JP2008299978A - Disk drive device and data reproducing method from disk - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To improve data read rate from a disk within limited retries. <P>SOLUTION: When reproducing a plurality of data sectors from a magnetic disk 11, an HDD 1 reads all the data sectors regardless of errors in ECC (Error Correction Code) of data read from the magnetic disk. Then, reproducing is retried only in a data sector having the ECC error. The processing time to accurately reproduce the plurality of data sectors can be reduced when the error occurs, and the data read rate can be increased with accuracy when the number of retries are limited. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a disk drive device and a method for reproducing data from a disk, and more particularly to retrying data reading from a disk.

  As data storage devices, devices using various types of media such as optical disks, magneto-optical disks, and flexible magnetic disks are known. Among them, hard disk drives (HDDs) are used as computer storage devices. It has become widespread and has become one of the storage devices that are indispensable in current computer systems. Furthermore, the use of HDDs such as a removable memory used in a moving image recording / reproducing apparatus, a car navigation system, a mobile phone, a digital camera, etc. is expanding more and more due to its excellent characteristics.

  Generally, a time from when the HDD receives a read command from the host to when data is transferred to the host is defined. Particularly in HDDs for AV equipment and the like, since it is necessary to read data within a short time, the number of retries is limited to a smaller number than usual. For this reason, the frequency of non-recoverable errors increases. In the case of uncompressed data, even if several data sectors are unrecoverable, there is no visual problem because it does not affect consecutive frames. For this reason, even if the data sector is an ECC error, there is a Read Continuous mode in which the data sector is transferred to the host without being corrected.

  However, the compression adopted by recent AV devices is mainly a compression method represented by MPEG, and stores the difference between the previous and next frames. For this reason, the loss of several data sectors has been visually affected. Recent AV system user demands are becoming stricter: "Request data is transferred within a unit time, and unrecoverable data is not allowed to be transferred". In fact, the situation is that “the improvement of the sustain transfer rate due to the high density of the HDD / the deterioration of the high density error rate”.

As the required transfer rate increases in the future as AV information becomes more high quality and diversified, "improvement of sustain transfer rate / deterioration of densification error rate due to higher density of HDD" will be very severe. It is expected to be. On the other hand, in a system having a RAID configuration or a system having a higher ECC in an AV server or the like, if the host device finds that the data sector is an unrecoverable data sector, The device can recover the error. Therefore, Patent Document 1 proposes a technique in which the HDD sends a bit map indicating whether each data sector has an ECC error to the host when the HDD transfers data to the host device.
JP 2001-331337 A

  With the above bit map, it is possible to accurately correct errors in the host device and reproduce more accurate data. However, in a device using a micro HDD such as a consumer use MP3 player, a mobile phone, or a Camcorder, it is common not to perform error correction using ECC at the upper level. As described above, when there is no error correction or RAID configuration at the upper level, it is required that the HDD reproduces as many data sectors as accurately as possible within a limited number of retries. Even if the host device has these error recovery means, it is preferable to reduce the number of error data sectors transferred to the host device as much as possible to reduce the processing burden on the host device.

  One aspect of the present invention is a disk drive device that reads data from a disk. The apparatus sequentially attempts to read a plurality of target data sectors using the head, a head for reading data from the magnetic disk, a buffer for storing data sectors read from the disk, and the head. A controller that stores sequentially read data sectors in a buffer and re-reads only data sectors that could not be accurately reproduced from the disk among a plurality of data sectors that were sequentially read; By trying again only the reading of data sectors that could not be reproduced correctly among the multiple data sectors that were sequentially read, the accurate data reproduction rate from the disk was increased within the limited number of retries. be able to.

Preferably, the controller includes a processor that operates according to firmware and a logic circuit that operates according to an instruction from the processor, and the logic circuit identifies the data sector that could not be accurately reproduced. A map is generated, and the data sector that could not be reproduced accurately by the map is tried again. This makes it possible to perform necessary processing without excessively demanding the processor.
Preferably, the plurality of target data sectors include a command-designated data sector and a different data sector. As a result, performance can be improved and an accurate data reproduction ratio can be increased.
The number of times of retrying reading of the different data sector is preferably equal to or less than the number of times of retrying reading of the data sector designated by the command. As a result, performance degradation can be suppressed.

In a preferred example, the controller stores a map that identifies data sectors stored in a buffer that are not accurately reproduced among the different data sectors, and stores the commanded data sectors to the host. After the transfer, re-reading of the data sector that could not be accurately reproduced among the different data sectors specified by the map is performed within the time free from the processing corresponding to the command. . As a result, an accurate data reproduction ratio can be increased while suppressing performance degradation.
In a preferred example, the buffer stores only correctly reproduced data sectors of the different data sectors. This can improve the cache hit rate while suppressing memory consumption.
In a preferred example, the buffer stores only the first consecutive data sector reproduced correctly among the different data sectors. This makes it possible to improve the cache hit rate while suppressing memory consumption with simpler control.

  A disk drive device according to another aspect of the present invention includes a head for reading data from the magnetic disk, a buffer for storing data sectors read from the disk, and data sectors sequentially read by the head. Regardless of whether or not the data was correctly reproduced, the sequentially read data sector is stored in the buffer, and a map for identifying the data sector that could not be accurately reproduced is generated and specified by the map. And a controller that attempts to read the data sector again. Accurate data from the disk within a limited number of retries by generating a map that identifies the data sector that could not be accurately reproduced and trying to read the data sector specified by the map again The reproduction ratio can be increased.

  Preferably, the map has information identifying data sectors that could not be correctly reproduced among commanded data sectors, and the buffer was reproduced correctly among the different data sectors. Save only the first continuous data sector. This makes it possible to improve the cache hit rate while suppressing memory consumption with simpler control.

  Another aspect of the present invention is a method for reproducing data from a disk in a disk drive device. This method sequentially attempts to read a desired plurality of data sectors from the magnetic disk. Data sectors sequentially read from the disk are stored in a buffer. Of the plurality of data sectors that have been sequentially read, only data sectors that could not be reproduced correctly are read again. By trying again only the reading of data sectors that could not be reproduced correctly among the multiple data sectors that were sequentially read, the accurate data reproduction rate from the disk was increased within the limited number of retries. be able to.

  According to the present invention, an accurate data reproduction ratio from a disk can be increased within a limited number of retries.

  Hereinafter, embodiments to which the present invention can be applied will be described. For clarity of explanation, the following description and drawings are omitted and simplified as appropriate. Moreover, in each drawing, the same code | symbol is attached | subjected to the same element and the duplication description is abbreviate | omitted as needed. In the following, a hard disk drive (HDD) that is an example of a disk drive device will be described.

  When reproducing a plurality of data sectors from the magnetic disk, the HDD of this embodiment reads all the data sectors regardless of the presence or absence of an ECC (Error Correction Code) error in the data read from the magnetic disk. Thereafter, reproduction is retried only for the data sector in which the ECC error has occurred. As a result, when an error occurs, the processing time for accurately reproducing a plurality of data sectors can be shortened. In particular, when the number of retries is limited, the accurate data reproduction rate is increased. Can do.

  First, the overall configuration of the HDD will be described. FIG. 1 is a block diagram schematically showing the overall configuration of the HDD 1. The HDD 1 includes a circuit board 20 that is fixed to the outside of the enclosure 10. On the circuit board 20, a read / write channel (RW channel) 21, a motor driver unit 22, a hard disk controller (HDC) and an MPU integrated circuit (HDC / MPU) 23, a semiconductor memory RAM 24, etc. It has a circuit. Within the enclosure 10, a spindle motor (SPM) 14 rotates the magnetic disk 11 at a predetermined angular velocity. The magnetic disk 11 is a disk for storing data. The motor driver unit 22 drives the SPM 14 according to control data from the HDC / MPU 23.

  Each head slider 12 includes a slider that floats on the magnetic disk, and a head element unit that is fixed to the slider and performs conversion (data read / write) between a magnetic signal and an electric signal. Each head slider 12 is fixed to the tip of the actuator 16. The actuator 16 is connected to a voice coil motor (VCM) 15, and moves in the radial direction on the magnetic disk 11 that rotates the head slider 12 by rotating about a rotation axis. The motor driver unit 22 drives the VCM 15 according to control data from the HDC / MPU 23. An arm electronic circuit (AE: Arm Electronics) 13 selects a head slider 12 that accesses (reads or writes) the magnetic disk 11 from a plurality of head element units 12 according to control data from the HDC / MPU 23, and Amplifies the write signal.

  In the read process, the RW channel 21 extracts servo data and user data from the read signal acquired from the AE 13 and performs a decoding process. The decoded data is supplied to the HDC / MPU 23. In the write process, the RW channel 21 code-modulates the write data supplied from the HDC / MPU 23, converts the code-modulated data into a write signal, and supplies the write signal to the AE 13. In the HDC / MPU 23, the HDC is a logic circuit, and the MPU operates according to the firmware loaded in the RAM 24. As the HDD 1 is activated, data required for control and data processing is loaded into the RAM 24 from the magnetic disk 11 or ROM (not shown). The HDC / MPU 23 is an example of a controller, and performs overall control of the HDD 1 in addition to necessary processing relating to data processing such as head positioning control, interface control, and defect management.

  The HDC / MPU 23 transfers the read data from the magnetic disk 11 acquired from the RW channel 21 to the host 51. Read data from the magnetic disk 11 is temporarily stored in a read buffer in the RAM 24 and then transferred to the host 51 via the HDC / MPU 23. The HDC / MPU 23 according to the present embodiment reads data from the magnetic disk 11 even when an ECC (Error Correcting Code) error occurs in the data sector in the reproduction process of a plurality of data sectors from the magnetic disk 11 in particular. Read subsequent data sectors without interruption. In the following, reproduction is processing including reading from the magnetic disk 11 and error correction processing of the data.

  For example, as shown in FIG. 2, when the HDC / MPU 23 reproduces the data sector 10 from the data sector 1, an ECC error occurs in the data sector 6. The conventional HDC / MPU 23 interrupts reading of the remaining data sectors 7 to 10 when an error occurs in the data sector 6 and starts an error recovery process for reproducing the data sector 6. The error recovery process retries reading by maintaining or changing the reading conditions in a preset procedure. The HDC / MPU 23 of this embodiment executes the subsequent reproduction of the data sectors 7 to 10 regardless of the error of the data sector 6. Thereafter, the HDC / MPU 23 re-reads the data sector 6 and performs necessary error correction processing.

  Details of the data reproduction processing of the HDC / MPU 23 will be described with reference to the block diagram of FIG. The HDC / MPU 23 includes an HDC 232 configured by a hardware circuit (logic circuit) and an MPU 231 that is a processor that operates according to firmware. The HDC 232 includes a host interface controller (HIC) 321, a memory manager 322, a drive manager 323, an ECC correction processing circuit 324 as an error correction processing unit, an SRAM 325, and a bit map 326 stored in a register. Bit map 326 indicates whether each of the target data sectors is being correctly reproduced.

  With reference to FIG. 3, the flow of transfer data in the data recording process and the reproduction process will be described. In FIG. 3, solid arrows indicate the flow of data transferred from the host 51 and recorded on the magnetic disk 11. A dotted arrow indicates a flow of data read from the magnetic disk 11 and transferred to the host 51. In each reproduction and recording process, data is transferred and processed in units of data sectors. Here, the data sector is a recording unit of user data on the magnetic disk 11.

  The data flow in the recording process will be described. The HIC 321 is an interface circuit between the host 51 and the HDD 1. In the recording process, the HIC 321 sends the user data transferred from the host 51 to the memory manager 322 via the data bus. The memory manager 322 temporarily stores the received user data in the buffer 241 in the DRAM 24. Thereafter, when there is a request from the drive manager 323, the data is read from the buffer 241 and transferred.

  The drive manager 323 sends the received data to the ECC correction processing circuit 324. The ECC correction processing circuit 324 generates an ECC from the data received from the drive manager 323. The ECC is a code for correcting an error in data recording on the magnetic disk 11. Although an explanation will be given assuming that an ECC and a CRC (Cyclic Redundancy Check) code are added as the ECC, only the ECC may be used. As the ECC and CRC codes, for example, Reed-Solomon codes can be used.

  The drive manager 323 acquires the ECC from the ECC correction processing circuit 324 and generates user data with ECC. In this way, the drive manager 323 supplies the user data to which the ECC is added to the RW channel 21 as NRZ (Non-Return to Zero) data via the data bus.

  Next, the flow of data transfer in the reproduction process will be described with reference to FIG. If there is no cache hit for the read command from the host 51, the corresponding data is read from the magnetic disk 11. In a read operation for reading data from the magnetic disk 11, the drive manager 323 sequentially receives data sectors to which ECC is added from the RW channel 21, and supplies this to the ECC correction processing circuit 324.

  The ECC correction processing circuit 324 stores the data sector received from the drive manager 323 in the SRAM 325 and executes necessary error correction processing. The ECC correction processing circuit 324 executes error correction processing for each data sector. Error correction processing is performed on the fly and can be performed without interrupting data transfer. Here, on-the-fly processing refers to performing processing without interrupting data transfer. The data sector that has undergone necessary error correction is transferred to the memory manager 322. The memory manager 322 temporarily stores the received data in the buffer 241, and then reads the data from the buffer 241 and transfers it to the HIC 321 in accordance with an instruction from the MPU 231. The HIC 321 transfers the data to the host 51.

  The ECC correction processing circuit 324 of the present embodiment transfers the data sector to the memory manager 322 even when the acquired data sector cannot be error-corrected by ECC. Therefore, in this embodiment, the transfer of each target data sector to the buffer 241 is not interrupted by an ECC error. The ECC correction processing circuit 324 sets a bit corresponding to each data sector in the bit map 326. Each bit indicates whether or not each data sector is correctly reproduced. For example, “0” indicates a data sector that has been correctly reproduced, and “1” indicates a data sector that has not been correctly reproduced.

  When the data of all target data sectors (including data that has not been correctly reproduced) is stored in the buffer 241, the MPU 231 executes a reproduction retry for the data sectors that have not been correctly reproduced. . Specifically, the MPU 231 refers to the bit map 326 and checks whether there is an error in each data sector. If even one error sector exists, the HDC 232 is instructed to retry. In response to an instruction from the MPU 231, the drive manager 323 refers to the bit map 326 and reads out an error sector from the magnetic disk 11. Subsequent processing is the same as described above. In this way, the reproduction of the data sector in which an error has occurred (transfer from the magnetic disk 11 to the buffer 241) is executed by the HDC 232 having a hardware configuration, not by the MPU 231, so that the MPU 231 is requested to have a large processing capacity. The processing can be executed without any problem.

  Note that the error sector may be selected by the memory manager 322 referring to the bit map 326, unlike the configuration performed by the drive manager 323 as described above. In this case, the memory manager 322 selects an error sector with reference to the bit map 326 from the data sectors sequentially transferred from the ECC correction processing circuit 324 and stores the data sector in the buffer 241. To do. When transferring data to the host 51, the HDD 1 also transfers the corresponding bit map 326. The host 51 uses this to perform necessary data recovery.

  Next, user data reproduction processing will be described with reference to the flowchart of FIG. The MPU 231 starts reading a plurality of target data sectors (S11). That is, the address of the target data sector is designated to the HDC 232 and reproduction is instructed. When a servo error occurs when reading user data from the magnetic disk 11 (Y in S12), the HDC 232 interrupts reading from the magnetic disk 11 and notifies the MPU 231 of the interruption. Thereby, the reproduction process is temporarily ended (S18).

  When an ECC error occurs in the data sector read from the magnetic disk 11 (Y in S13), the HDC 232 continues processing such as data reading from the magnetic disk 11 and error correction as described with reference to FIG. (S14). The data sector that has undergone the necessary processing is stored in the buffer 241 (S15). The ECC correction processing circuit 324 sets a bit in the bit map 326.

  When all the data sectors are stored in the buffer 241 without causing a servo error, the MPU 231 refers to the bit map 326 to determine whether all the data sectors have been correctly reproduced (S16). If any data sector is not correctly reproduced (N in S16), the HDC / MPU 23 performs a retry. There is typically an upper limit on the number of retries. Specifically, there is a case where the upper limit is determined by limiting the time allowed for reproduction, or the upper limit is set as the number of retries. In the example of FIG. 4, the MPU 231 determines whether the number of times has reached the upper limit number before starting a retry (S17). Typically, there is an upper limit on the number of retries for the corresponding command. If the number of retries has not reached the upper limit, the MPU 231 instructs the start of re-reading of the error sector (S19). Thereafter, the same processing is repeated.

  In the example of FIG. 4, the HDC / MPU 23 interrupts the reproduction process in response to the occurrence of a servo error. However, as shown in FIG. 5, even if a servo error occurs and the data sector cannot be read from the magnetic disk 11, the HDC / MPU 23 may continue the reproduction process without interruption. . The HDC / MPU 23 tries to reproduce to the last data sector, and then retries the reproduction of all data sectors that have not been reproduced correctly, including the data sector that could not be read from the magnetic disk 11.

  A specific example of data reproduction processing from the magnetic disk 11 will be described below. In the model of FIG. 6, the HDD 1 plays back data sectors 0 to 15 in response to an AV mode read command from the host 51. X indicates a data sector in which an ECC error has occurred. In addition, the upward x order in FIG. 6 matches the retry order. That is, in the first reading, data sectors 0 to 2 and 6 to 10 cause an ECC error. For example, the data sectors 0 to 2 are more likely not to be accurately reproduced by the reproduction read because the head slider 12 is unstable. The other data sectors are correctly reproduced by the first reading.

  In the next read (first retry), data sectors 4, 8, 12, and 13 cause an ECC error. Since FIG. 6 is a model diagram, it shows the result when the ECC process is performed, and does not indicate whether or not it is actually performed. In the third and fourth reading (second and third retry), the data sectors 7, 8, and 12 cause an ECC error. Data sector 8 causes an error no matter how many times it is read again. Other data sectors can ultimately be reproduced accurately.

  An example of conventional reproduction processing will be described with reference to FIG. In the AV command, it is assumed that the number of reads is limited to four. On the first read, data sector 0 causes an error. The HDD stops reading and executes data sector 0 error recovery processing (retry). In the first retry, data sectors 0 to 3 can be accurately reproduced, but data sector 4 causes an error. The HDD interrupts reading and starts error recovery processing for the data sector 4. At this time, data sectors 0 to 3 exist in the buffer.

  In the second retry, data sectors 4 to 6 are correctly reproduced, but data sector 7 causes an error. In the last read, which is the third retry, the HDD stores each data sector in the buffer regardless of the presence or absence of an ECC error. As a result, data sectors 7, 8, and 12 become error sectors. The HDD generates a bit map and transfers it to the host together with the reproduced data. The host performs error recovery using ECC or RAID parity.

  Next, an example of processing according to the present invention will be described with reference to FIG. The HDC / MPU 23 accurately reproduces the data sectors 3 to 5 and 11 to 15 in the first reproduction process. At the first retry, data sectors 0 to 3 and 6 to 10 are tried to be reproduced. In the first retry, all data sectors other than data sector 8 are correctly reproduced. Thereafter, the HDC / MPU 23 executes the second and third retries. Data sector 8 is not accurately reproduced to the end, and bit map 326 indicates that only data sector 8 is an error sector. The HDC / MPU 23 transfers the data sectors 0 to 15 and the bit map 326 to the host 51. Thus, in the conventional method, data sectors 7, 8 and 12 are error sectors, but in the method of the present invention, only data sector 8 is an error sector.

  Next, another example model will be described. The description method of FIG. 9 is the same as that of FIG. The upper limit number of retries in the error recovery process is three. An example of conventional reproduction processing will be described with reference to FIG. On the first read, data sector 0 causes an error. The HDD stops reading and executes data sector 0 error recovery processing (retry). In the first retry, data sectors 0 to 3 can be accurately reproduced, but data sector 4 causes an error. The HDD interrupts reading and starts error recovery processing for the data sector 4.

  In the second retry, data sectors 4 to 6 are correctly reproduced, but data sector 7 causes an error. In the last read, which is the third retry, the HDD stores each data sector in the buffer regardless of the presence or absence of an ECC error. As a result, data sectors 7, 8, and 12 become error sectors. The HDD generates a bit map and transfers it to the host together with the reproduced data. The host performs error recovery using ECC or RAID parity.

  Next, an example of processing according to the present invention will be described with reference to FIG. The HDC / MPU 23 accurately reproduces the data sectors 3 to 5 and 11 to 15 in the first reproduction process. At the first retry, data sectors 0 to 3 and 6 to 10 are tried to be reproduced. In that first retry, all remaining data sectors are correctly reproduced. Bit map 326 indicates that all data sectors are correct data sectors.

  The HDC / MPU 23 transfers the data sectors 0 to 15 and the bit map 326 to the host 51. Thus, in the conventional method, the data sectors 7, 8 and 12 are error sectors, but in the method of the present invention, there is no error sector. As can be understood from the above description, the method of the present invention is particularly effective when an error does not always occur in a data sector, but an error does not always occur in the data sector.

  Next, the operation timing in the HDC / MPU 23 in the reproduction processing of this embodiment will be described. FIG. 12 shows a timing chart when all sectors of data sectors S to S + 7 can be accurately reproduced. S to S + 7 indicate data sector addresses, respectively. The sector address SCTAdr is data read at the position of each sector of the magnetic disk 11. When the SCTtoRead signal is asserted, the data sector read from the magnetic disk 11 is taken into the HDC 232. The SCTtoRead signal is asserted when each bit in the bit map indicates 1. The ReadGate signal that controls reading of data from the magnetic disk 11 is asserted for each data sector when the SCTtoRead signal is asserted.

  In the first read process, all data sectors are read. All bits of the bit map BitMap (1) are 1, and the SCTtoRead signal and ReadGate signal are asserted in all data sectors. Thereafter, each data sector is stored in the address indicated by the cache address CacheAdr in the buffer 241. In this example, since all data sectors are accurately reproduced, all bits of the bit map BitMap (2) after buffer storage are zero.

  Next, an example in which data sectors S + 2, S + 4, and S + 6 cause an ECC error will be described with reference to FIG. The first reading from the magnetic disk 11 reads all data sectors S to S + 7. Of the read data sectors, data sectors S + 2, S + 4, and S + 6 cannot be corrected by the ECC, and fail. Therefore, the third, fifth and seventh bits are 1 in the bit map BitMap (2) after buffer storage.

  FIG. 14 is a timing chart of the first retry after reading in FIG. The bit map BitMap (1) in reading is a copy of the bit map BitMap (2) in FIG. 13 and the third, fifth and seventh bits are 1. The SCTtoRead signal and ReadGate signal change according to this bit map BitMap (1), are asserted in each of the data sectors S + 2, S + 4, and S + 6, and are deassertd in the other data sectors. Of the data sectors S + 2, S + 4, and S + 6 read from the magnetic disk 11, the data sector S + 4 again indicates an error. For this reason, the fifth bit is set to 1 in the bit map BitMap (2) after buffer storage.

  FIG. 15 is a timing chart of the second retry. The bit map BitMap (1) in reading is a copy of the bit map BitMap (2) in FIG. The SCTtoRead signal and ReadGate signal change according to this bit map BitMap (1), are asserted in the data sector S + 4, and are deasserted in the other data sectors. In this retry, the data sector S + 4 is accurately reproduced, and the fifth bit is set to 0 in the bit map BitMap (2) after storing the buffer.

  The HDD 1 reproduces data in accordance with a read command from the host 51. The data sector to be reproduced can include a data sector before or after the address in addition to the data sector specified by the read command. In a typical HDD, the data sector following the command designation address is also read from the magnetic disk 11 and stored in a buffer in advance. As a result, the cache hit rate of subsequent commands can be increased. Such a process is called a prefetch process, which is particularly useful for a command that instructs to reproduce a long continuous data sector, such as an AV command.

  A retry method in reproducing a data sector different from the data sector designated by the command, such as the pre-read data sector, can be the same as that of the data sector designated by the command. However, even though all data sectors specified by the command are correctly reproduced, repeating the retry for reproducing different data sectors causes a decrease in performance. Therefore, when the command-designated data sector is correctly reproduced, it is preferable to end the reproduction process without retrying for another data sector. Alternatively, it is preferable that the upper limit number of retries for data sectors not designated by the command is equal to or less than the upper limit number of data sectors designated by the command.

  Some of the data sectors that are not commanded may not be reproduced correctly within a limited number of retries. FIG. 16A shows an example in which an ECC error has occurred in a part of the pre-read data sector. The command designated data sectors are data sectors 1 to 4, and the pre-read data sectors are data sectors 5 to 10. FIG. 16B shows that when the HDD 1 transfers the command designated data sector to the host 51, the data sectors 7 and 8 are error sectors.

  The pre-read data sector needs to be held in the buffer 241 in preparation for a future command. On the other hand, when the error sector is stored in the buffer 241, it is necessary to store it in the bit map 326 memory indicating which data sector is the error sector. If there is a margin in the semiconductor memory area, the error sectors 7 and 8 are stored in the buffer 241 and the HDC / MPU 23 is released from processing corresponding to the command from the host 51 and is in an idle state. It is preferable to perform error recovery processing for the error sectors 7 and 8 when the error occurs. When the error is recovered, the bit map indicating the error becomes unnecessary and can be erased from the memory.

  It is preferable that the number of saved data sectors or bit maps different from the data sector designated by the command is equal to or less than a preset upper limit number. When the number of bit maps is limited, the number of data sectors stored is also limited. As a result, it is possible to prevent the memory area from being consumed by unnecessary data sectors or bit maps.

  Alternatively, as shown in FIG. 16 (c), data sectors 5, 6, 9, and 10 that are accurately reproduced among the pre-read data sectors can be stored in the buffer 241. That is, the area of the data sectors 7 and 8 can be overwritten by other data. By storing only accurate data sectors in the buffer, it is not necessary to store a bit map indicating an error sector in the memory, and a memory area can be secured.

  Alternatively, only the first continuous data that has been accurately reproduced in the pre-read data sector can be stored in the buffer 241. In the example of FIG. 16D, the buffer 241 stores data sectors 5 and 6 and does not store error sectors 7 and 8 and accurate data sectors 9 and 10. By doing so, the cache hit rate can be improved without requiring complicated control processing.

  Although the present invention has been described using preferred embodiments, the present invention is not limited to the above-described embodiments, and various modifications can be made without departing from the scope of the present invention. Of course. For example, in the above-described embodiment, the HDD has been described as an example. However, the present invention is applied to a disk drive device using another disk such as an optical disk or a magneto-optical disk, or a disk drive apparatus to which the disk can be attached or detached. Also good. The MPU 231 may change the retry method according to the mode of the command from the host 51. Specifically, the HDD 1 may execute a retry of the present invention in the case of an AV command, and may execute a retry for each data sector in the case of another command.

1 is a block diagram schematically showing an overall configuration of a hard disk drive according to an embodiment. It is a figure which shows typically the example of the reproduction | regeneration processing concerning this Embodiment. In this Embodiment, it is a block diagram which shows typically the component which performs a reproduction | regeneration process. It is a flowchart which shows the flow of the reproduction | regeneration processing concerning this Embodiment. It is a flowchart which shows the flow of the reproduction | regeneration processing concerning this Embodiment. It is a figure which shows typically the precondition of an example of the reproduction | regeneration processing concerning this Embodiment. It is a figure which shows typically the example of the conventional reproduction | regeneration processing in the precondition of FIG. It is a figure which shows typically the reproduction | regeneration processing concerning this embodiment in the precondition of FIG. It is a figure which shows typically the precondition of the other example of the reproduction | regeneration processing concerning this Embodiment. It is a figure which shows typically the example of the conventional reproduction | regeneration processing in the precondition of FIG. It is a figure which shows typically the reproduction | regeneration processing concerning this embodiment in the precondition of FIG. It is a timing chart explaining the operation timing in HDC / MPU in the reproduction | regeneration processing concerning this Embodiment. It is a timing chart explaining the operation timing in HDC / MPU in the reproduction | regeneration processing concerning this Embodiment. It is a timing chart explaining the operation timing in HDC / MPU in the reproduction | regeneration processing concerning this Embodiment. It is a timing chart explaining the operation timing in HDC / MPU in the reproduction | regeneration processing concerning this Embodiment. In this Embodiment, it is a figure which shows typically the preservation | save data in the prefetch data sector.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Hard disk drive, 10 Housing, 11 Magnetic disk 12 Head slider, 13 Arm electronics, 14 Spindle motor 15 Voice coil motor, 16 Actuator, 20 Circuit board 21 RW channel, 22 Motor driver unit 23 Hard disk・ Controller / MPU, 24 DRAM
311 Host interface controller 231 MPU, 232 Hard disk controller 241 Buffer, 321 Host interface controller 322 Memory manager, 323 Drive manager 324 ECC correction processing circuit, 325 SRAM, 326 bit map

Claims (19)

A disk drive device for reading data from a disk,
A head for reading data from the magnetic disk;
A buffer for storing data sectors read from the disk;
Sequential attempts to read a plurality of data sectors of interest using the head, store the data sectors sequentially read from the disk in a buffer, and accurately read from the disk among the plurality of data sectors that were sequentially read. A controller that only tries to read data sectors that could not be replayed;
A disk drive device. The controller includes a processor that operates according to firmware, and a logic circuit that operates according to an instruction from the processor,
The logic circuit generates a map that identifies the data sector that could not be accurately reproduced, and re-reads the data sector that could not be accurately reproduced according to the map;
The disk drive device according to claim 1. The target plurality of data sectors includes a command specified data sector and a different data sector.
The disk drive device according to claim 1. The number of times to retry reading the different data sector is less than or equal to the number of times to retry reading the data sector specified by the command,
The disk drive device according to claim 3. The controller stores a map identifying data sectors stored in a buffer that are not accurately reproduced among the different data sectors, and transfers the command-specified data sectors to the host; Re-reading of data sectors that could not be accurately reproduced among the different data sectors specified by the map within the time freed from processing corresponding to the command;
The disk drive device according to claim 3. The buffer stores only correctly reproduced data sectors of the different data sectors;
The disk drive device according to claim 3. The buffer stores only the first consecutive data sector reproduced correctly among the different data sectors.
The disk drive device according to claim 3. A disk drive device for reading data from a disk,
A head for reading data from the magnetic disk;
A buffer for storing data sectors read from the disk;
Regardless of whether or not the data sector sequentially read out by the head can be accurately reproduced, the data sector read out sequentially is stored in the buffer and cannot be reproduced accurately. Generating a map identifying the data, and retrying to read the data sector identified by the map;
A disk drive device. The sequentially read data sectors include a command-specified data sector and a different data sector.
The disk drive device according to claim 8. The number of times to retry reading the different data sector is less than or equal to the number of times to retry reading the data sector specified by the command,
The disk drive device according to claim 9. The controller performs re-reading of data sectors that could not be accurately reproduced among the different data sectors specified by the map within a time free from processing corresponding to a command.
The disk drive device according to claim 9. The buffer stores only correctly reproduced data sectors of the different data sectors;
The disk drive device according to claim 9. The map has information that identifies data sectors that could not be reproduced correctly among commanded data sectors;
The buffer stores only the first consecutive data sector that has been accurately reproduced among the different data sectors.
The disk drive device according to claim 9. A method for reproducing data from a disk in a disk drive device,
Sequential attempts to read a desired plurality of data sectors from the magnetic disk,
Store data sectors sequentially read from the disk in a buffer;
Of the plurality of data sectors that have been read sequentially, only the data sectors that could not be accurately reproduced are retried.
Method. The target plurality of data sectors includes a command specified data sector and a different data sector.
The method according to claim 14. The number of times to retry reading the different data sector is less than or equal to the number of times to retry reading the data sector specified by the command,
The method of claim 15. Storing a map identifying the data sectors stored in the buffer that are not accurately reproduced among the different data sectors;
Re-reading of data sectors that could not be accurately reproduced among the different data sectors specified by the map within the time freed from processing corresponding to the command;
The method of claim 15. Of the different data sectors, only those data sectors that have been accurately reproduced within a predetermined number of re-reads are stored in the buffer.
The method of claim 15. Only the first consecutive data sector reproduced correctly among the different data sectors is stored in the buffer,
The method of claim 15.

Images