Patent Application
20080270672
The instant nonprovisional patent application claims priority to Japanese Patent Application No. 2007-115986 filed Apr. 25, 2007 and which is incorporated by reference in its entirety herein for all purposes.
Disk drive devices using various kinds of disks, such as an optical disk, a magneto-optical disk, and a flexible magnetic disk, have been known in the art. In particular, a hard disk drive (HDD) has been widely used as a storage device of a computer and has been one of indispensable storage devices for current computer systems. Moreover, the HDD has found widespread application such as a removable memory used in a moving image recording/reproducing apparatus, a car navigation system, a cellular phone, or a digital camera, as well as the computer, due to its outstanding characteristics.
Magnetic disks used in the HDD have a plurality of data tracks formed concentrically. Each data track has a plurality of data sectors recorded thereon. On the magnetic disk, a plurality of servo data are discretely recorded in the circumferential direction. A head element portion of a head slider supported by a swinging actuator accesses a desired data sector according to address information of the servo data, which allows data writing to and data retrieving from a data sector.
If a data sector on the recording surface is a defective sector and data writing thereto or data retrieving therefrom cannot be performed, the sector is reassigned to another sector. That is, if a host accesses (reads from or writes to) the defective sector which is the reassignment source sector, the HDD accesses the reassignment destination sector corresponding to the defective sector. This leads to inhibition of using a defective sector within the HDD with no necessity of address management by the host.
With expansion of use range of PC applications and HDDs in recent years, opportunities for storing AV data in HDDs have been increasing. Hence, a stream command is defined corresponding to continuous data such as AV data. In reproducing AV data, data continuity (data transfer rate) has a higher priority than a marginal data error (accuracy). Therefore, an option is prepared that does not start an error recovery, but accesses the next sector even if an error has occurred during data retrieval or data write. This is referred to as continuous read or continuous write.
When the host requests accesses to continuous sectors as in accesses in AV data, an access to the reassignment destination address needs more time than an access to the reassignment source address because the reassignment destination sector is located at a position away from the reassignment source sector. Therefore, a technique has been known that accesses the reassignment source sector instead of the reassignment destination sector in the accesses in response to an AV command even if there is a reassigned sector (for example, refer to Japanese Patent Publication No. 2004-70979 “Patent Document 1”). This allows the AV data to be transferred at a required constant rate.
As in the above conventional technique, retrieving data at the reassignment source sector can maintain the transfer rate and continuity of the AV data. However, if a data sector has been reassigned, new data is written in the reassignment destination sector and data in the reassignment source sector will never be updated. This results in that the HDD always sends the host incorrect data in response to the AV command. Even in a mode where the transfer rate has higher priority than the accuracy in data, it is preferable to transfer more accurate data to the host. If a plurality of data sectors designated by the AV command include many reassigned data sectors, accuracy in data to be transferred to the host will be significantly impaired.
Embodiments of the present invention help to improve accuracy in data and maintain data rate in a disk drive device which sends data at a reassignment source address to a host in response to a specific command. According to the embodiment of
Embodiments of the present invention relate to a disk drive device and a method for accessing a disk thereof, and more particularly to control in a disk drive device for sending reassignment source address data to a host in response to a specific read command.
An aspect of embodiments of the present invention is a method for accessing a disk in a disk drive device. This method performs a data write operation to a reassignment source address and a reassignment destination address corresponding to the reassignment source address in response to a write command to the reassignment source address. It sends data at the reassignment destination address in response to a normal read command, and also sends data at the reassignment source address in response to a predetermined read command. Performing a data write operation to a reassignment source address improves accuracy in data. This aspect includes a method for performing a data write operation only in the case that predetermined conditions are satisfied. The data write operation includes the case that the operation does not write data on the disk because of errors or the like.
In one example, if an error occurs in the data write operation to the reassignment source address, the method omits an error recovery and ends the data write operation to the reassignment source address. The omission of the error recovery results in preventing performance degradation.
In one example, the maximum number of steps for an error recovery corresponding to the data write operation to the reassignment source address is smaller than the maximum number of steps for an error recovery corresponding to the data write operation to the reassignment destination address. This achieves accuracy in data and good performance.
In one example, the method changes the data write operation to the reassignment source address depending on a reason for reassignment of the reassignment source address. Furthermore, if the reason for reassignment is a write error, it is preferable that the method omits an error recovery in the data write operation to the reassignment source address or sets the maximum number of steps for the error recovery to be smaller than the maximum number of steps in a case that the reason for reassignment is a read error. This achieves accuracy in data and good performance.
In one example, the method determines whether or not the data write operation to the reassignment source address should be executed depending on the kind of the write command. This achieves accuracy in data and good performance.
In one example, the write command to the reassignment source address includes a plurality of addresses, and the method determines whether or not the data write operation to the reassignment source address should be executed or changes the maximum number of steps for the error recovery depending on the number of the reassignment source addresses included in the plurality of addresses. This achieves accuracy in data and good performance.
A disk drive device according to another aspect of embodiments of the present invention comprises a head for accessing a disk, a moving mechanism for moving the head above the disk, and a controller for sending data at a reassignment destination address retrieved with the head in response to a normal read command and sending data at a reassignment source address retrieved with the head in response to a predetermined read command. The controller performs data write operations to the reassignment destination address corresponding to the reassignment source address and the reassignment source address in response to a write command to the reassignment source address. Performing a data write operation to the reassignment source address can improve accuracy in data. This aspect includes a device for performing a data write operation only in the case that predetermined conditions are satisfied. The data write operation includes the case that the operation does not write data on the disk because of errors or the like.
A disk drive device according to yet another aspect of embodiments of the present invention comprises a head for accessing a disk, a moving mechanism for moving the head, and a controller for controlling the moving mechanism. The controller retrieves data at a reassignment destination address using the head in response to a normal read command, and further determines whether to retrieve data at a reassignment source address or to retrieve data at a reassignment destination address based on the number of reassignment source addresses included in the addresses designated by a predetermined read command. Determining whether to retrieve data at a reassignment source address or to retrieve data at a reassignment destination address based on the number of reassignment source addresses included in the addresses designated by a predetermined read command achieves accuracy in data and good performance.
In one example, the number of reassignment source addresses is the maximum number of continuous reassignment source addresses in the designated addresses, and the controller retrieves data at the reassignment destination address using the head if the maximum number of the continuous reassignment source addresses exceeds a reference number. Or, the number of reassignment source addresses is a total number of the reassignment source addresses included in the designated addresses, and the controller retrieves data at the reassignment destination address using the head if the total number exceeds a reference number. This achieves more appropriate determination.
Embodiments of the present invention can maintain a necessary data rate and improve the accuracy in data in a disk drive device which sends a host data at a reassignment source address in response to a specific command.
Hereinafter, particular embodiments to which the present invention has been applied will be described. For clarity of explanation, the following description and the accompanying drawings contain omissions and simplifications as appropriate. Throughout the drawings, like components are denoted by like reference numerals, and their repetitive description is omitted if not necessary for the sake of clarity. Hereinbelow, embodiments of the present invention will be described by way of example of a hard disk drive (HDD) as an example of a disk drive device. If an address designated by a read command contains a reassignment source data sector, the HDD according to one embodiment accesses a reassignment destination data sector in response to a normal read command and accesses a reassignment source data sector instead of the reassignment destination sector in response to a specific read command.
First, an entire configuration of an HDD is outlined.
Each head slider 12 includes a slider flying over the magnetic disk and a head element portion which is fixed to the slider and converts magnetic signals and electric signals (retrieves and writes data). Each head slider 12 is fixed to a tip end of an actuator 16. The actuator 16, which is coupled to the voice coil motor (VCM) 15, pivots about a pivotal shaft to move the head sliders 12 above the magnetic disk 11 in its radial direction. The motor driver unit 22 drives the VCM 15 in accordance with control data from the HDC/MPU 23. An arm electronics (AE) 13 selects a head slider 12 to access (read from or write to) the magnetic disk 11 from a plurality of head sliders 12 in accordance with control data from the HDC/MPU 23 and amplifies read/write signals.
The RW channel 21, in a read operation, extracts servo data and user data from the read signals obtained from the AE 13 to perform a decoding process. The decoded data are supplied to the HDC/MPU 23. In a write operation, the RW channel 21 code-modulates write data supplied from the HDC/MPU 23 and further converts the code-modulated data into write signals to supply them to the AE 13. In the HDC/MPU 23, the HDC is a logic circuit and the MPU operates in accordance with firmware loaded in the RAM 24. With starting up of the HDD 1, data required for control and data processing is loaded from the magnetic disk 11 or a ROM (not shown) onto the RAM 24. The HDC/MPU 23 is an example of a controller and performs entire control of the HDD 1 in addition to necessary processes concerning data processing such as head positioning control, interface control, and defect management.
The HDC/MPU 23 transmits read data from the magnetic disk 11 obtained from the RW channel 21 to the host 51. The read data from the magnetic disk 11 is temporarily stored in a read buffer in the RAM 24 before transferred to the host 51 via the HDC/MPU 23. The write data from the host 51 is temporarily stored in a write buffer in the RAM 24 via the HDC/MPU 23 before transferred to the magnetic disk 11 via the HDC/MPU 23 at a specific timing. The HDC/MPU 23 according to one embodiment particularly controls data rewrite operation in accordance with the number of data write operations to the magnetic disk 11.
Hereinafter, detailed explanation will be given on control for the access to the magnetic disk in the HDD 1 according to one embodiment. In one embodiment, the “access” is a generic concept of read from and write to the magnetic disk 11. The HDD 1 of the present embodiment can handle a stream command for AV data and the like. If the host 51 instructs a continuous access for a stream read command, the HDD 1 transfers data to the host 51 ignoring an ECC error. Further, the HDD 1 of the present embodiment reads a data sector at the reassignment source address instead of the data sector at the reassignment destination address in response to the stream read command for which a continuous access is ordered. In the present specification, this function of the continuous read is referred to as “read continuous”.
The host 51, for example, enables the read continuous with a field within the stream read command. The HDD 1 reads the reassignment source data sector instead of the reassignment destination data sector in the address designated by the stream read command.
Specifically, as shown in a flowchart of
If the received read command is not a stream read command or the read continuous has been disabled (N in S12), the HDC/MPU 23 retrieves data from the magnetic disk 11 as a normal operation. That is, if an ECC error occurs, it performs an off-line error correction or an error recovery. Or, if a reassigned data sector is present, it reads the reassignment destination data sector (S14). Thus, continuity and transfer rate in data transfer can be maintained by reading the reassignment source data sector in response to a predetermined read command.
If a reassigned data sector is present, the HDD 1 according to one embodiment performs a data write operation (including the case that it does not write to the magnetic disk 11 due to an error) to the reassignment source data sector in addition to the reassignment destination data sector. Writing data to the reassignment source data sector improves the accuracy in data to be transferred in the continuous read. The HDD 1 can perform write operations to the reassignment destination data sector and the reassignment source data sector in the same way. In the present specification, the “write operation” includes a correction operation and an error recovery if an error occurs as well as an actual write operation of magnetic data to the magnetic disk 11.
However, the reassignment source data sectors are usually defective sectors having any problem so that there is high probability that a data write cannot be completed to the reassignment source data sector or it takes long time to write data due to an error occurrence, which may impair the performance of the HDD 1. Therefore, the data write operations are preferably different between the reassignment destination sector and the reassignment source data sector. In a preferred example, the HDD 1 performs an error recovery in a different way if an error has occurred.
The HDC/MPU 23 controls the error recovery. Specifically, as shown in a block diagram of
The ERP tables 241 and 242 contain steps for error recovery recorded therein, particularly such as a retry step of write operation under the same conditions, a retry step with a changed parameter of the RW channel 21, and the like. The HDC/MPU 23 performs the ERP steps sequentially from the ERP step with higher priority until the error recovery or to the final ERP step. In a preferred embodiment, the maximum number of the execution steps in the ERP table for reassignment source 242 is smaller than the maximum number of execution steps in the ERP table for reassignment destination 241.
The maximum number of ERP steps is the number of steps from the beginning to the last ERP steps to be actually carried out. For example, 128 is assigned to the maximum number of execution steps in the ERP table for reassignment destination 241 and 64 is assigned to the number of execution steps in the ERP table for reassignment source table 242. Assigning the smaller number of the execution steps in the ERP table for reassignment source table 242 avoids wasting much time for a write operation to the reassignment source data sector so that the performance degradation is prevented.
Now referring to a flowchart of
The HDC/MPU 23 attempts to write data to the reassignment source address. If the HDC/MPU 23 starts a write operation to the reassignment source address (S26) and an error arises (Y in S27), the HDC/MPU 23 selects the ERP table for reassignment source 242 (S28). If it performs an error recovery in accordance with the ERP table for reassignment source 242 (S29) and completes the error recovery to write data normally, the data write operation to the reassignment source address ends. If the error is not recovered by the last step, the HDC/MPU 23 ends the operation without data write to the reassignment source address. If no errors arise (N in S27), the HDC/MPU 23 writes data normally and ends the data write operation to the reassignment source address.
Since it is permitted even if the data of the reassignment source data sector are not updated, there arises no problem even if the data write operation ends in case that the error has not been recovered. From the view point of the performance, it is preferable that the operation ends at the earliest timing as possible. Smaller number of ERP execution steps and end of the operation without error recovery in case that the error is not recovered improve the accuracy in data and maintain the performance.
In the above example, if an error arises in the write operation to the reassignment source data sector, an error recovery is performed referring to a table with smaller number of the maximum execution steps than the one in normal. If the priority is placed in the performance, omission of the error recovery can shorten the operation time. Specifically, as shown in the flowchart of
The examples explained referring to
The reasons for reassignment are particularly a read error, a write error, a command from the host 51, and the like. The HDC/MPU 23 registers a corresponding address in the RDM 243 if an error cannot be recovered in a read operation (read error) or in a write operation (write error), or in accordance with a command from the host 51. The HDC/MPU 23 refers to the reason for reassignment and determines whether or not it performs a data write operation to the reassignment source address, if the write command designates the reassignment source address.
Detailed operation steps will be explained referring to a flowchart of
If the HDC/MPU 23 receives a write command (S41), it performs a write operation to a reassignment destination address (S42). The HDC/MPU 23 determines the reason for reassignment referring to the RDM 243 (S43). If the reason for reassignment is a read error (Y in S44), the HDC/MPU 23 performs a write operation to the reassignment source address (including the case that an actual write is not made due to an error) (S45). If the reason for reassignment is not a read error (N in S44), the HDC/MPU 23 ends the operations without starting a write operation to the reassignment source address.
The above example determines whether or not the data write operation is necessary to the reassignment source address depending on the reason for reassignment. On the contrary, the way of error recovery may be changed depending on the reason for reassignment. In one example, the HDC/MPU 23 refers to an ERP table containing a large number of registered execution steps in a read error and an ERP table containing a small number of registered execution steps in other errors.
Specifically, as shown in a flowchart of
In another example, the HDC/MPU 23 determines the necessity of data write operation to the reassignment source address depending on the number of reassigned data sectors in the address designated by the write command. If the number of the reassignment source addresses in the designated address exceeds a reference number, the HDC/MPU 23 performs a data write operation to the reassignment source address. If it does not exceed, it does not perform the operation. The number of data sectors to be counted may be the number of all reassignment source data sectors in the address designated by the command or the maximum number of the continuous data sectors in the address designated by the command.
Now referring to a flowchart of
The number of addresses to be compared with the reference number may be the maximum number of continuous sectors or the number of total sectors as described above. In another case, whether or not a write operation is performed to the reassignment source data sector may be determined depending on the kind of the write command. For example, if the write command is a stream command, it may be determined to perform a write operation to the reassignment source data sector.
Based on the number of reassignment source data sectors in the designated address in the write command, the HDC/MPU 23 may change the ERP table in a data write operation to the reassignment source address. It corresponds to changing the criteria in the operation explained referring to
Specifically, as shown in
In a read continuous, it may be desirable to select either the reassignment source data sector or the reassignment destination sector depending on predetermined conditions and transfer the selected data to the host 51, instead of reading the reassignment source data sector or reading the reassignment destination sector at all times. This achieves both of accuracy in transferred data and sustainment of transfer rate.
In an example, the HDC/MPU 23 selects either the reassignment source address or the reassignment destination address depending on the number of the reassignment source addresses (number of the data sectors) in the designated address by the read command corresponding to the read continuous. Specifically, if the number of reassignment source addresses in the designated address exceeds the reference number, the HDC/MPU 23 reads the reassignment destination data sector; and if it does not exceed the reference number, the HDC/MPU 23 reads the reassignment source data sector. If the number of reassigned data sectors is large, accessing the reassignment destination achieves both accuracy in data and sufficient transfer rate. The number of the reassignment source addresses to be compared with the reference number may be the total address number or the maximum number of the continuous addresses as in the above-described example.
Now referring to
If the number of the reassignment source addresses exceeds the reference number (Y in S84), the HDC/MPU 23 reads the reassignment destination data sector (S85). In this step, it may perform a correction operation of the ECC error or ignore the error. If the number of reassignment source addresses does not exceed the reference number (N in S84), the HDC/MPU 23 reads the reassignment source data sector instead of the reassignment destination and transfers the data to the host 51 (S86). In this step, the HDC/MPU 23 ignores the ECC error.
If the received read command is not a stream read command or the read continuous has been disabled (N in S82), the HDC/MPU 23 retrieves data from the magnetic disk 11 as a normal operation. Namely, if an ECC error occurs, it performs an error correction or an error recovery off-line. If any reassigned data sector is present, it reads the reassignment destination data sector (S85).
As set forth above, embodiments of the present invention have been described by way of example of particular embodiments but are not limited to the above embodiments. A person skilled in the art can easily modify, add, and convert each element in the above embodiments within the scope of the present invention. For example, the control according to one embodiment can be applied to a disk drive device using disks other than magnetic disks. The components in each of the above examples may be combined or may be used separately. For example, the selection of the reassignment source address or reassignment destination address may be performed independent from the ECC error handling.
| Number | Date | Country | Kind |
|---|---|---|---|
| 2007-115986 | Apr 2007 | JP | national |
