the fifth embodiment, processes such as registration in the nonvolatile memory device performed after the verification are performed as follows.
Any of the processes of Embodiments 1 to 3 may be performed.

(Embodiment 6) Next, claim 1 of the present invention will be described.
A sixth embodiment corresponding to the inventions described in 6, 17, and 18 will be described with reference to FIGS. 1, 8, and 17. FIG. The configuration of the magnetic disk drive in the sixth embodiment is the same as the configuration shown in FIG. 1 in the first embodiment. In FIG. 1, a control circuit 3 includes a CPU (not shown) and a ROM (program recording medium) which are not shown. The disk access program is recorded, and the CPU executes the disk access program for the magnetic disk device, whereby the magnetic disk device 1 performs the following operation.

In the magnetic disk device 1 of the sixth embodiment, large-capacity continuous data such as AV data written on the magnetic disk 2 is read (step S51 in FIG. 17).
Control circuit 3 performs verification (step S5).
2), for example, if an uncorrectable error is detected in sector 4 (see step S53), sector 4 is BAD
As SECTOR 4B (see step S54), the physical position on the magnetic disk 2 and the error level are registered in the error map of the nonvolatile storage device 5 as nonvolatile storage means (see step S55).

When a read command is received from the upper level device 6 (see step S56), the control circuit 3 refers to the error map of the nonvolatile memory device 5 and places an error in the range of data to be read, as shown in FIG. It is determined whether or not there is a sector 4 having a pattern (see step S57).
When it is necessary to perform read access to the sector 4 having an error (see step S58), control is performed so that skip operation is performed without passing data from the sector 4 without reading data. Then, reading is started again from the data of the next sector (see step S59), and is continued until the data to be read ends (see step S60).

As described above, in the magnetic disk drive of the sixth embodiment, when a large capacity continuous data such as AV data written on the magnetic disk is verified and an uncorrectable error is detected, BADS for sector 4
When ECTOR 4B is set, an error map is registered, and when a read command is received from the host device 6,
By referring to the error map, it is determined whether or not there is an error sector in the range of data to be read. When reading access to the error sector is to be performed, skipping operation is performed without reading data from sector 4 and passing. In this case, reading is started again from the next sector data, so even if there is a sector with an uncorrectable error in the data to be read, there is no error Continuity can be ensured. In the sixth embodiment,
The processing such as registration in the non-volatile storage device performed after the verification may be the processing of any of the first and second embodiments.

(Embodiment 7) Next, claim 1 of the present invention will be described.
A seventh embodiment corresponding to the inventions described in 9, 20, and 21 will be described with reference to FIG. 1, FIG. 9 and FIG. The configuration of the magnetic disk drive according to the seventh embodiment is the same as the configuration shown in FIG. 1 in the first embodiment. In FIG. 1, a control circuit 3 includes a CPU (not shown) and a ROM (program recording medium) which are not shown, and a ROM for a magnetic disk device based on the disk access method of the magnetic disk device according to the seventh embodiment. The disk access program is recorded, and the CPU executes the disk access program for the magnetic disk device, whereby the magnetic disk device 1 performs the following operation.

In the magnetic disk drive 1 according to the seventh embodiment, large-capacity continuous data such as AV data written on the magnetic disk 2 is read (step S61 in FIG. 18).
The control circuit 3 performs verification (step S6).
2), for example, if an uncorrectable error is detected in the sector 4 (see step S63), the sector 4 is BAD
As SECTOR 4B (see step S64), the physical position on the magnetic disk 2 and the error level are registered in an error map in the nonvolatile storage device 5 as nonvolatile storage means (see step S65).

The control circuit 3 controls to skip the sector 4 immediately after that, as shown in FIG.
66), after requesting the buffer 9 to send data for rewriting (see step S67), using the data sent from the buffer 9 to create an error map from the next skipped sector. Rewriting is performed while controlling not to write to sector 4 by referring to the data (see step S68).

As described above, in the magnetic disk drive of the seventh embodiment, when an uncorrectable error is detected, the sector 4 is set to BADSECTOR 4B, an error map is registered, and the sector 4 is skipped immediately after that. And a request to transfer data for rewriting to the buffer 9, and by using the data sent from the buffer 9, control is performed so as not to write to the sector 4 from the next sector skipped. However, since rewriting is performed, even if there is an uncorrectable error during normal use, large-capacity continuous data such as AV data can be maintained without loss and continuous access can be maintained. In the seventh embodiment, the processing such as registration in the non-volatile storage device performed after the verification may be the processing of any of the first and second embodiments.

(Embodiment 8) Next, claim 2 of the present invention will be described.
An eighth embodiment corresponding to the inventions described in 2, 23, and 24 will be described with reference to FIGS. 1, 10, and 19. FIG. The configuration of the magnetic disk drive according to the eighth embodiment is the same as the configuration shown in FIG. 1 in the first embodiment. In FIG. 1, a control circuit 3 includes a CPU (not shown) and a ROM (program recording medium) which are not shown, and a ROM for a magnetic disk drive based on the disk access method of the magnetic disk drive according to the eighth embodiment. The disk access program is recorded, and the CPU executes the disk access program for the magnetic disk device, whereby the magnetic disk device 1 performs the following operation.

In the magnetic disk drive 1 of the eighth embodiment, large-capacity continuous data such as AV data written on the magnetic disk 2 is read (step S71 in FIG. 19).
10), as shown in FIG. 10, the control circuit 3 performs verification (see step S72). For example, when an uncorrectable error is detected in the sector 4 (see step S73), the sector 4 is set as BADSECTOR 4B. (See step S74), the physical position on the magnetic disk 2 and the error level are registered in the error map of the nonvolatile storage device 5 as nonvolatile storage means (step S74).
75). Thereafter, the control circuit 3 proceeds to rewriting while referring to the error map (see step S76).
Data to be written to sector 4 is error replacement track 1
0 is written as replacement data 11 (see step S77).

Next, upon receiving a read command from the host device 6 (see step S78), the control circuit 3 first refers to the error map to determine whether there is a sector of BADSECTOR 4B within the range of data to be read. It is determined whether or not such a sector 4 exists (see step S80).
The control circuit 3 issues an instruction to pre-read the replacement data 11 into the buffer 9 (see step S81). In accordance with the prefetch instruction, the prefetched data is arranged in the buffer 9 in a normal order together with data from other normal sectors (see step S82). Then, the host device 6 reads the data in the buffer 9 until the end as usual (see steps S83 and S84). As a result, the occurrence of an error at the time of reading access to continuous data such as AV data is eliminated.
The bandwidth required by large-capacity continuous data such as V data and the continuity of access can be ensured.

As described above, in the magnetic disk device of the eighth embodiment, as described above, the verification is performed on the large-capacity continuous data written on the magnetic disk 2, and an uncorrectable error is detected. In this case, the sector 4 is BAD
As SECTOR 4B, its physical position and error level are registered in an error map, and then rewriting is performed with reference to the error map. Data to be written in the error sector 4 is written as replacement data 11 in the error replacement track 10. When a read command is received from the upper level device 6, it is determined whether or not there is a sector of BADSECTOR 4B in the range of data to be read by referring to the error map.
The replacement data 11 is instructed to be prefetched into the buffer 9, and the prefetched data is arranged in a normal order together with data from other normal sectors in the buffer 9. Since the data in the buffer 9 is read as described above, the occurrence of an error at the time of reading access to continuous data such as AV data is eliminated, the rotation waiting time accompanying the processing for the error of the higher-level device is eliminated, and the large-capacity continuous The effect that the band requested by the data and the continuity of the access can be secured is obtained. Note that, in the eighth embodiment, the processing such as registration in the non-volatile storage device performed after the verification may be the processing of any of the first and second embodiments.

(Embodiment 9) Next, Embodiment 2 of the present invention will be described.
A ninth embodiment corresponding to the inventions described in 5, 26, and 27 will be described with reference to FIGS. The configuration of the device according to the ninth embodiment is the same as that of the first embodiment.
Is the same as that shown in FIG. In FIG. 1, a control circuit 3 includes a CPU (not shown) and a ROM
(Program recording medium).
In this embodiment, a disk access program for a magnetic disk device based on the disk access method for a magnetic disk device according to the ninth embodiment is recorded, and the CPU executes the disk access program for a magnetic disk device to 1 performs the following operation.

In the magnetic disk device 1 according to the ninth embodiment, after the original data of large-capacity continuous data such as AV data to be written on the magnetic disk 2 is written (see step S91 in FIG. 20), FIG. As shown in (1), the control circuit 3 issues an instruction to write the copy data to another track (see step S92). In FIG. 11, track 7 is original data and track 8 is copy data. After the writing of both data is completed, the control circuit 3 verifies the track 7 (see step S93). If an uncorrectable error is found in the sector 4 (see step S94), the sector 4 is set to BADSE.
As the CTOR 4B (see step S95), the physical position on the magnetic disk 2 and the error level are registered in the error map of the nonvolatile storage device 5 as nonvolatile storage means (see step S96).

Next, upon receiving a read command from the host device 6 (see step S97), the control circuit 3 refers to the error map and determines whether or not the range of data to be read includes an erroneous sector 4. Judgment (see step S98), and if included (step S98)
Before the read access, an instruction to pre-read the copy data 12 corresponding to the data of the sector 4 on the track 7 into the buffer 9 is issued (see step S100). The copy data 12 corresponding to the data of the sector 4 on the track 8 is arranged in the original order during the pre-reading (see step S101). And the host device 6
Reads the data in the buffer 9 until the end as usual (see steps S102 and S103).

As described above, in the magnetic disk device of the ninth embodiment, after the original data of large-capacity continuous data such as AV data to be written on the magnetic disk 2 is written, the control circuit 3 copies the copy data. An instruction to write to another track is issued, and after the writing of both data is completed, the control circuit 3 verifies the track 7 of the original data, and if there is an uncorrectable error in the sector 4, the sector 4 is replaced with the BADSECTOR 4B. When the error map of the physical position and the error level is registered, and a read command is received from the higher-level device 6, the control circuit 3 refers to the error map and finds a sector having an error within the range of data to be read. 4 is included, and if so, the cell on track 8 is read before read access. An instruction to pre-read the copy data 12 corresponding to the data of the sector 4 into the buffer 9 is issued, and an instruction to pre-read the copy data 12 corresponding to the data of the sector 4 on the track 8 to the buffer 9 is issued. Since the copy data 12 corresponding to the data of the sector 4 on the top 8 is arranged in the original order during the pre-reading, and the host device 6 reads the data from the buffer 9 as usual, the AV data etc. In the read access to the large-capacity continuous data, no error occurs, and the required bandwidth and continuity of access can be secured.

In the ninth embodiment, the processing such as registration in the nonvolatile memory device performed after the verification may be the same as the processing in the first or second embodiment.

[0092]

As described above, according to the magnetic disk device, the disk access method, and the disk access program recording medium for the magnetic disk device according to the first, second and third aspects of the present invention, Immediately after the end of write access to large-capacity continuous data such as AV data, verification is performed on large-capacity continuous data such as written AV data. If an uncorrectable error is detected in a sector, for example, control is performed. The circuit uses the sector as a BADSECTOR to register the physical position on the magnetic disk and the error level in the error map in the non-volatile storage device. It is possible to detect uncorrectable errors registered in the map and process them Therefore, it is possible to eliminate the rotation waiting time associated with the error processing of the higher-level device, which occurs in the conventional magnetic disk device, and it is possible to maintain the bandwidth required for large-capacity continuous data such as AV data and the continuity of access. it can.

According to the magnetic disk device, the disk access method and the disk access program recording medium for the magnetic disk device according to the fourth, fifth and sixth aspects of the present invention, it is possible to store AV data or the like on the magnetic disk. After the end of the large-capacity continuous write access, verification is performed on large-capacity continuous data such as continuously written AV data, and for example, a correctable ECC error other than an uncorrectable error is detected in a sector. In this case, the physical position of the sector on the magnetic disk and the error level are registered in an error map in the nonvolatile storage device, and after the access is completed, the host device is notified that an ECC error has occurred. Therefore, even if an error on the magnetic disk is an ECC error that can be corrected at a certain point in time, While repeating the write access,
Although an uncorrectable error may occur, the host device receives a notification that there is a correctable ECC error, determines what to do with the error, and before the error becomes uncorrectable. , A rewrite command is sent to the control circuit, and the control circuit receives the command and performs rewrite,
Eliminating the occurrence of errors at the time of read access eliminates the rotation waiting time associated with the error processing of the higher-level device, which occurs in the conventional magnetic disk device. The bandwidth required for large-capacity continuous data such as AV data, and the continuity of access Can be kept.

According to the magnetic disk device, the disk access method, and the disk access program recording medium for the magnetic disk device according to the seventh, eighth, and ninth aspects of the present invention, the magnetic disk has a large capacity for storing data and the like. After ending the continuous data write access, the host device
Before the control circuit verifies large-capacity continuous data such as written AV data, the control circuit sends an EC
Set the level of C error. The control circuit verifies the written data. When an ECC error is detected in the sector, the control circuit determines the level of the error. The physical location on the magnetic disk and the level of error
Register an error map in the nonvolatile storage device. afterwards,
The control circuit refers to the error map and automatically performs processing such as rewriting on the data in the sector. As a result, the bandwidth required by large-capacity continuous data such as AV data,
Reliability can be maintained.

Further, claims 10, 11 and 1 of the present invention
According to the magnetic disk device, the disk access method, and the disk access program recording medium for the magnetic disk device described in 2, the magnetic disk is used as a recording medium, and the magnetic disk is accessed to read and write various information data. When transferring large-capacity continuous data such as AV data that requires a certain band between a magnetic disk device and a higher-level device connected to the outside, a write command is received from the higher-level device during normal use. After a write access to the large-capacity continuous data is completed, when verifying the large-capacity continuous data continuously, a certain one on the magnetic disk is
Immediately after the writing of the large-capacity continuous data to one track is completed, the verify is performed on the same track. Thereafter, the head of the magnetic disk drive SEEKs to another track on the magnetic disk, Verify is performed on the same track immediately after write access of large-capacity continuous data to the track is completed. In other words, verify is performed on a track-by-track basis after write access is completed. Therefore, the SEEK time can be minimized as compared with a case where a write command is received from a higher-level device and all large-capacity continuous data such as AV data to be written is completely written, and then verification is performed in units of a plurality of tracks. , The most efficient verification can be performed.

Further, claims 13, 14 and 1 of the present invention
According to the magnetic disk device, the disk access method, and the disk access program recording medium for the magnetic disk device described in 5, the magnetic disk is used as the recording medium, and the magnetic disk is accessed to read and write various information data. When transferring large-capacity continuous data such as AV data that requires a certain bandwidth between a magnetic disk device and a higher-level device connected to the outside, a write command is received from the higher-level device during normal use. After the write access to the large-capacity continuous data is completed, the magnetic disk device receives a write command from a higher-level device and writes the large-capacity continuous data to be written in the continuous verification of the large-capacity continuous data. After that, automatically without any interaction with the host device Verify is performed on the written large-capacity continuous data, and the magnetic disk device finishes the verification before the next write command comes from a higher-level device, and when the next write command is received from the higher-level device, Similarly, since the magnetic disk performs the automatic verification until the next write command, during normal use, the magnetic disk device receives a write command from a higher-level device and performs a write operation on the large-capacity continuous data. After the write access is completed, a verify operation for continuously writing the large-capacity continuous data is performed in a period between a write command to the magnetic disk from a host device and the write command. Automatically occurs, which occurs in the conventional magnetic disk drive, The duration of the period included command and the write command can be effectively used.

Further, claims 16, 17 and 18 of the present invention.
According to the magnetic disk device, the disk access method, and the disk access program recording medium for the magnetic disk device, a magnetic disk is used as a recording medium, and the magnetic disk is accessed to read and write various information data. In a magnetic disk device, when transferring large-capacity continuous data such as AV data requesting a certain band between the magnetic disk device and an external device connected to the outside, the magnetic disk device is configured such that during normal use, the magnetic disk device After receiving a write command from the controller and ending write access to the large-capacity continuous data, continuously performing verification on the large-capacity continuous data, and if an uncorrectable error is detected by the verification, the The physical location of the erroneous sector on the magnetic disk, The error level is registered in the error map of the non-volatile storage means. After that, when a read command is received from a higher-level device and read access is made to large-capacity continuous data such as the AV data, the non-volatile storage means is used. By referring to the error map, the location of the erroneous sector can be known in advance. Therefore, as read access to the uncorrectable erroneous sector, what Can be skipped by just passing without reading, and normal read access can be performed from the next normal sector after the sector having the uncorrectable error. Receiving a read command of large-capacity continuous data such as the AV data in a range where the magnetic disk performs read access. However, if there is an uncorrectable error, the conventional method notifies the host device of the occurrence of the error, and the read access operation was interrupted during the error processing of the host device. By referring to the error map of the non-volatile storage means in which the physical position of the sector having the uncorrectable error on the magnetic disk and its level are registered, the sector having the uncorrectable error is determined in advance. When reading and accessing the sector, it is possible to skip without reading anything and perform a skip operation to start reading again from the next normal sector. The continuity of access to the operation can be maintained.

Further, claims 19, 20 and 21 of the present invention
According to the magnetic disk device, the disk access method, and the disk access program recording medium for the magnetic disk device, a magnetic disk is used as a recording medium, and the magnetic disk is accessed to read and write various information data. In a magnetic disk device, when transferring large-capacity continuous data such as AV data requesting a certain bandwidth between the magnetic disk device and a host device connected to the outside, the magnetic disk is written from the host device during normal use. Upon receiving a command, the write access to the large-capacity continuous data such as the AV data is completed, and immediately after that, an uncorrectable error is detected by verification of the large-capacity continuous data such as the AV data, Physical sector on magnetic disk of sectors with uncorrectable errors The position and the level of the error are registered in an error map of the nonvolatile storage means, and the magnetic disk device starts a rewrite operation from a sector immediately after detecting the uncorrectable error. At this time, the buffer stores the data to be written. , The rewriting is performed using the data in the buffer. Here, the buffer is a buffer having a capacity for storing data for several tracks, When an uncorrectable error occurs in large-capacity continuous data such as AV data in the range written by the magnetic disk device upon receiving a command, the error is detected after the completion of the write access. The physical position and level on the magnetic disk are detected by the verification performed on the By reading the error map of the non-volatile storage means at the time of read access, skipping the sector having the uncorrectable error and performing rewriting from the next normal sector. The continuity of access to large-capacity continuous data such as the AV data can be ensured.

Further, claims 22, 23 and 24 of the present invention.
According to the magnetic disk device described in (1), a magnetic disk is used as a recording medium, and the magnetic disk device accesses the magnetic disk and reads / writes data of various information. so,
When transferring large-capacity continuous data such as AV data that requires a certain bandwidth, during normal use, the magnetic disk receives a write command from a host device, and the write access of the large-capacity continuous data such as the AV data ends. If an uncorrectable error is detected by the verification of the large-capacity continuous data performed immediately after that, the physical position of the sector having the uncorrectable error on the magnetic disk and the error level Is registered in the error map of the non-volatile storage means, and thereafter, the magnetic disk device shifts to rewriting, but the data to be rewritten uses the data held in the buffer. By referring to the map, the position of the sector having the uncorrectable error can be known in advance, and the The data to be written is written as replacement data in another track called an error replacement track, and at the time of read access, the replacement data is read out to the buffer first, and data is read from the sector having the uncorrectable error. The replacement data is read directly from the buffer at the timing at which the data should be read, so that the write command from the host device is received and the size of the AV data or the like written on the magnetic disk is received. If there is an uncorrectable error in the continuous capacity data, the AV
After the end of write access to large-capacity continuous data such as data, the magnetic disk can be detected by continuous verification of large-capacity continuous data such as the AV data, and the sector of the sector having an uncorrectable error can be detected. The physical position above and the level of the error are registered in the error map of the non-volatile storage means, and then the rewriting performed by the magnetic disk device refers to the error map of the non-volatile storage means, Since the position of the sector having the uncorrectable error is known in advance, the data to be written in the sector is written as replacement data in an error replacement track, and the read data is read out to a buffer at the time of read access. ,
Since the data of the sector having the uncorrectable error is read from the buffer which has been read ahead at the timing when the data is originally read, large-capacity continuous data such as AV data is not lost. Required bandwidth and access continuity can be secured.

Further, claims 25, 26 and 2 of the present invention
7. According to the magnetic disk device, the disk access method, and the disk access program recording medium for the magnetic disk device described in 7, the magnetic disk is used as a recording medium, and the magnetic disk is accessed to read and write various information data. When transferring large-capacity continuous data such as AV data requesting a certain band between the external device and a higher-level device connected to the external device, during normal use, the magnetic disk is transferred from the higher-level device. Upon receiving a write command, write access is made to the large-capacity continuous data, and the copy data is also written to another track, and by performing verification on the original data of the large-capacity continuous data performed immediately after the end of the write access, Uncorrectable error detected If registers the physical position on the magnetic disk of the sector with the uncorrectable error, the level of error in the error map of the nonvolatile memory means,
The magnetic disk receives a read command from a higher-level device and, before performing read access to the original data of the large-capacity continuous data, refers to an error map of the non-volatile means and checks the uncorrectable error. The copy data corresponding to the data of a certain sector is pre-read in a buffer, and at the time of read access, the error map of the nonvolatile storage means is referred to, and the data is not read from the sector having the uncorrectable error. ,
At the timing at which the data should be read, the read-ahead copy data in the buffer is directly read, and in response to a write command from a higher-level device,
When there is an uncorrectable error in the original data of the large-capacity continuous data written on the magnetic disk,
After the write access to the original data and its copy data is completed, the uncorrectable error is detected by successive verification of the original data, and the detected sector having the uncorrectable error is detected. The physical position on the magnetic disk and the error level are registered in the error map of the non-volatile storage means, and by referring to the error map, the copy data corresponding to the uncorrectable error data is read. The buffer can be read first, and at the time of read access, data is not read from the sector having the uncorrectable error by referring to the error map of the nonvolatile storage means, and at the timing when the data should be read, By reading the data in the buffer, Without chipping, the requested bandwidth of the high-capacity continuous data such as the AV data, it is possible to ensure the continuity of the access.

As described above, according to the magnetic disk device of the present invention, the disk access method thereof, and the disk access program recording medium for the magnetic disk device,
Even if there is an uncorrectable error in large-capacity continuous data such as data, it can be detected during normal use by verification performed immediately after the end of write access. Further, by registering the detected result in an error map, it is possible to perform processing for errors such as rewriting before read access, and to eliminate the occurrence of errors at the time of read access.
Therefore, the rotation waiting time associated with the error processing of the host device is eliminated, and the band required for large-capacity continuous data such as AV data and the continuity of access can be secured.

[Brief description of the drawings]

FIG. 1 is an explanatory diagram of the configuration and operation of a magnetic disk drive according to Embodiments 1 to 9 of the present invention.

FIG. 2 is a schematic diagram when an error occurs during read access in a conventional magnetic disk device.

FIG. 3 is an explanatory diagram of an operation in the magnetic disk device according to the first embodiment of the present invention.

FIG. 4 is an explanatory diagram of an operation in the magnetic disk device according to the second embodiment of the present invention.

FIG. 5 is an explanatory diagram of an operation in the magnetic disk drive according to the third embodiment of the present invention.

FIG. 6 is an explanatory diagram of an operation in the magnetic disk drive according to the fourth embodiment of the present invention.

FIG. 7 is an explanatory diagram of an operation in the magnetic disk drive according to the fifth embodiment of the present invention.

FIG. 8 is an explanatory diagram of an operation in the magnetic disk drive according to the sixth embodiment of the present invention.

FIG. 9 is an explanatory diagram of an operation in the magnetic disk drive according to the seventh embodiment of the present invention.

FIG. 10 is an explanatory diagram of an operation in the magnetic disk drive according to the eighth embodiment of the present invention.

FIG. 11 is an explanatory diagram of an operation in the magnetic disk drive according to the ninth embodiment of the present invention.

FIG. 12 is a flowchart of an operation in the magnetic disk drive according to the first embodiment of the present invention.

FIG. 13 is a flowchart of an operation in the magnetic disk drive according to the second embodiment of the present invention.

FIG. 14 is a flowchart of an operation in the magnetic disk drive according to the third embodiment of the present invention.

FIG. 15 is a flowchart of an operation in the magnetic disk drive according to the fourth embodiment of the present invention.

FIG. 16 is a flowchart of an operation in the magnetic disk drive according to the fifth embodiment of the present invention.

FIG. 17 is a flowchart of an operation in the magnetic disk drive according to the sixth embodiment of the present invention.

FIG. 18 is a flowchart of an operation in the magnetic disk drive according to the seventh embodiment of the present invention.

FIG. 19 is a flowchart of an operation in the magnetic disk drive according to the eighth embodiment of the present invention.

FIG. 20 is a flowchart of an operation in the magnetic disk drive according to the ninth embodiment of the present invention;

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Magnetic disk device 2 Magnetic disk 3 Control circuit 4 Sector 5 Non-volatile storage device 6 High-order device 7, 8 Track 9 Buffer 10 Error replacement track 11 Replacement data 12 Copy data

────────────────────────────────────────────────── ───

[Procedure amendment]

[Submission date] May 25, 2001 (2001.5.2)
5)

[Procedure amendment 1]

[Document name to be amended] Statement

[Correction target item name] 0060

[Correction method] Change

[Correction contents]

By this verification, for example, not only an uncorrectable error but also a correctable ECC ( E
If an error is detected (see step S13), the control circuit 3 determines the physical position of the sector 4 on the magnetic disk 2 for the uncorrectable error and the correctable ECC error. The error level is registered in the error map of the nonvolatile storage device 5.

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) G11B 20/18 572 G11B 20/18 572F

Claims (27)

[Claims] 1. A magnetic disk device that uses a magnetic disk as a recording medium and accesses the magnetic disk to read and write various data, wherein: write access means for performing data write access to the magnetic disk; Read access means for performing data read access to the magnetic disk; verify means for verifying data on the magnetic disk; non-volatile storage means for registering an error map of the magnetic disk; detected by the verify Error map registration means for registering the position and the result of the error on the magnetic disk in an error map of the nonvolatile storage means, and transferring data between the magnetic disk device and a higher-level device connected to the outside thereof. When performing, the data is Verifying the data on the magnetic disk after terminating the data write access,
When an uncorrectable error is detected, the write access unit, the verify unit, the read access unit, and the write access unit register the position and the result of the error on the magnetic disk in an error map of the nonvolatile storage unit. And a sequence control means for controlling the error map registration means. 2. A disk access method for a magnetic disk device in which a magnetic disk is used as a recording medium and accesses the magnetic disk to read and write various data. When performing the data transfer, after verifying the data on the magnetic disk after completing the data write access to the magnetic disk, if an uncorrectable error is detected, Registering the position and the result of the error in an error map of a non-volatile storage means. 3. A recording medium in which a magnetic disk is used as a recording medium, and a control program for a magnetic disk device for reading and writing various data by accessing the magnetic disk is provided. When transferring data between the above, after completing the data write access to the magnetic disk, verify the data to the magnetic disk, if an uncorrectable error is detected, A disk access control program recording medium for a magnetic disk device, wherein a position and a result of the error on the magnetic disk are registered in an error map of a nonvolatile storage unit. 4. The magnetic disk drive according to claim 1, wherein the sequence control unit is configured to correct the ECC which can be corrected by the verification unit in addition to the uncorrectable error.
Even when an error is detected, the nonvolatile memory means and the error map registration means are controlled to register the position and the result of the error on the magnetic disk in an error map of the nonvolatile memory means, and A magnetic disk device, after completion of the verification, notifying the host device that an ECC error has occurred. 5. The disk access method of a magnetic disk drive according to claim 2, wherein a correctable ECC error other than the uncorrectable error is detected by the verification even when the error is detected. A disk access method for a magnetic disk device, comprising: registering a position on a disk and a result in an error map of a non-volatile storage unit, and notifying an upper-level device that an ECC error has occurred after completion of the verification. 6. The control program recording medium for a magnetic disk device according to claim 3, wherein the correctable ECC error other than the uncorrectable error is detected by the verification even if the error is detected. Registering the position on the magnetic disk and the result in an error map of the non-volatile storage means, and notifying an upper-level device that an ECC error has occurred after completion of the verification; Program recording medium. 7. The magnetic disk drive according to claim 1, wherein the sequence control means registers an ECC in an error map by the higher-level device before performing the verification.
When an error level is set, and when the ECC error detected by the verifying means has reached the set level, the sequence control means stores the position of the error on the magnetic disk and the result in the nonvolatile memory. The non-volatile storage unit, the error map registration unit, and the write access unit are controlled so as to be registered in an error map of a storage unit and to automatically perform processing such as rewriting on the magnetic disk. Magnetic disk device. 8. The disk access method for a magnetic disk drive according to claim 2, wherein a level of an ECC error to be registered in an error map is set by a higher-level device before performing the verifying, and the ECC detected by the verifying is set. When the error has reached the set level, the position and the result of the error on the magnetic disk are registered in an error map of the nonvolatile storage means, and processing such as automatically rewriting the magnetic disk is performed. A disk access method for a magnetic disk drive. 9. The control program recording medium for a magnetic disk device according to claim 3, wherein a level of an ECC error to be registered in an error map is set by a host device before the verification is performed, and the level is detected by the verification. When the ECC error has reached the set level, the position and the result of the error on the magnetic disk are registered in an error map of the non-volatile storage means, and processing such as automatically rewriting the magnetic disk is performed. A disk access control program recording medium for a magnetic disk device. 10. The magnetic disk drive according to claim 1, wherein the sequence control unit controls the read access unit and the verification unit so as to perform the verification operation in units of one track. Magnetic disk drive. 11. The disk access method of a disk device according to claim 2, wherein the verifying operation is performed for each track. 12. The disk access control program recording medium for a magnetic disk device according to claim 3, wherein the verifying operation is performed for each track. Medium. 13. The magnetic disk drive according to claim 1, wherein the sequence control unit controls the read access unit and the verify unit so that the verify operation is automatically performed without intervention of a host device. A magnetic disk drive characterized by the above-mentioned. 14. The disk access method for a magnetic disk device according to claim 2, wherein the verifying operation is performed automatically without the intervention of a host device. 15. The disk access control for a magnetic disk device according to claim 3, wherein in the verification, the operation is automatically performed without the intervention of a host device. Program recording medium. 16. The magnetic disk drive according to claim 1, wherein, when an uncorrectable error is detected by the verifying unit, the sequence control unit stores the position of the error on the magnetic disk and the result in a nonvolatile manner. The non-volatile storage means and the read access means so that the read access is performed by referring to the error map at the time of read access by the read access means and skipping the sector having the error. A magnetic disk drive. 17. The disk access method for a magnetic disk device according to claim 2, wherein in the verification, when an uncorrectable error is detected, a position and a result of the error on the magnetic disk and a result are stored in a nonvolatile storage unit. And performing a read access by skipping the erroneous sector by referring to the error map at the time of read access. 18. The control program recording medium for a magnetic disk device according to claim 3, wherein in the verification, when an uncorrectable error is detected, a position and a result of the error on the magnetic disk are stored in a nonvolatile memory. A disk access control program recording medium for a magnetic disk drive, wherein the disk access control program is registered in an error map of the means, refers to the error map at the time of read access, and skips the erroneous sector to perform read access. 19. The magnetic disk drive according to claim 1, wherein, when an uncorrectable error is detected by the verifying unit, the sequence control unit determines a position on the magnetic disk of the error with respect to data and a result. Is registered in the error map of the non-volatile storage means, and the non-volatile storage means, the error map registration means and the error map registration means are rewritten from the sector immediately after the errored sector using the data of the buffer. A magnetic disk drive for controlling write access means. 20. The disk access method for a magnetic disk device according to claim 2, wherein in the verification, when an uncorrectable error is detected, the position of the error on the magnetic disk and the result are stored in a nonvolatile storage means. And rewriting from the next sector after the erroneous sector using the data in the buffer immediately after registration in the error map. 21. The control program recording medium for a magnetic disk device according to claim 3, wherein in the verification, when an uncorrectable error is detected, a position and a result of the error on the magnetic disk are stored in a non-volatile memory. A disk access control program recording medium for a magnetic disk device, wherein the information is registered in an error map of the means, and immediately after that, rewriting is performed from a sector next to the sector having the error using data in the buffer. 22. The magnetic disk drive according to claim 1, further comprising: a buffer for temporarily storing data exchanged between the magnetic disk and the host device, wherein the sequence control unit includes the verification unit. When an uncorrectable error is detected, the position and the result of the error on the magnetic disk are registered in an error map of the nonvolatile storage means, and the error map is referred to at the time of rewriting, and the data in the buffer is read. The data of the erroneous sector is written into a replacement track of the magnetic disk by using, and at the time of read access, the data of the replacement track is pre-read into the buffer, and the data of the erroneous sector is directly read from the buffer. The nonvolatile storage means, the error map registration means, the write access Means for controlling the read access means and a buffer, it magnetic disk apparatus according to claim. 23. The disk access method of a magnetic disk drive according to claim 2, wherein in the verifying, when an uncorrectable error is detected, a position and a result of the error on the magnetic disk are stored in a nonvolatile storage means. The error map is registered in the error map, the error map is referred to at the time of rewriting, the data of the erroneous sector is written into the replacement track using the data of the buffer, and the data of the replacement track is pre-read into the buffer at the time of read access, The data of the sector having an error is directly read from a buffer, wherein the data is read from the buffer. 24. The control program recording medium for a magnetic disk device according to claim 3, wherein in the verification, when an uncorrectable error is detected, a position and a result on the error magnetic disk are stored in a nonvolatile storage means. The error map is registered in the error map, the error map is referred to at the time of rewriting, the data of the erroneous sector is written to the replacement track using the data of the buffer, and the data of the replacement track is pre-read to the buffer at the time of read access, 3. The disk access control program recording medium for a magnetic disk device according to claim 1, wherein the data of the erroneous sector is read directly from a buffer. 25. The magnetic disk drive according to claim 1, further comprising: a buffer for temporarily storing data exchanged between the magnetic disk and the host device; When performing data transfer with a higher-level device connected outside the device, the original data and its copy data are written to another track on the magnetic disk at the time of the data write access, and the original data is verified. Thus, when an uncorrectable error is detected, the position and the result of the error on the magnetic disk are registered in an error map of the non-volatile storage means, and the error map is referred to by referring to the error map. Copy data is pre-read into the buffer, and during read access, Chromatography data, as read directly from the buffer, the write access means, nonvolatile storage means, the error map registration means controls the read access unit and a buffer, magnetic disk apparatus according to claim. 26. The disk access method for a magnetic disk drive according to claim 2, wherein when data is transferred to or from an externally connected higher-level device, the original data and the original data are written during the data write access. When an uncorrectable error is detected by writing the copy data to another track and verifying the original data, the position of the error on the magnetic disk and the result are registered in an error map of the nonvolatile storage means, and rewritten. The data of the erroneous sector is read ahead of the buffer using copy data, and the data of the erroneous sector is read directly from the buffer at the time of read access. Disk access method. 27. The control program recording medium for a magnetic disk device according to claim 3, wherein when transferring data to / from a higher-level device externally connected thereto, when writing the data, If the uncorrectable error is detected by verifying the original data, the position and the result of the error on the magnetic disk are registered in an error map of the nonvolatile storage means, Referring to the error map at the time of rewriting, prefetching data of the erroneous sector into the buffer using copy data, and directly reading the data of the erroneous sector from the buffer at the time of read access. Disk access control program recording medium for disk device.