Magnetic disk device and method

Abstract

To provide a magnetic disk device capable of efficiently writing data.SOLUTION: Each track of a magnetic disk is provided with a long-distance sector having a length in a circumferential direction that spans two or more servo sectors. A controller executes an acquisition operation to acquire one or more evaluation quantities based on the track pitch of each of two or more servo sectors included in a portion adjacent to the long-distance sector. The controller executes a protection operation to protect data of the adjacent track when the total value of one or more valuation quantities exceeds a first threshold value.SELECTED DRAWING: Figure 11

Description

[Technical field]

[0001] The present embodiment relates to a magnetic disk device and method. [Background technology]

[0002] Conventionally, a method of arranging data on a magnetic disk is known in which data is written to an area longer than the unit size transmitted to and received from a host. Such an area is referred to as a long-distance sector. A long-distance sector has a length that spans multiple servo sectors in the circumferential direction. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] U.S. Pat. No. 1,065,0860 Summary of the Invention [Problem to be solved by the invention]

[0004] An object of one embodiment is to provide a magnetic disk device that can write data efficiently. [Means for solving the problem]

[0005] According to one embodiment, a magnetic disk device includes a magnetic disk, a magnetic head, and a controller. The magnetic disk includes a plurality of tracks. A plurality of servo sectors in which servo information is recorded are arranged at intervals in the circumferential direction on the plurality of tracks. Each track is provided with a plurality of long-distance sectors, which are areas in which data is written. Each long-distance sector has a circumferential length that spans two or more servo sectors among the plurality of servo sectors. The magnetic head writes data to and reads data from each of the plurality of long-distance sectors. The controller operates as follows in a write operation to a first portion, which is a portion on a first track that is one of the plurality of tracks and is adjacent to one long-distance sector of a second track adjacent to the first track. That is, the controller executes an acquisition operation to acquire one or more evaluation amounts based on the track pitch in each of two or more servo sectors included in a range between two ends in the circumferential direction of the first portion. The one or more evaluation amounts correspond to a squeeze amount of the track pitch based on a set value of the track pitch. The controller executes a protection operation to protect data of the second track when a first sum, which is a sum of one or more evaluation quantities, exceeds a first threshold value. [Brief description of the drawings]

[0006] [Figure 1] FIG. 1 is a schematic diagram illustrating an example of the configuration of a magnetic disk device according to the first embodiment. [Diagram 2] FIG. 2 is a schematic diagram showing an example of the configuration of the magnetic disk according to the first embodiment. [Diagram 3] FIG. 3 is a schematic diagram for explaining the SMR method used in the magnetic disk device according to the first embodiment. [Figure 4] FIG. 4 is a diagram for explaining an example of the configuration of a long distance sector according to the first embodiment. [Diagram 5] FIG. 5 is a diagram for explaining an example of the interleaving operation according to the first embodiment. [Figure 6]FIG. 6 is a diagram for explaining an example of a deinterleaving operation according to the first embodiment. [Figure 7] FIG. 7 is a diagram showing an example of the positional relationship between the long distance sector and each servo sector according to the first embodiment. [Figure 8] FIG. 8 is a diagram for explaining the setting of DDOL at the time when writing to the previous track is completed in the first embodiment. [Figure 9] FIG. 9 is a diagram for explaining a write operation to a certain long distance sector on a certain track in the first embodiment. [Figure 10] FIG. 10 is a diagram for explaining an example of a protection operation according to the first embodiment. [Figure 11] FIG. 11 is a diagram illustrating an example of a write operation performed by the controller according to the first embodiment on a write-targeted long-distance sector. [Figure 12] FIG. 12 is a flowchart showing an example of an operation of calculating a squeeze evaluation amount in a passed servo sector according to the first embodiment. [Figure 13] FIG. 13 is a diagram illustrating an example of a squeeze evaluation amount for each data sector calculated in the magnetic disk device according to the first embodiment. [Figure 14] FIG. 14 is a schematic diagram for explaining the operation of the sector slip according to the first embodiment. [Figure 15] FIG. 15 is a schematic diagram for explaining the CMR method used in the magnetic disk device of the first modification. [Figure 16] FIG. 16 is a diagram for explaining an example of the configuration of a long distance sector according to the second modification. [Figure 17] FIG. 17 is a diagram for explaining the setting of DOL when writing to a certain track in the third modification. [Figure 18] FIG. 18 is a diagram for explaining a write operation to a long distance sector on a certain track in the third modification. [Figure 19]FIG. 19 is a diagram for explaining an example of a protection operation according to the third modification. [Figure 20] FIG. 20 is a diagram illustrating an example of a write operation to a certain track by the controller according to the second embodiment. [Figure 21] FIG. 21 is a flowchart for explaining the operation of switching control of DDOL according to the fourth modification. [Figure 22] FIG. 22 is a diagram for explaining an example in which the sector ends of long distance sectors are not aligned between tracks. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0007] Hereinafter, a magnetic disk device and a method according to embodiments will be described in detail with reference to the accompanying drawings. Note that the present invention is not limited to these embodiments.

[0008] (First embodiment) FIG. 1 is a schematic diagram showing an example of the configuration of a magnetic disk device 1 according to the first embodiment.

[0009] The magnetic disk device 1 is connected to a host 2. The magnetic disk device 1 can receive access commands, such as a write command and a read command, from the host 2.

[0010] The magnetic disk device 1 includes a magnetic disk 11 having a recording surface formed on its surface. The magnetic disk device 1 writes and reads data to the magnetic disk 11 (more precisely, the recording surface of the magnetic disk 11) in response to an access command. Note that the magnetic disk device 1 may include multiple magnetic disks 11, but in this embodiment, for the sake of simplicity of explanation and illustration, the magnetic disk device 1 includes one magnetic disk 11.

[0011] Data is written and read via a magnetic head 22. Specifically, in addition to the magnetic disk 11, the magnetic disk device 1 includes a spindle motor 12, a motor driver IC (Integrated Circuit) 21, a magnetic head 22, an actuator arm 15, a voice coil motor (VCM) 16, a lamp 13, a head IC 24, a read / write channel (RWC) 25, a RAM 27, a Flash Read Only Memory (FROM) 28, a buffer memory 29, a hard disk controller (HDC) 23, and a processor 26.

[0012] The magnetic disk 11 is rotated at a predetermined rotation speed by a spindle motor 12 attached to the rotation shaft of the magnetic disk 11. The spindle motor 12 is driven by a motor driver IC 21.

[0013] The motor driver IC 21 controls the rotation of the spindle motor 12 and the rotation of the VCM 16 .

[0014] The magnetic head 22 has a write element 22w and a read element 22r for writing and reading data to and from the magnetic disk 11. The magnetic head 22 is attached to the tip of an actuator arm 15. The magnetic head 22 is moved in the radial direction of the magnetic disk 11 by a VCM 16 driven by a motor driver IC 21.

[0015] When the magnetic disk 11 is not rotating, the magnetic head 22 is moved onto the ramp 13. The ramp 13 is configured to hold the magnetic head 22 at a position separated from the magnetic disk 11.

[0016] During reading, the head IC 24 amplifies and outputs a signal read from the magnetic disk 11 by the magnetic head 22 and supplies it to the RWC 25. The head IC 24 also amplifies a signal corresponding to the data to be written that is supplied from the RWC 25 and supplies it to the magnetic head 22.

[0017] The HDC 23 controls data transmission and reception between the host 2 and the HDC 20 via the I / F bus, controls the buffer memory 29, and performs error correction processing on the read data.

[0018] The buffer memory 29 is used as a buffer for data transmitted and received between the host 2. For example, the buffer memory 29 is used to temporarily store data to be written to the magnetic disk 11 or data read from the magnetic disk 11.

[0019] The buffer memory 29 is, for example, a volatile memory capable of high-speed operation. The type of memory constituting the buffer memory 29 is not limited to a specific type. The buffer memory 29 can be, for example, a dynamic random access memory (DRAM), a static random access memory (SRAM), or a combination of these.

[0020] The RWC 25 modulates the data to be written that is supplied from the HDC 23 and supplies it to the head IC 24. The RWC 25 also demodulates the signal that is read from the magnetic disk 11 and supplied from the head IC 24, and outputs it to the HDC 23 as digital data.

[0021] The processor 26 is, for example, a CPU (Central Processing Unit). A RAM 27, a FROM (Flash Read Only Memory) 28, and a buffer memory 29 are connected to the processor 26.

[0022] The FROM 28 is a non-volatile memory. Firmware (program data) and various operation parameters are stored in the FROM 28. The firmware may be stored in the magnetic disk 11.

[0023] The RAM 27 is configured by, for example, DRAM, SRAM, or a combination of these. The RAM 27 is used as an operating memory by the processor 26. The RAM 27 is used as an area into which firmware is loaded and an area in which various management data is held.

[0024] The processor 26 performs overall control of the magnetic disk device 1 in accordance with firmware stored in the FROM 28 or the magnetic disk 11. For example, the processor 26 loads firmware from the FROM 28 or the magnetic disk 11 into the RAM 27, and controls the motor driver IC 21, the head IC 24, the RWC 25, the HDC 23, and the like in accordance with the loaded firmware.

[0025] The configuration including the RWC 25, the processor 26, and the HDC 23 can also be regarded as the controller 30. The controller 30 may be configured as a SoC (System-On-a-Chip). The controller 30 does not necessarily have to be configured as a SoC. The controller 30 may include other elements (such as a RAM 27, a ROM 28, a buffer memory 29, or the RWC 25) in addition to the above.

[0026] 2 is a schematic diagram showing an example of the configuration of the magnetic disk 11 according to the first embodiment. This diagram shows an example of the rotation direction of the magnetic disk 11. The magnetic head 22 moves relatively to the magnetic disk 11 as the magnetic disk 11 rotates. Therefore, the write / read direction, i.e., the direction in which data is written or read by the magnetic head 22 along the circumferential direction, is opposite to the rotation direction of the magnetic disk 11.

[0027] In the manufacturing process, servo information is written on the magnetic disk 11 by, for example, a servo writer or by self-servo writing (SSW). Referring to Fig. 2, radially arranged servo areas 42 are shown as an example of the arrangement of servo areas in which servo information is written. Data areas 43 in which data can be written are provided between the servo areas 42.

[0028] Based on servo information, a plurality of concentric tracks 41 are set in the radial direction of the magnetic disk 11. A plurality of data areas 43 are provided along the tracks 41, and a plurality of data sectors in which data is written are arranged.

[0029] Hereinafter, an area of ​​the track 41 that is partitioned by the servo areas 42 will be referred to as a servo sector SV. In other words, it can be considered that a plurality of servo sectors SV are arranged on the track 41 at intervals in the circumferential direction.

[0030] Here, expressions related to the positional relationship in the circumferential direction used in this specification are defined. When there is a first region and a second region near the first region along the write / read direction, and the first region is a region that the magnetic head 22 passes through before the second region, the direction from the second region to the first region is expressed as "front". Also, the direction from the first region to the second region is expressed as "rear". Also, when there is a first region and a plurality of second regions along the write / read direction, the second region that is closest to the first region in front of the first region among the plurality of second regions may be expressed as the second region "just before" the first region. Also, when there is a first region and a plurality of second regions along the write / read direction, the second region that is closest to the first region behind the first region among the plurality of second regions may be expressed as the second region "just after" the first region.

[0031] Known methods for writing data to a magnetic disk include a method called SMR (Shingled Magnetic Recording) and a method called CMR (Conventional Magnetic Recording). These methods are collectively called recording methods.

[0032] In the first embodiment, the controller 30 can write data to the magnetic disk 11 in the SMR format.

[0033] 3 is a schematic diagram for explaining the SMR method used in the magnetic disk device 1 according to the first embodiment. In the SMR method, when data (referred to as first data) of a certain track 41 is written and then data (referred to as second data) of a track 41 adjacent to the track 41 in the radial direction is written, each track 41 is arranged so that the second data overlaps a part of the first data. In other words, according to the SMR method, data of one track 41 of two tracks 41 adjacent to each other in the radial direction of the magnetic disk 11 is written overlapping a part of data of the other track of the two tracks 41.

[0034] For example, data on track #2 is written so as to overlap a portion of data already written on track #1. Data on track #3 is written so as to overlap a portion of data already written on track #2. In other words, according to the SMR method, data on one track is repeatedly overlapped with a portion of data already written on an adjacent track.

[0035] As a result, each track width TW is narrowed below the width (WHw) of the write element 22w, thereby improving the recording density.

[0036] However, in the SMR system, because the track width TW is narrower than the width WHw of the write element 22w, when a portion of the data for multiple tracks is updated, data on tracks adjacent to the updated data is destroyed. In order to prevent data destruction, the data for multiple tracks including the portion of the data is updated collectively. The area of ​​multiple tracks that is updated collectively is called a band.

[0037] According to the SMR method, it is determined that writing can be performed only from a predetermined one of the outer side and the inner side of the magnetic disk to the other predetermined one for the multiple tracks 41 in one band. In the example shown in FIG. 3, writing is performed for each track 41 from the outer side to the inner side. The controller 30 may be configured to perform writing for each track 41 from the inner side to the outer side. Also, the order of writing may be set individually for each band.

[0038] In the following, as an example, a description will be given of a case where writing is performed in units of tracks 41 from the outer periphery toward the inner periphery in the SMR method.

[0039] In the first embodiment, a plurality of long distance sectors are provided on each track 41. Each long distance sector is an area in which data is written. Each long distance sector has a circumferential length spanning two or more servo sectors SV.

[0040] 4 is a diagram for explaining an example of the configuration of a long distance sector according to the first embodiment, in which the servo sector SV is omitted.

[0041] A large number of data sectors DS are arranged on one track 41. A data sector ID is assigned to each data sector DS in the track 41. A data sector DS that is assigned an ID of "X" is represented as data sector DS#X.

[0042] 4, 16 data sectors DS are arranged on one track 41. The 16 data sectors DS are given numerical information corresponding to their positional order along the track 41 as IDs.

[0043] Each data sector DS has a capacity corresponding to the unit size of data transferred between the host 2 and the magnetic disk device 1 .

[0044] For example, if the host 2 supports 4K sectors, data is transferred between the host 2 and the magnetic disk device 1 in 4K byte units. In such a case, each data sector DS has a capacity corresponding to 4K bytes. More specifically, the controller 30 performs a predetermined data processing such as encoding using an error correction code on the 4K byte data unit from the host 2, and writes the data unit after the predetermined data processing to the magnetic disk 11. The predetermined data processing makes the size of the data unit larger than 4K bytes. Each data sector DS has a capacity capable of storing a data unit having a size larger than 4K bytes through the predetermined data processing.

[0045] A long distance sector is made up of a plurality of data sectors DS arranged consecutively along the track 41. Similarly to the data sectors DS, each long distance sector is given an ID, which is numerical information corresponding to the positional order along the track 41.

[0046] 4, one long distance sector is made up of four data sectors DS. That is, long distance sector #0 is made up of data sectors DS#0 to DS#3, long distance sector #1 is made up of data sectors DS#4 to DS#7, long distance sector #2 is made up of data sectors DS#8 to DS#11, and long distance sector #3 is made up of data sectors DS#12 to DS#15.

[0047] The remaining data sector #16 is set as a sector where parity is written, that is, as a parity sector. The parity written to data sector DS#16 protects the group of data units written to data sectors DS#0 to DS#15 from errors. In other words, the parity written to data sector DS#16 protects data in units of tracks 41.

[0048] The method of calculating the parity is not limited to a specific method. In one example, the parity is generated by performing XOR for each bit position on a group of data units written to data sectors DS#0 to DS#15. Note that, when an interleaving operation is performed, the parity is calculated before the interleaving operation, which will be described later.

[0049] Hereinafter, data written to one data sector DS or the data before error correction encoding is referred to as a data unit. A set including four data units written to a long distance sector is referred to as a data unit set. Error correction based on an error correction code that is executed on a data unit encoded with the error correction code is referred to as a first error correction. Error correction using parity is referred to as a second error correction.

[0050] The second error correction is performed when the first error correction fails during a read operation.

[0051] In the first embodiment, the controller 30 is further configured to be able to perform interleaving and de-interleaving operations on the set of data units.

[0052] 5 is a diagram for explaining an example of an interleave operation according to the first embodiment. This diagram shows an example of an interleave operation executed on a data unit set written to a certain long-distance sector #m. The long-distance sector #m is configured by data sectors #n to #(n+3). Note that both n and m are integers.

[0053] The data unit set written to long distance sector #m is made up of, for example, data units #0 to #3.

[0054] In the interleaving operation, the controller 30 divides each of the data units #0 to #3 into four sub-data units. Then, the controller 30 changes the order of the arrangement of a total of 16 sub-data units generated by dividing the data units #0 to #3, and writes them to the long distance sector #m. Specifically, the controller 30 distributes the four sub-data units generated from the data unit #0 to the data sectors #n to #(n+3). Similarly, the controller 30 distributes the four sub-data units generated from the data unit #1 to the data sectors #n to #(n+3), distributes the four sub-data units generated from the data unit #2 to the data sectors #n to #(n+3), and distributes the four sub-data units generated from the data unit #3 to the data sectors #n to #(n+3).

[0055] That is, according to the interleaving operation, one data unit is distributed among a plurality of data sectors DS that constitute a long distance sector.

[0056] 6 is a diagram for explaining an example of a deinterleaving operation according to the first embodiment. This diagram shows an example of a deinterleaving operation on data read from the long distance sector #m.

[0057] In the long distance sector #m, each data unit is distributed in sub-data units in an area equivalent to four data sectors DS. In the deinterleave operation, the controller 30 restores the arrangement of each sub-data unit to the original arrangement for the data unit set read from the long distance sector #m and whose arrangement has been changed by the sub-data unit. This restores the data unit set before the interleave operation, in which data unit #0, data unit #1, data unit #2, and data unit #3 are arranged in this order.

[0058] In this way, the interleaving and deinterleaving operations are performed, and thus the long-distance sector #m becomes more resistant to burst errors. For example, even if a burst error portion is included in data read from a certain data sector DS, the burst error portion is dispersed into a plurality of data units by the deinterleaving operation. Since each data unit is individually error-corrected, the dispersed burst error portion is corrected by the first error correction for each data unit.

[0059] Burst errors can also occur when the track width is narrowed by writing to adjacent tracks. Narrowing the track width by writing to adjacent tracks is called squeeze write. It becomes difficult to read data without errors from a section where the track width has been significantly narrowed by squeeze write, and as a result, burst errors can occur in data read from that section. By performing interleaving and deinterleaving operations, even if a burst error with a length equal to or less than the data sector DS occurs due to a local squeeze write with a length equal to or less than the data sector DS, the burst error can be corrected by the first error correction.

[0060] As described above, the long distance sector has a length spanning multiple servo sectors SV in the circumferential direction.

[0061] 7 is a diagram showing an example of the positional relationship between the long distance sector and each servo sector SV according to the first embodiment. Each servo sector SV is actually an area with a certain width in the circumferential direction, and each servo sector SV and each data area are mutually exclusive. However, in order to prevent the diagram from becoming complicated, the position of each servo sector SV is simply shown by dotted lines.

[0062] 7, the range between the two circumferential ends of long-distance sector #m includes four servo sectors SV#p to SV#(p+3). Servo sector SV#p exists at the beginning of data sector DS#n (more precisely, immediately before data sector DS#0), servo sector SV#(p+1) exists in data sector DS#(n+1), servo sector SV#(p+2) exists in data sector DS#(n+2), and servo sector SV#(p+3) exists in data sector DS#(n+3).

[0063] The controller 30 estimates the current position of the magnetic head 22 based on the servo information read from a servo sector SV when the magnetic head 22 (more precisely, the read element 22r) passes over that servo sector SV, and executes positioning control of the magnetic head 22 based on the estimated current position. The positioning control of the magnetic head 22 includes seek control for moving the magnetic head 22 to a target track 41, and tracking control for maintaining the magnetic head 22 on the target track 41.

[0064] In a write operation for writing data to the long distance sector #m, the controller 30 executes tracking control based on the servo information read from at least four servo sectors SV#p to SV#(p+3).

[0065] Specifically, the controller 30 estimates the current position of the magnetic head 22 in each of the four servo sectors SV#p to SV#(p+3) and determines whether the magnetic head 22 is in an on-track state. If it is determined that the magnetic head 22 is in an on-track state, the controller 30 executes writing of data.

[0066] If it is determined that the magnetic head 22 is not on-track, i.e., if it is determined that the magnetic head 22 is off-track, the controller 30 immediately stops writing. Then, if it is determined that the magnetic head 22 is on-track when the magnetic disk 11 has rotated one or more times and approaches the same circumferential position, the controller 30 resumes writing. This operation of stopping writing and resuming writing after waiting for a rotation is called a write retry operation.

[0067] Whether the magnetic head 22 is on-track or not is determined based on whether the magnetic head 22 is located in a write permission range provided in the radial direction. The write permission range is set for each track 41. The boundary line of the write permission range is called the drift off level (DOL). Hereinafter, the drift off level will be abbreviated as DOL.

[0068] There are various methods for setting the DOL. According to the SMR method, in one band, the DOL of the track 41 to be written later is set based on the position of data of the track 41 that is written earlier among two adjacent tracks 41. That is, the DOL is dynamically set based on the position of the written data. Such a dynamically set DOL is called DDOL (Dynamic Drift Off Level).

[0069] Hereinafter, when one of two adjacent tracks 41 has already been written and data is to be written to the other track 41, the track 41 to which data has already been written is referred to as the previous track.

[0070] In the first embodiment, the DDOL is set based on not only the position of the data of the previous track but also the total value of the squeeze evaluation amount. The squeeze evaluation amount is an evaluation amount indicating how much the distance between the track center of the previous track and the track center of the write target track 41 (i.e., the track pitch) is narrowed based on the set value of the track pitch. As an example, the squeeze evaluation amount is numerical information that becomes larger as the actual track pitch, i.e., the distance between the track center of the previous track and the track center of the write target track 41, becomes narrower.

[0071] As described above, the long-distance sector has a length that spans a plurality of servo sectors SV, and furthermore, the data unit set written to the long-distance sector is interleaved. Even in a situation where the actual track pitch is significantly narrowed in a certain portion of a certain long-distance sector, and a burst error occurs in that portion due to squeeze writing during a subsequent read operation, if data can be read without error from other portions of the long-distance sector, it is possible to obtain error-free data by the first error correction.

[0072] In other words, in order to estimate as accurately as possible whether error-free data can be obtained by the first error correction, it is necessary to make a judgment based on the squeeze evaluation value for all of the multiple servo sectors SV included in the range between the two ends of the long-distance sector, rather than just a portion of the multiple servo sectors SV.

[0073] Therefore, when performing a write operation on a certain long-distance sector (referred to as a long-distance sector to be written), the controller 30 executes the following operation to protect data of another long-distance sector (referred to as an adjacent long-distance sector) located at the same circumferential position as the long-distance sector to be written on the previous track. That is, the controller 30 calculates the squeeze evaluation amount at the position of each sector (each servo sector SV or each data sector DS) based on the servo data read from each servo sector SV in the range from the beginning to the end of the long-distance sector to be written. The squeeze evaluation amount at the position of each sector (each servo sector SV or each data sector DS) is referred to as an acquisition operation. In the acquisition operation, if there is a sector (servo sector SV or data sector DS) that has not yet been passed, the controller 30 predicts the squeeze evaluation amount of the sector that has not yet been passed. Then, the controller 30 sums up the squeeze evaluation amounts in all sectors (servo sectors SV or data sectors DS) included in the range between the two ends of the long-distance sector to be written, and compares the sum of the squeeze evaluation amounts with a threshold value (referred to as threshold value Th1).

[0074] The threshold value Th1 is preset to a value as large as possible within a numerical range that ensures that an error-free data unit set can be obtained from the adjacent long-distance sector by the first error correction. Therefore, if the total value of the squeeze evaluation amount is smaller than the threshold value Th1, it can be estimated that it is possible to obtain error-free data from the adjacent long-distance sector. Also, if the total value of the squeeze evaluation amount is larger than the threshold value Th1, it can be estimated that it may be difficult to obtain error-free data from the adjacent long-distance sector.

[0075] If the sum of the squeeze evaluation quantities is greater than the threshold value Th1, the controller 30 executes a protection operation to protect the data in the adjacent long distance sector.

[0076] The controller 30 executes a protection operation to tighten the write permission range. Specifically, the tightening of the write permission range means tightening the DDOL. If the magnetic head 22 exceeds the DDOL, the controller 30 immediately stops writing and executes a write retry operation. If the magnetic head 22 has not reached the DDOL, the controller 30 continues the write operation.

[0077] Hereinafter, the servo sectors SV included in the range between the two ends of the long-distance sector are simply referred to as the servo sectors SV included in the long-distance sector. In other words, according to the example shown in Fig. 7, the four servo sectors SV#p to SV#(p+3) correspond to the servo sectors SV included in the long-distance sector #m.

[0078] In the first embodiment, the controller 30 calculates the squeeze evaluation amount for each data sector DS. The controller 30 may be configured to calculate the squeeze evaluation amount for each servo sector SV.

[0079] An example of the protection operation according to the first embodiment will be described with reference to Figures 8 to 10. Here, track #(k-1) and track #k are taken as two adjacent tracks 41, and the operation when a write operation is performed on long distance sector #m on track #k will be described. In other words, when track #k is used as a reference, track #(k-1) corresponds to the previous track.

[0080] In the example described in this specification, the sector ends are aligned between the multiple long-distance sectors arranged in the cross-track direction, but the technology described in this specification can be applied even if the sector ends are not aligned between the multiple long-distance sectors arranged in the cross-track direction. Although the examples in FIG. 8 to FIG. 10 show an example in which the long-distance sectors in the track #(k-1) and the track #k are aligned, the beginning and end of the long-distance sectors do not necessarily have to be aligned. For example, they may be shifted in units of 4k sectors or servo sector SV. As shown in FIG. 22, the long-distance sector #m of the track #k may be composed of DS#(n+2) to DS#(n+5). In this case, the long-distance sector #m-1 of the track #k is DS#(n-2) to DS#(n+1). In this case, the calculation of the total value of the squeeze evaluation amount is calculated in units of the adjacent long-distance sector #m of the track #(k-1). Therefore, the sum of the squeeze evaluation amounts is calculated in the write operation to the long distance sectors #m-1 and #m on the track #k. In other words, the sum is calculated in the write operation to the portion of the track #k adjacent to the adjacent long distance sector of the track #(k-1), that is, the portion from DS#(n) to DS#(n+3) on the track #k.

[0081] For simplicity, the following description will be given assuming that the long distance sectors on track #(k-1) and track #k are aligned, i.e., that a write operation is performed on long distance sector #m on track #k. In other words, the write to long distance sector #m on track #k is aligned with the adjacent long distance sector #m on track #(k-1) that is the protection target.

[0082] FIG. 8 is a diagram for explaining the setting of DDOL at the time when writing to the previous track (that is, track #(k-1)) is completed in the first embodiment.

[0083] 8 shows the transition of the position error signal PES#(k-1) during a write operation on track #(k-1). The position error signal PES#(k-1) represents the actual trajectory of the magnetic head 22 during a write operation on track #(k-1). In other words, the position error signal PES#(k-1) indicates the write position of the data on track #(k-1).

[0084] In the initial state, the controller 30 calculates a DDOL for a write operation on track #k based on the write position of data on track #(k-1), i.e., the position error signal PES#(k-1). k Specifically, the controller 30 sets the DDOL at a position offset from the position error signal PES#(k-1) by a predetermined fixed length L1 toward the inner periphery. k Set.

[0085] FIG. 9 is a diagram for explaining a write operation to long distance sector #m on track #k in the first embodiment.

[0086] The controller 30 writes data in sequence from the head of the long-distance sector #m on the track #k. Each time the magnetic head 22 passes a servo sector SV, the controller 30 executes an acquisition operation to calculate the squeeze evaluation amount for each sector, calculates a sum SQsum1 of the squeeze evaluation amounts, and compares the sum SQsum1 of the squeeze evaluation amounts with a threshold value Th1.

[0087] 9, as indicated by the position error signal #k, the magnetic head 22 passes through the servo sectors SV#p and SV#(p+1) while writing data, and then reaches the servo sector SV#(p+2).

[0088] When the magnetic head 22 reads the servo information from the servo sector SV#(p+2), the controller 30 calculates the squeeze evaluation quantity SQ n, the squeeze estimate SQ in data sector DS#(n+1) n+1 , and the squeeze estimate SQ in data sector DS#(n+2) n+2 is calculated based on the read servo information.

[0089] There are various methods for calculating the squeeze evaluation amount.

[0090] In one example, the controller 30 calculates the actual track pitch TP for each servo sector SV based on the position error signal PES#(k-1), the position error signal PES#k, and the track pitch set value TPset. Then, the controller 30 calculates a value (referred to as the squeeze amount) obtained by subtracting the actual track pitch TP from the track pitch set value TPset for each servo sector SV. Then, the controller 30 determines the squeeze amount in the servo sector SV closest to one data sector DS as the squeeze evaluation amount for that data sector DS.

[0091] In another example, the controller 30 uses the maximum squeeze amount in all servo sectors SV passed from the servo sector SV immediately before a data sector DS to the servo sector DV immediately after the data sector DS as the squeeze evaluation amount for the data sector DS.

[0092] The controller 30 predicts the squeeze evaluation amount in the remaining data sector DS#(n+3) that the magnetic head 22 has not yet passed over. The predicted value SQE of the squeeze evaluation amount in the data sector DS#(n+3) n+3 There are various methods for calculating the value of .

[0093] In one example, the maximum value of the squeeze evaluation amount in all data sectors DS that the magnetic head 22 has already passed among the data sectors DS included in the long distance sector #m to be written is the predicted value SQE n+3 It is said.

[0094] In another example, the average value of the squeeze evaluation amounts of all data sectors DS that the magnetic head 22 has already passed among the data sectors DS included in the long distance sector #m to be written is the predicted value SQE n+3 It is said.

[0095] In yet another example, a trained neural network model configured to output an estimated value of the squeeze amount or squeeze evaluation amount in one or more sectors (servo sectors SV or data sectors DS) arranged immediately after a predetermined number of sectors (servo sectors SV or data sectors DS) arranged in succession in the circumferential direction when the squeeze amount or squeeze evaluation amount in the predetermined number of sectors is input is implemented as an electronic circuit in a storage area. The neural network model has three layers, for example, an input layer, an intermediate layer, and an output layer. Each layer has a large number of neurons, and the neurons are connected with a certain weight. Learning refers to adjusting how the weight is set in each connection so that the output error is minimized. For example, inside the controller 30, there is a servo logic unit having a synchronization circuit that obtains the squeeze amount for each servo sector SV, and transmits it to a servo channel unit inside the controller 30. The servo channel unit is implemented with a neural network model as a circuit. The controller 30 calculates a predicted value SQE of the squeeze evaluation amount based on the trained neural network model. n+3 Get the.

[0096] In yet another example, the controller 30 is provided with a circuit that, when a squeeze amount or squeeze evaluation amount in a predetermined number of sectors (servo sectors SV or data sectors DS) arranged consecutively in the circumferential direction is input, outputs an estimated value of the squeeze amount or squeeze evaluation amount in one or more sectors (servo sectors SV or data sectors DS) arranged immediately after the predetermined number of sectors using a regression equation. The controller 30 calculates a predicted value SQE of the squeeze evaluation amount based on a prediction using the learned regression equation. n+3 Get the.

[0097] In this way, in the acquisition operation, the controller 30 obtains one or more squeeze evaluation amounts (four squeeze evaluation amounts SQ in this case) for the long distance sector #m to be written based on the actual track pitches TP of all servo sectors SV included in the long distance sector #m to be written. n , S.Q. n+1 , S.Q. n+2 , S.Q.E. n+3 ) to get the

[0098] The controller 30 calculates the squeeze evaluation quantity SQ n , the squeeze estimate SQ in data sector DS#(n+1) n+1 , the squeeze estimate SQ in data sector DS#(n+2) n+2 , and the squeeze metric SQE in data sector DS#(n+3) n+3 When these values ​​are acquired, the controller 30 calculates the sum SQsum1 of the squeeze evaluation quantities. Then, the controller 30 compares the sum SQsum1 of the squeeze evaluation quantities with a threshold value Th1.

[0099] If the sum SQsum1 of the squeeze evaluation quantities does not reach the threshold Th1, the controller 30 continues writing to the write target long distance sector #m. If the sum SQsum1 of the squeeze evaluation quantities exceeds the threshold Th1, the controller 30 executes a protection operation.

[0100] The process to be performed when the sum SQsum1 of the squeeze evaluation amounts is equal to the threshold Th1 can be arbitrarily designed by the designer. For example, the controller 30 may continue writing or may execute a protection operation. Here, the controller 30 continues writing when the sum SQsum1 of the squeeze evaluation amounts is equal to the threshold Th1.

[0101] In the example shown in FIG. 9, the sum SQsum1 of the squeeze evaluation quantities exceeds the threshold Th1, which causes the controller 30 to execute a protective operation.

[0102] FIG. 10 is a diagram for explaining an example of a protection operation according to the first embodiment.

[0103] When the magnetic head 22 reaches the servo sector SV#(p+2), the sum SQsum1 of the squeeze evaluation amounts exceeds the threshold Th1, and the controller 30 executes a protection operation. In the protection operation, the controller 30 performs a DDOL operation after the servo sector SV#(p+2). k 10, the controller 30 tightens the DDOL at a position after the servo sector SV#(p+2) and offset inward from the position error signal PES#(k-1) by a length L2 that is longer than L1. k This sets the DDOL k is moved to the inner periphery.

[0104] DDOL k To tighten the DDOL k The purpose of this is to move the DDOL to the inner side. k As a result of the tightening of the DDOL, the conditions for determining whether or not a vehicle is on track will become stricter. k In the portion where the track width is reduced, the track width of the previous track is prevented from being narrowed, and as a result, it is possible to prevent the data unit set written in the long distance sector #m of the previous track from becoming difficult to read. In other words, the data unit set written in the long distance sector #m of the previous track is protected.

[0105] Here, a technique to be compared with the first embodiment will be described. The technique to be compared with the first embodiment will be referred to as a comparative example. According to the comparative example, the maximum value of the squeeze amount calculated based on the passed servo sectors included in the long distance sector is regarded as the squeeze amount in the entire long distance sector. The controller determines whether or not to execute a write retry operation based on a comparison between the maximum value of the squeeze amount and a certain threshold value. However, according to the comparative example, the write retry operation occurs frequently, reducing the efficiency of the write operation.

[0106] In contrast, in the first embodiment, a protection operation is performed for writing to a long-distance sector based on the sum SQsum1 of the squeeze evaluation amounts in all data sectors DS included in the long-distance sector. Even if the track width of the previous track is significantly narrowed in a part of the long-distance sector of the previous track, the write operation can be continued as long as the sum SQsum1 of the squeeze evaluation amounts does not reach the threshold value Th1. Therefore, unlike the comparative example, a decrease in the efficiency of the write operation due to frequent write retry operations is suppressed. In other words, the magnetic disk device 1 can write data efficiently.

[0107] Next, the operation of the magnetic disk device 1 according to the first embodiment will be described.

[0108] FIG. 11 is a diagram showing an example of a write operation performed by the controller 30 according to the first embodiment on a write-target long-distance sector.

[0109] First, the controller 30 executes writing up to just before the first servo sector SV included in the long distance sector to be written (S101). Then, when the magnetic head 22 reads the servo information from the servo sector SV (S102), the controller 30 starts the acquisition operation.

[0110] In the acquisition operation, the controller 30 first calculates the squeeze evaluation amount for each passed data sector DS included in the long distance sector to be written (S103).

[0111] Then, the controller 30 calculates the sum of the squeeze evaluation amounts of all the passed data sectors DS included in the write target long distance sector (S104). The sum obtained by the process of S104 is denoted as SQsum2.

[0112] Furthermore, the controller 30 predicts the squeeze evaluation amount for each data sector DS that has not yet been passed and is included in the long distance sector to be written (S105).

[0113] When the squeeze evaluation amounts for all data sectors DS included in the long distance sector to be written are obtained by the acquisition operation, the controller 30 calculates the sum SQsum1 of these squeeze evaluation amounts (S106).

[0114] The controller 30 determines whether the sum SQsum1 is greater than a threshold value Th1 (S107).

[0115] If the sum SQsum1 is greater than the threshold Th1 (S107: Yes), the controller 30 executes DDOL contraction (S108). Specifically, the controller 30 calculates a length L2 based on the following formula (1), and moves the DDOL to a position offset by the length L2 toward the inner periphery from the position error signal of the previous track. L2=TPset-(Th1-SQsum2)+Cov...(1)

[0116] TPset is the set value of the track pitch. Cov is a margin set in consideration of overrun. When an off-track state occurs, the magnetic head 22 may move beyond the DDOL toward the previous track during the period from when the off-track state occurs until writing stops. The distance that the magnetic head 22 may move in the radial direction beyond the DDOL before writing stops is called overrun. Cov is the length corresponding to the expected overrun.

[0117] If the sum SQsum1 is greater than the threshold Th1, it is expected that it will become difficult to read the data of the adjacent long distance sector if the write operation is continued. The allowable amount of squeeze to prevent the data of the adjacent long distance sector from becoming difficult to read is expressed by (SQsum2-Th1). Therefore, if the magnetic head 22 does not pass over a line that is separated by a length ((TPset-(Th1-SQsum2)) from the position error signal of the previous track, the controller 30 can prevent the data of the adjacent long distance sector from becoming difficult to read. However, since an overrun may actually occur, the DDOL is set at a position offset by Cov toward the inner periphery from the line determined by the length ((TPset-(Th1-SQsum2)) in consideration of the overrun.

[0118] If the sum SQsum1 is not greater than the threshold Th1 (S107: No), or after the process of S108, the controller 30 determines whether the magnetic head 22 is located on the outer circumferential side (in other words, on the previous track side) than the DDOL (S109).

[0119] If the magnetic head 22 is located on the outer circumferential side of the DDOL (S109: Yes), it is estimated that the magnetic head 22 is in an off-track state. Therefore, the controller 30 stops writing and postpones (postpones) writing until the magnetic disk 11 rotates once (S110). Then, when the magnetic disk 11 rotates once and the magnetic head 22 again approaches the circumferential position where writing stopped, the controller 30 executes the process of S109 again.

[0120] If the magnetic head 22 is not located on the outer periphery side of the DDOL (S109: No), it can be assumed that the magnetic head 22 is in an on-track state. Therefore, the controller 30 executes a write. Specifically, the controller 30 first determines whether the last passed servo sector SV is the last servo sector SV included in the long-distance sector to be written (S111).

[0121] If the last servo sector SV passed is not the last servo sector SV included in the long distance sector to be written (S111: No), the controller 30 executes writing up to just before the next servo sector SV (S112), and the control transitions to S102.

[0122] If the last passed servo sector SV is the last servo sector SV included in the long-distance sector to be written (S111: Yes), the controller 30 executes writing up to the end of the long-distance sector to be written (S113), and the write operation ends.

[0123] In the series of operations shown in FIG. 11, the sum SQsum1 is an example of a first sum. The threshold Th1 is an example of a first threshold. The sum SQsum2 is an example of a second sum. Each servo sector SV included in the long-distance sector to be written is an example of a first servo sector. Among the multiple servo sectors SV included in the long-distance sector to be written, the servo sector SV that the magnetic head 22 has passed through is an example of a second servo sector. Among the multiple servo sectors SV included in the long-distance sector to be written, the servo sector SV that the magnetic head 22 has not yet passed through is an example of a third servo sector.

[0124] In addition, in the series of operations shown in FIG. 11, the process of S112 or S113 executed after the determination in the determination process of S109 is No corresponds to a write retry operation.

[0125] FIG. 12 is a flowchart showing an example of the operation of calculating the squeeze evaluation amount in the passed servo sector SV, that is, the process of S103 shown in FIG. 11, according to the first embodiment.

[0126] First, the controller 30 selects one of the data sectors DS included in the long distance sector to be written and through which the magnetic head 22 has passed (S201). The selected data sector DS is referred to as the target data sector DS.

[0127] The controller 30 calculates the minimum value TPmin of the track pitch TP in all servo sectors SV passed among all servo sectors SV from the servo sector SV immediately before the target data sector DS to the servo sector SV immediately after the target data sector DS (S202).

[0128] In the first embodiment, when the target data sector DS exists across one or more data areas, the worst value (i.e., the maximum value) of the squeeze amount calculated in all servo sectors SV located immediately before and immediately after the one or more data areas is regarded as the squeeze amount in the target data sector DS. This makes it possible to evaluate the squeeze amount in the worst case, that is, the case in which the track width of the adjacent long distance sector in the target data sector DS is narrowest.

[0129] To this end, in the example shown in FIG. 12, when the target data sector DS exists across one or more data areas, first, the minimum track pitch value TPmin of all servo sectors SV located immediately before and after the one or more data areas is calculated.

[0130] Following S202, the controller 30 determines whether the DDOL has been tightened (S203). That is, the controller 30 determines whether the DDOL has been moved to a position offset from the position error signal of the previous track by a length L2 toward the inner periphery.

[0131] If the DDOL has been tightened (S203: Yes), the controller 30 determines whether the minimum value TPmin is smaller than L2 (S204).

[0132] If the controller 30 detects that the magnetic head 22 is located on the outer periphery side of the DDOL (i.e., on the previous track side), the controller 30 cannot continue writing. However, since writing is continuing at the time when the process of S103 is executed, it is considered that the magnetic head 22 is not actually located on the outer periphery side of the DDOL even if the minimum value TPmin is smaller than L2. Therefore, if the minimum value TPmin is smaller than L2 (S204: Yes), the controller 30 replaces the minimum value TPmin with the value of L2 (S205).

[0133] If the DDOL has not been tightened (S203: No), if the minimum value TPmin is not smaller than L2 (S204: No), or after processing of S205, the controller 30 calculates the squeeze evaluation quantity SQ in the target data sector DS using the following equation (2) (S206). SQ = TPset - TPmin (2)

[0134] The controller 30 judges whether the calculation of the squeeze evaluation quantity SQ for all passed data sectors DS included in the long distance sector to be written is completed (S207). In this specification, the passed data sectors DS include not only the data sectors DS over whose entire area the magnetic head 22 has passed, but also the data sectors DS over whose partial area the magnetic head 22 has passed.

[0135] If there remains a data sector DS for which the squeeze evaluation amount SQ has not yet been calculated among all the passed data sectors DS included in the long distance sector to be written (S207: No), the controller 30 selects one data sector DS for which the squeeze evaluation amount SQ has not yet been calculated as a new target data sector DS (S208). Then, the control transitions to S202.

[0136] When the calculation of the squeeze evaluation amount SQ for all passed data sectors DS included in the long distance sector to be written is completed (S207: Yes), the operation of calculating the squeeze evaluation amount SQ for the passed data sectors DS ends.

[0137] Fig. 13 is a diagram showing an example of the squeeze evaluation amounts SQ and SQE for each data sector DS calculated in the magnetic disk device 1 according to the first embodiment. In the example shown in this figure, the squeeze evaluation amount SQ in the passed data sector DS is calculated according to the operation shown in Fig. 12. Also, the maximum value of the squeeze evaluation amounts SQ in all passed data sectors DS is used as the predicted value SQE of the squeeze evaluation amount in the data sector DS that has not yet been passed.

[0138] 13, the columns of data sectors DS#n through DS#(n+3) show the squeeze evaluation amounts SQ and SQE calculated when the magnetic head 22 is located at each of servo sectors SV#p through SV#(p+3). The numerical information in parentheses shows the actual track pitch TP detected at each servo sector SV. The cells with diagonal hatching show the predicted value SQE of the squeeze evaluation amount.

[0139] FIG. 13 also shows the sums SQsum1 and SQsum2 calculated when the magnetic head 22 is located at each of the servo sectors SV#p to SV#(p+3).

[0140] In the example shown in this figure, the set value TPset of the track pitch is 50 nm, the threshold value Th1 is 24 nm, and the margin Cov for overrun is 2 nm.

[0141] As shown in the row of servo sector SV#p, the track pitch TP in servo sector SV#p when the magnetic head 22 reached servo sector SV#p was 50 nm. Since the magnetic head 22 has reached the beginning of data sector DS#n, the controller 30 selects data sector DS#n as the target data sector DS and calculates the minimum track pitch TPmin for data sector DS#n. Since only servo sector SV#p has passed through the servo sectors SV#p immediately before data sector DS#n to SV#(p+1) immediately after data sector DS#n, the controller 30 obtains 50 nm, which is the track pitch TP in servo sector SV#p, as the minimum track pitch TPmin. Since the DDOL has not been shrunk, the controller 30 obtains 0 nm as the squeeze evaluation amount SQ in data sector DS#n by subtracting 50 nm, which is the minimum track pitch TPmin, from 50 nm, which is the set value TPset of the track pitch.

[0142] In addition, since the maximum value of the squeeze evaluation amount SQ in all the data sectors DS passed in the long distance sector to be written is 0 nm, the predicted value SQE n+1 , S.Q.E. n+2 , S.Q.E. n+3 are all set to 0 nm.

[0143] At this time, the sum SQsum1 of the squeeze evaluation quantities is 0 nm, which does not exceed the threshold value Th1, and therefore the protective operation is not started.

[0144] As shown in the row of servo sector SV#(p+1), the track pitch TP at servo sector SV#(p+1) when the magnetic head 22 reached servo sector SV#(p+1) was 44 nm. When the controller 30 selected data sector DS#n as the target data sector DS, it acquired 44 nm as the minimum value TPmin. Therefore, the controller 30 subtracts 44 nm, which is the minimum value TPmin of the track pitch, from 50 nm, which is the set value TPset of the track pitch, to obtain the squeeze evaluation quantity SQ n to get 6nm.

[0145] When the controller 30 selects the data sector DS#(n+1) as the target data sector DS, it obtains 44 nm as the minimum value TPmin. Therefore, the controller 30 subtracts 44 nm, which is the minimum value TPmin of the track pitch, from 50 nm, which is the set value TPset of the track pitch, to obtain the squeeze evaluation quantity SQ n+1 to get 6nm.

[0146] Since the maximum value of the squeeze evaluation amount SQ in all data sectors DS passed in the long distance sector to be written is 6 nm, the predicted value SQE n+2 , S.Q.E. n+3 All are set to 6nm.

[0147] At this time, the sum SQsum1 of the squeeze evaluation quantities is 24 nm, which does not exceed the threshold value Th1, and therefore the protective operation is not started.

[0148] As shown in the row for servo sector SV#(p+2), the track pitch TP at servo sector SV#(p+2) when the magnetic head 22 reached servo sector SV#(p+2) was 43 nm. When the controller 30 selected data sector DS#n as the target data sector DS, it acquired 44 nm as the minimum value TPmin. Therefore, the controller 30 subtracts 44 nm, which is the minimum value TPmin of the track pitch, from 50 nm, which is the set value TPset of the track pitch, to obtain the squeeze evaluation quantity SQ n to get 6nm.

[0149] When the controller 30 selects the data sector DS#(n+1) as the target data sector DS, it obtains 43 nm as the minimum value TPmin. Therefore, the controller 30 subtracts 43 nm, which is the minimum value TPmin of the track pitch, from 50 nm, which is the set value TPset of the track pitch, to obtain the squeeze evaluation quantity SQ n+1 to get 7nm.

[0150] Since the maximum value of the squeeze evaluation amount SQ in all the data sectors DS that have passed in the long distance sector to be written is 7 nm, the predicted value SQE n+3 is determined to be 7 nm.

[0151] At this time, the sum SQsum1 of the squeeze evaluation amounts is 28 nm, which exceeds the threshold Th1. Therefore, a protection operation is started. Specifically, the controller 30 calculates the length L2 using the formula (1), and sets the DDOL at a position offset by the length L2 toward the inner periphery from the position indicated by the position error signal of the previous track.

[0152] In the example shown in Fig. 13, the track pitch setting value TPset is 50 nm, the threshold value Th1 is 24 nm, and the margin Cov for overrun is 2 nm. The sum SQsum2 when the magnetic head 22 reaches the servo sector SV#(p+2) is 20 nm. Therefore, the controller 30 obtains 48 nm as the length L2 by using the formula (1). The controller 30 sets the DDOL at a position offset by 48 nm toward the inner periphery from the position indicated by the position error signal of the previous track.

[0153] When the magnetic head 22 reaches the servo sector SV#(p+2), it is located at a position offset by 43 nm toward the inner periphery from the position indicated by the position error signal of the previous track. In other words, the magnetic head 22 is located on the outer periphery side of the post-contraction DDOL. Therefore, the controller 30 stops writing, and when the magnetic head 22 reaches the servo sector SV#(p+2) again, it judges whether to resume writing based on the DDOL.

[0154] 13, writing is resumed when the magnetic head 22 reaches the servo sector SV#(p+2) again. Then, when the magnetic head 22 reaches the servo sector SV#(p+1), the track pitch TP in the servo sector SV#(p+3) is 48 nm, as shown in the row of the servo sector SV#(p+3). The controller 30 sequentially selects the data sectors DS#n, DS#(n+1), and DS#(n+2) as the target data sectors DS, thereby obtaining the squeeze evaluation quantity SQ n 6nm, squeeze evaluation quantity SQ n+1 7nm, squeeze evaluation quantity SQ n+2 to get 7nm.

[0155] During the write retry operation, the controller 30 may obtain the squeeze evaluation quantities SQ of the data sectors DS#n, DS#(n+1), and DS#(n+2) in the same manner as before the write retry operation.

[0156] Alternatively, the controller 30 may store the squeeze evaluation quantities SQ of data sectors DS#n, data sectors DS#(n+1), and data sectors #(n+2) calculated before the write retry operation in any one of the memory areas, and read these squeeze evaluation quantities SQ from this memory area during the write retry operation.

[0157] In a write retry operation, when the controller 30 selects the data sector DS#(n+3) as the target data sector DS, it obtains 43 nm as the minimum value TPmin. However, the minimum value TPmin is smaller than 48 nm, which is the length L2. Therefore, the controller 30 sets 48 nm, which is the value of the length L2, as the minimum value TPmin. After that, the controller 30 uses the formula (2) to obtain the squeeze evaluation quantity SQ n+3 to get 2nm.

[0158] At this time, the sum SQsum1 of the squeeze evaluation amounts is 24 nm, which does not exceed the threshold value Th1. Therefore, the protection operation is not started and writing continues.

[0159] In the above description, the controller 30 tightens the DDOL as a protective operation, but the example of the protective operation is not limited to this.

[0160] In one example, the controller 30 may be configured to immediately stop writes without tightening the DDOL and perform a write retry operation as a protective action.

[0161] In another example, the controller 30 may execute a sector slip as a protective action.

[0162] FIG. 14 is a schematic diagram for explaining the operation of sector slip according to the first embodiment. As shown in this figure, in the sector slip, the data sector DS constituting the long distance sector to be written is relocated to a position after the initially set position. Alternatively, the sector slip may be performed in units of long distance sectors. That is, in addition to DS#n+3, written DS#n, DS#n+1, and DS#n+2 are slipped. In this case, the DS#n after the slip is placed at the write position of DS#0 in the long distance sector next to DS#n+3 that was originally to be written. That is, the sector slip is performed in units of long distance sectors.

[0163] 11, the controller 30 waits for the magnetic disk 11 to rotate (i.e., performs the process of S110) until the magnetic head 22 does not go beyond the DDOL toward the outer periphery. An upper limit may be set for the number of times the magnetic disk 11 is waited for rotation. When the number of times the magnetic disk 11 is waited for rotation reaches the upper limit, the controller 30 may execute a sector slip.

[0164] The upper limit of the number of rotational delays of the magnetic disk 11 may be different when the DDOL is tightened and when the DDOL is not tightened. For example, the upper limit when the DDOL is tightened is set to a value smaller than the upper limit when the DDOL is not tightened.

[0165] As described above, according to the first embodiment, the controller 30 operates as follows in a write operation to one of the long-distance sectors. That is, the controller 30 executes an acquisition operation to acquire one or more squeeze evaluation amounts based on the track pitch TP for each of two or more servo sectors SV included in the long-distance sector to be written. In the acquisition operation, the controller 30 calculates a squeeze evaluation amount for the servo sector SV that the magnetic head 22 has passed based on the position error signal of the previous track and the position error signal of the track 41 including the long-distance sector to be written. In the acquisition operation, the controller 30 predicts a squeeze evaluation amount for the servo sector SV that the magnetic head 22 has not yet passed. When the sum SQsum1, which is the sum of the squeeze evaluation amounts of all the long-distance sectors to be written, exceeds the threshold value Th1, the controller 30 executes a protection operation to protect the data of the previous track.

[0166] Therefore, unlike the comparative example, a decrease in efficiency of a write operation caused by frequent write retry operations is suppressed. In other words, the magnetic disk device 1 is capable of writing data efficiently.

[0167] Furthermore, according to the first embodiment, the controller 30 tightens the write permission range in the protection operation.

[0168] Thus, the data on the previous track is protected.

[0169] Also, according to the first embodiment, the sum SQsum2 of the squeeze evaluation amounts is calculated based on all servo sectors SV passed by the magnetic head 22 among all two or more servo sectors SV included in the long distance sector to be written. Then, in the protection operation, the controller 30 calculates the length L2 based on the sum SQsum2, and sets the DDOL based on the length L2.

[0170] Thus, the data on the previous track is protected.

[0171] In the example shown in FIG. 4, the parity sector does not belong to any of the long-distance sectors. The parity sector may be included in the long-distance sector at the end of the track 41. For example, in FIG. 4, the parity sector may be included in the long-distance sector #3. The long-distance sector #3 is different from any of the long-distance sectors #0 to #2 and is composed of five data sectors DS#12 to DS#16. For the long-distance sector #3, the controller 30 calculates a sum SQsum1, which is the sum of the squeeze evaluation amounts of the entire long-distance sector, from the five data sectors DS#12 to DS#16, and compares the sum SQsum1 with a threshold value Th1. However, a value different from the threshold value Th1 used for the long-distance sector #3 is used.

[0172] As described above, according to the first embodiment, the controller 30 predicts a squeeze evaluation amount based on a servo sector SV that the magnetic head 22 has not passed, based on a squeeze evaluation amount based on a servo sector SV that the magnetic head 22 has passed.

[0173] The first embodiment can be modified in various ways. Some modified examples of the first embodiment will be described below. In each modified example of the first embodiment, differences from the first embodiment will be described. Explanations of the same items as the first embodiment will be omitted or will be briefly described.

[0174] (Variation 1) In the first embodiment, the SMR method is used as the recording method. In the first modification, the magnetic disk device 1 writes data using the CMR method.

[0175] FIG. 15 is a schematic diagram for explaining the CMR method used in the magnetic disk device 1 of the first modification. As shown in this figure, according to the CMR method, each track is arranged so as not to overlap with adjacent tracks in the radial direction. In other words, the CMR method is a method in which data of two tracks adjacent to each other in the radial direction of the magnetic disk 11 is written so as not to overlap each other. According to the CMR method, the width of each track is the same as the width (WHw) of the write element 22w, so data at any position can be updated. Therefore, according to the CMR method, although the recording density is lower than that of the SMR method, high random access performance can be obtained.

[0176] The controller 30 is capable of carrying out the operations described with reference to FIGS.

[0177] (Variation 2) In the second modification, the configuration of the long distance sector is different from that in the first embodiment.

[0178] 16 is a diagram for explaining an example of the configuration of a long distance sector according to Modification 2. Note that in this diagram, the servo sector SV is omitted.

[0179] In the second modification, the controller 30 collects a plurality of data units received from the host 2 into one data block, and encodes the data block using an error correction code. One long-distance sector has a capacity capable of storing one encoded data block.

[0180] The magnetic disk 11 is provided with long-distance sectors in which parity is stored. The controller 30 generates parity based on a plurality of encoded data blocks stored in a plurality of long-distance sectors. The controller 30 then stores the generated parity in a specific long-distance sector.

[0181] The parity may be generated based on a plurality of data blocks stored in a plurality of long-distance sectors included in one track 41. The parity may also be generated based on a plurality of data blocks stored in a plurality of long-distance sectors included in a plurality of tracks 41. In the example shown in FIG. 16, the parity is generated based on five data blocks stored in five long-distance sectors #0 to #4 included in track #q and track #(q+1). The generated parity is stored in the long-distance sector immediately following long-distance sector #4.

[0182] Even if the long range sector has the configuration illustrated in FIG. 16, the controller 30 can still perform the operations described with reference to FIG.

[0183] In the example shown in Fig. 11, the controller 30 calculates the squeeze evaluation amount SQ or the predicted value SQE of the squeeze evaluation amount for each data sector DS. In the third modification, the controller 30 calculates the squeeze evaluation amount SQ or the predicted value SQE of the squeeze evaluation amount for each servo sector SV included in the long distance sector. In the second modification, the controller 30 may obtain the squeeze amount as the squeeze evaluation amount, or may obtain the squeeze evaluation amount by performing any processing on the squeeze amount.

[0184] (Variation 3) As described above, according to the SMR method, the position of each track 41 is set so that the track width TW is narrower than that of the CMR method. Therefore, the DDOL is adopted so that the actual track pitch does not become narrower than necessary.

[0185] When DDOL is used, a position error signal of the previous track is required to set DDOL, which means that the controller 30 needs to record the position error signal of the previous track in any one of the storage areas.

[0186] On the other hand, the CMR method does not require such strict control of the track pitch as the SMR method, since the track width TW is wider than that of the SMR method. In other words, instead of the DDOL, a DOL that is fixedly set based on the design position of each track 41, that is, the ideal radial position, can be used. When the DOL is set based on the design position of each track 41, the controller 30 does not need to record other position error signals of the track 41 that is the access target.

[0187] In the third modification, a controller 30 configured to write data using the CMR method and set the DOL based on the design position of each track 41 will be described.

[0188] 17 is a diagram for explaining the setting of DOL when writing to track #k in the third modification. As shown in the figure, when writing to track #k, the controller 30 sets the DOL at a position offset by a predetermined distance from the set track center of track #(k-1). k Set.

[0189] FIG. 18 is a diagram for explaining a write operation to long distance sector #m on track #k in the third modification.

[0190] The controller 30 writes data in sequence from the beginning of the long-distance sector #m on the track #k. The controller 30 calculates the sum SQsum1 of the squeeze evaluation amounts and compares the sum SQsum1 of the squeeze evaluation amounts with a threshold value Th1 every time the magnetic head 22 passes over a servo sector SV.

[0191] Here, in the third modification, the controller 30 uses the squeeze amount as the squeeze evaluation amount. The amount of movement of the magnetic head 22 from the set track center of the track #k toward the outer periphery corresponds to the squeeze amount. Therefore, the controller 30 obtains the amount of movement of the magnetic head 22 from the set track center of the track #k toward the outer periphery as the squeeze evaluation amount.

[0192] 18, writing is performed up to just before the servo sector SV#(p+2), and the magnetic head 22 has completed reading the servo information from the servo sector SV#(p+2). Therefore, the controller 30 calculates the squeeze evaluation quantity SQ p , the squeeze evaluation quantity SQ in servo sector SV#(p+1) p+1 , and the squeeze evaluation quantity SQ in servo sector SV#(p+2) p+2 is acquired based on the servo information read from each servo sector SV. For the servo sector SV#(p+3) that has not yet been passed, the controller 30 estimates the squeeze evaluation amount in the same manner as in the first embodiment.

[0193] The controller 30 calculates the squeeze metric SQ p , Squeeze valuation quantity SQ p+1 , Squeeze valuation quantity SQ p+2 , and the squeeze evaluation quantity estimate SQE in servo sector SV#(p+3) p+3 When these values ​​are acquired, the controller 30 calculates the sum SQsum1 of the squeeze evaluation quantities. Then, the controller 30 compares the sum SQsum1 of the squeeze evaluation quantities with a threshold value Th1.

[0194] If the sum SQsum1 of the squeeze evaluation quantities does not reach the threshold Th1, the controller 30 continues the write operation on the write-target long-distance sector #m.

[0195] In the example shown in FIG. 18, the sum SQsum1 of the squeeze evaluation quantities exceeds the threshold Th1, which causes the controller 30 to execute a protective operation.

[0196] FIG. 19 is a diagram for explaining an example of a protection operation according to the third modification.

[0197] When the magnetic head 22 reaches the servo sector SV#(p+2), if the sum SQsum1 of the squeeze evaluation amounts exceeds the threshold Th1, the controller 30 executes a protective operation. k Tighten.

[0198] Specifically, as shown in FIG. 19, the controller 30 performs DOL k This moves the DOL k In the portion where the track width is reduced, the track width of the previous track is prevented from being narrowed, and as a result, it is possible to prevent the data unit set written in the long distance sector #m of the previous track from becoming difficult to read. In other words, the data unit set written in the long distance sector #m of the previous track is protected.

[0199] Second embodiment In the description of the second embodiment, the same reference numerals as those in the first embodiment are used for the respective components, and the same matters as those in the first embodiment will be described briefly or omitted.

[0200] Error correction using parity, i.e., second error correction, is performed when the first error correction fails.

[0201] Specifically, the controller 30 reads the data unit to be read from the magnetic disk 11, and executes the first error correction on the read data unit. If the first error correction on the data unit to be read fails, the controller 30 reads data for one track 41 including the data unit to be read and the parity, and executes error correction using the parity, i.e., the second error correction, on the data unit to be read.

[0202] In the second embodiment, as in the case of the first error correction, in order to estimate as accurately as possible whether error-free data can be obtained by the second error correction, the controller 30 makes a judgment based on the total squeeze evaluation amount based on all of the multiple servo sectors SV provided on the track 41, rather than on a portion of the multiple servo sectors SV, i.e., the total squeeze evaluation amount for each track 41.

[0203] The sum of the squeeze evaluation amounts is compared with a threshold value (denoted as threshold value Th2), and a protective action is executed based on the result of the comparison. As an example of the protective action, in the second embodiment, DDOL tightening is executed.

[0204] The ability of the second error correction depends on the number of sectors (referred to as error sectors) that cannot be corrected by the first error correction and the extent to which the degree of error in the error sector exceeds the ability of the first error correction. For example, even if the number of error sectors is only two and the squeeze evaluation amount is 10 nm, if there are 20 error sectors, the squeeze evaluation amount may be less than 5 nm.

[0205] Therefore, in the second embodiment, the controller 30 determines the threshold value Th2 based on the number of error sectors. Note that the total value of the squeeze evaluation amount in units of tracks 41 is greater than the total value of the squeeze evaluation amount in units of long distance sectors, so that the threshold value Th2 is greater than the threshold value Th1.

[0206] FIG. 20 is a diagram showing an example of a write operation performed by the controller 30 according to the second embodiment on the write target track 41.

[0207] The controller 30 calculates the number of error sectors passed during a write operation on the write target track 41 and the total value of the squeeze evaluation amount in those error sectors (S301).

[0208] In one example, the error sector is a data sector included in a long distance sector in which the sum SQsum1 of the squeeze evaluation amount exceeds the threshold value Th1 in the track 41. The error sector may be a long distance sector in which the sum SQsum1 of the squeeze evaluation amount exceeds the threshold value Th1.

[0209] Next, the controller 30 predicts the number of error sectors that have not yet passed and their squeeze evaluation values ​​(S302).

[0210] In the case of long distance sectors, in the example shown in the figure, the number of error sectors increases by four every time the number of long distance sectors whose sum value SQsum1 of the squeeze evaluation amount exceeds the threshold value Th1 increases by one. That is, in the write operation for the write target track 41, the number of error sectors that have passed and the number of error sectors that have not yet passed may both increase by four. When a parity sector is included, the number of error sectors may increase by five.

[0211] The controller 30 calculates the total number of error sectors in the entire write target track 41 and the total squeeze evaluation amount (represented as SQsum3) in the entire write target track 41 (S303). The controller 30 also determines the threshold value Th2 based on the total number of error sectors in the entire write target track 41 (S304).

[0212] The controller 30 determines whether the sum SQsum3 of the squeeze evaluation amounts of all the write-target tracks 41 is greater than a threshold value Th2 (S305).

[0213] If the sum SQsum3 is greater than the threshold Th2 (S305: Yes), the controller 30 performs a DDOL tightening operation as an example of a protection operation (S306). Specifically, the controller 30 calculates a length L3 based on the following formula (3), and moves the DDOL to a position offset from the position error signal of the previous track by the length L3 toward the inner periphery. L3 = TPset - Cecc + Cov (3)

[0214] Cecc is an index value of the error correction capability by the second error correction.

[0215] If the sum SQsum3 is not greater than the threshold Th2 (S305: No), or after the process of S306, the controller 30 determines whether the magnetic head 22 is located on the outer circumferential side (in other words, the previous track side) of the DDOL (S307).

[0216] If the magnetic head 22 is located on the outer circumferential side of the DDOL (S307: Yes), it is estimated that the magnetic head 22 is in an off-track state. Therefore, the controller 30 stops writing and postpones (postpones) writing until the magnetic disk 11 rotates once (S308). Then, when the magnetic disk 11 rotates once and the magnetic head 22 again approaches the circumferential position where writing stopped, the controller 30 executes the process of S307 again.

[0217] If the magnetic head 22 is not located on the outer periphery side of the DDOL (S307: No), it can be assumed that the magnetic head 22 is in an on-track state. Therefore, the controller 30 executes a write. Specifically, the controller 30 first determines whether the last passed servo sector SV is the last servo sector SV included in the track 41 to be written (S309).

[0218] If the last servo sector SV passed is not the last servo sector SV included in the write target track 41 (S309: No), the controller 30 executes writing up to just before the next servo sector SV (S310). Then, control transitions to S301.

[0219] If the last passed servo sector SV is the last servo sector SV included in the write target track 41 (S309: Yes), the controller 30 executes writing up to the end of the write target track 41 (S311). Then, the write operation ends.

[0220] In the series of operations shown in FIG. 20, the sum SQsum3 is an example of a third sum.

[0221] In this way, according to the second embodiment, the controller 30 calculates the sum SQsum3 of the squeeze evaluation amounts of the entire write target track 41. Then, when the sum SQsum3 exceeds the threshold value Th2, the controller 30 executes a protection operation to protect the data of the previous track 41.

[0222] Therefore, as in the first embodiment, a decrease in efficiency of a write operation caused by frequent write retry operations is suppressed, which means that the magnetic disk device 1 can write data efficiently.

[0223] As in the first embodiment, in the second embodiment, a specific example of the protection operation is not limited to tightening the DDOL.

[0224] Furthermore, according to the second embodiment, the controller 30 calculates the total number of error sectors in all of the write target tracks 41. Then, the controller 30 calculates the threshold value Th2 based on the total number of error sectors in all of the write target tracks 41.

[0225] Therefore, it is possible to reduce the possibility that the second error correction will fail.

[0226] (Variation 4) When the DDOL is controlled to be tightened using the threshold value Th1 to ensure that an error-free data unit set can be obtained from the adjacent long-distance sector by the first error correction, the frequency of write retry operations may increase due to off-track caused by the DDOL after the tightening. In other words, the write performance may decrease.

[0227] In contrast, if the DDOL tightening control using the threshold value Th1 is not performed, the frequency of execution of the second error correction due to the failure of the first error correction increases, and the read performance decreases. Instead, the decrease in write performance caused by the high frequency of write retry operations is suppressed.

[0228] Therefore, the controller 30 according to the fourth modification is configured to be able to switch between the control of DDOL contraction using the threshold value Th1 and the control of DDOL contraction using the threshold value Th2. By switching between the two types of control for DDOL contraction, the controller 30 balances the degree of degradation in write performance and the degree of degradation in read performance, thereby achieving both high write performance and high read performance.

[0229] 21 is a flowchart for explaining the operation of switching DDOL control according to Modification 4. This diagram shows an example of the operation of switching DDOL control during a write operation to one track 41. That is, the series of operations shown in this diagram are executed individually for each track 41.

[0230] In the initial setting, the controller 30 is set to execute DDOL control using a threshold value Th1 (S401). That is, the controller 30 is set to tighten the DDOL according to the result of comparison between the sum SQsum1 of the squeeze evaluation amount in the long distance sector and the threshold value Th1.

[0231] When the controller 30 executes a write operation (S402), it determines whether or not a write stop due to off-track has occurred during the write operation (S403).

[0232] If a write stop occurs (S403: Yes), the controller 30 determines whether or not the current setting is to execute DDOL control using the threshold value Th2 (S404).

[0233] If the setting is not currently set to execute DDOL control using threshold value Th2 (S404: No), that is, if the setting is currently set to execute DDOL control using threshold value Th1, the controller 30 determines whether the number of times DDOL contraction has been executed has reached the set value (S405).

[0234] When the number of times that DDOL contraction has been performed reaches the set value (S405: Yes), the controller 30 changes the setting to execute DDOL control using threshold value Th1 (S406). That is, the controller 30 changes the setting to stop the control of contracting DDOL in response to the sum value SQsum1 exceeding threshold value Th1, and executes the control of contracting DDOL in response to the sum value SQsum3 exceeding threshold value Th2.

[0235] If no write stop has occurred (S403: No), the controller 30 ends the operation without switching the DDOL control.

[0236] If it is currently set to execute DDOL control using threshold value Th2 (S404: Yes), or if the number of times DDOL contraction has been executed has not reached the set value (S405: No), or after S406, control transitions to S403.

[0237] In this way, when the number of times that the sum value SQsum1 exceeds the threshold value Th1 reaches a set value, the controller 30 stops controlling the DDOL to be tightened in response to the sum value SQsum1 exceeding the threshold value Th1, and instead executes control to tighten the DDOL in response to the sum value SQsum3 exceeding the threshold value Th1.

[0238] Therefore, it is possible to achieve both high write performance and high read performance.

[0239] (Variation 5) In the initial setting, the controller 30 is set to execute DDOL control using a threshold value Th2. When the number of error sectors that have passed through exceeds a specified value during a write operation on the write target track 41, the controller 30 changes the setting so as to execute control to tighten the DDOL in response to the sum value SQsum1 exceeding the threshold value Th1.

[0240] (Variation 6) Although the configuration in which data is written in an area longer than the size of the unit transmitted to and from the host is described as a long distance sector, this proposal can also be applied to a configuration in which the host unit has a length spanning multiple servo sectors in the circumferential direction and data is written in the same unit. For example, host 2 transmits data in 16K sector units, and after interleaving, one long distance sector is composed of four data sectors DS.

[0241] Although some embodiments of the present invention have been described, these embodiments are presented as examples and are not intended to limit the scope of the invention. These novel embodiments can be implemented in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their modifications are included in the scope and spirit of the invention, and are included in the scope of the invention and its equivalents described in the claims. [Explanation of symbols]

[0242] 1 magnetic disk device, 2 host, 11 magnetic disk, 12 spindle motor, 15 actuator arm, 16 VCM, 21 motor driver IC, 22 magnetic head, 22r read element, 22w write element, 23 HDC, 24 head IC, 25 RWC, 26 processor, 27 RAM, 28 FROM, 29 buffer memory, 30 controller, 41 track, 42 ​​servo area, 43 data area.

Claims

1. a magnetic disk including a plurality of tracks, a plurality of servo sectors on which servo information is recorded being arranged at intervals in a circumferential direction on the plurality of tracks, each track being provided with a plurality of long distance sectors, each of the plurality of long distance sectors being an area in which data is written and having a circumferential length spanning two or more of the plurality of servo sectors; a magnetic head for writing data to and reading data from each of the plurality of long distance sectors; In a write operation to a first portion, the first portion being a portion on a first track that is one of the plurality of tracks and adjacent to one long distance sector of a second track adjacent to the first track, executing an acquisition operation for acquiring one or more evaluation quantities corresponding to a squeeze amount of the track pitch based on a set value of the track pitch, based on a track pitch in each of two or more first servo sectors, the two or more servo sectors being included in a range between two ends in a circumferential direction of the first portion; executing a protection operation for protecting data of the second track when a first sum value, which is a sum value of the one or more evaluation quantities, exceeds a first threshold value; A controller that runs A magnetic disk device comprising:

2. the track pitch is the distance between the track center of the first track and the track center of the second track; 2. The magnetic disk drive according to claim 1.

3. the acquiring operation includes calculating the evaluation amount for a second servo sector, which is a first servo sector through which the magnetic head has passed, based on a position error signal of at least the first track, among the two or more first servo sectors; 2. The magnetic disk drive according to claim 1.

4. the acquiring operation further includes predicting the evaluation amount for a third servo sector, which is a first servo sector that the magnetic head has not yet passed over, among the two or more first servo sectors.

4. The magnetic disk drive according to claim 3.

5. The controller performs a radial write permission range tightening in the protection operation.

2. The magnetic disk drive according to claim 1.

6. The controller: calculating a second sum value which is a sum value of one or more of the evaluation amounts based on all the first servo sectors corresponding to the second servo sector among the two or more first servo sectors; In the protection operation, a boundary of a radial write permission range is set at a radial position according to the second sum value.

5. The magnetic disk drive according to claim 4.

7. In the protection operation, the controller stops the write operation when the first sum exceeds the first threshold value.

2. The magnetic disk drive according to claim 1.

8. In the acquisition operation, the controller predicts the evaluation amount based on the first servo sector corresponding to the third servo sector based on the evaluation amount based on the first servo sector corresponding to the second servo sector.

5. The magnetic disk drive according to claim 4.

9. the second track includes a parity sector in which parity for error correction in a track unit is stored; The controller: performing the obtaining operation and the comparison of the first sum value with the first threshold value for each of a plurality of second long distance sectors that are the plurality of long distance sectors in the second track; calculating a third sum, the third sum being a sum of the evaluation values ​​over the entire second track; performing a protection operation to protect data of the second track when the third sum exceeds a second threshold; Further execute 2. The magnetic disk drive according to claim 1.

10. The controller: calculating a total number of error sectors in the entire second track; determining the second threshold based on the total number; and 10. The magnetic disk drive according to claim 9, further comprising:

11. The controller: performing radial write permission contraction if the first sum exceeds the first threshold; when the number of times that the first sum exceeds the first threshold reaches a set value, the control of tightening the write permission range in response to the first sum exceeding the first threshold is stopped, and the control of tightening the write permission range in response to the third sum exceeding the second threshold is executed.

10. The magnetic disk drive according to claim 9.

12. The controller: Calculating the total number of error sectors passed by the magnetic head; if the total number exceeds a prescribed value, tightening a write permission range in a radial direction in response to the first total number exceeding the first threshold value; 10. The magnetic disk drive according to claim 9, further comprising:

13. the controller calculates the total number of error sectors in the entire second track by incrementing the error sector count by the number of data sectors included in the long-distance sector; The magnetic disk drive according to claim 10.

14. A method for controlling a magnetic disk device having a magnetic disk including a plurality of tracks, a plurality of servo sectors on which servo information is recorded being arranged at intervals in a circumferential direction on the plurality of tracks, each track being provided with a plurality of long distance sectors, each of the plurality of long distance sectors being an area in which data is written and having a circumferential length spanning two or more of the plurality of servo sectors, comprising: The method includes a write operation to a first portion, the first portion being a portion on a first track that is one of the plurality of tracks and adjacent to one long distance sector of a second track that is adjacent to the first track, the write operation comprising: executing an acquisition operation, the acquisition operation being an operation of acquiring one or more evaluation quantities corresponding to a squeeze amount of the track pitch based on a set value of the track pitch, based on a track pitch in each of two or more first servo sectors, the two or more servo sectors being included in a range between two ends in a circumferential direction of the first portion; executing a protection operation for protecting data on the second track in response to a first sum value, which is a sum value of the one or more evaluation quantities, exceeding a first threshold value; The method includes:

15. The acquisition operation includes: calculating the evaluation amount based on at least a position error signal of the first track for a second servo sector that is a first servo sector that a magnetic head has passed among the two or more first servo sectors; predicting the evaluation amount for a third servo sector which is a first servo sector that the magnetic head has not yet passed among the two or more first servo sectors; The method of claim 14 further comprising:

16. performing a radial write permission range tightening in the protection operation; 15. The method of claim 14, comprising:

17. calculating a second sum value which is a sum value of one or more of the evaluation amounts based on all the first servo sectors corresponding to the second servo sector among the two or more first servo sectors; In the protection operation, setting a boundary of a radial write permission range at a radial position according to the second sum value; 16. The method of claim 15 further comprising:

18. When the first sum exceeds the first threshold value in the protection operation, the write operation is stopped.

15. The method of claim 14, comprising:

19. predicting, in the acquisition operation, the evaluation amount based on a first servo sector corresponding to the third servo sector, based on the evaluation amount based on a first servo sector corresponding to the second servo sector; 16. The method of claim 15 further comprising:

20. the second track includes a parity sector in which parity for error correction in a track unit is stored; performing the obtaining operation and the comparison of the first sum value with the first threshold value for each of a plurality of second long distance sectors that are the plurality of long distance sectors in the second track; calculating a third sum, the third sum being a sum of the evaluation values ​​over the entire second track; performing a protection operation to protect data of the second track when the third sum exceeds a second threshold; The method of claim 14 further comprising:

Images