KR100305309B1 - Data reading method and data storage device - Google Patents

Abstract

An object of the present invention is to reduce the occurrence of a read error of a storage device due to bus noise during data transfer between the host device and the storage device.

To achieve this object, when a read error due to bus noise is detected, data transmission to the host device 10 is once stopped, and the read is performed in a state in which the bus noise which is the cause of the read error is removed. When the buffer 6 becomes full, the data transfer is started again when the reading of data on the command from the host device 10 ends.

Description

Data reading method and data storage device

The present invention relates to a method and a data reading device for reading data from a storage device. More specifically, the present invention relates to a data reading method and a data reading device characterized by a method of transferring data between a host device and a storage device when a read error occurs due to bus noise.

Along with the rapid spread of information equipment, notebook computers are becoming smaller and lighter. As a result of this influence, in a notebook computer, various peripheral devices are densely arranged in the narrow case. In such a case, various kinds of noise are generated, which is a poor environment for peripheral devices. In particular, the hard disk device is susceptible to noise since the magnetic head following the recording track on the magnetic disk detects subtle magnetic flux changes and reads the data recorded on the magnetic disk.

In a hard disk device, the nearest noise source is a data bus. For example, if data is being read in parallel with data transfer to a host device (in this case, an MPU of a laptop computer), a reading error occurs every time a sector is read. For example, normally, 128 sectors may be read in one rotation of the disc, but an error may occur in each sector, which may take 128 revolutions.

For this reason, in notebook computer personal computers and the like, various countermeasures by hardware such as strengthening of shields are generally adopted.

However, depending on the specifications of the host device, the above-described measures by the hardware may be insufficient, and in such a case, the read error of the hard disk device (HDD) increases.

Therefore, an object of the present invention is to reduce a read error of a storage device due to bus noise during data transfer between the host device and the storage device.

Moreover, an object of this invention is to implement | achieve reading error reduction only by the software of a memory | storage device.

A reading method according to the present invention is a method for reading data from a storage device, comprising: a reading step of reading data from a storage device, a data transfer step of outputting data read in the reading step to a host device via a bus; An error detection step of detecting occurrence of a read error due to a bus noise, and stopping the data transfer to and from reading from the storage device at the same time when the occurrence of the read error is detected by the error detection step. It includes a control step.

The data reading apparatus according to the present invention further includes a storage medium, reproducing means for reproducing data stored in the storage medium, data transmission means for outputting data reproduced by the reproducing means to a host device via a bus, and bus noise. Error detection means for detecting the occurrence of a read error by means of controlling the operation of the data transfer means and simultaneously performing data transfer to the host device and read from the storage device when the occurrence of the read error is detected by the error detection step. It is provided with the control means for stopping a thing.

BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a block diagram showing the structure of a disk drive apparatus according to an embodiment of the present invention.

2 is a flowchart showing error recovery processing by the disk drive apparatus according to the embodiment of the present invention.

3 is a flowchart showing error recovery processing by the disk drive apparatus according to the embodiment of the present invention.

<Explanation of symbols for the main parts of the drawings>

1: magnetic disk

2: buffer

3: head drive mechanism

4: servo system

5 read / write meter

6: buffer

7: bus

8: control unit

10: information processing device (host device)

1 is a block diagram showing the structure of a disk drive apparatus according to an embodiment of the present invention. This disk drive device applies the present invention to a small disk drive device used for a notebook computer, for example.

The disk drive apparatus includes a magnetic disk 1 for storing data, a head 2 for reading and writing data to and from the magnetic disk 1, and a head drive mechanism 3 for driving the head 2; And a servo system 4 for controlling the head drive mechanism 3, a read / write system 5 for controlling recording / reading by the head 2, and a buffer 6 for holding write / read data. And a control unit 8 for controlling the operation of the entire disk drive device and controlling the input / output of data to and from the external device (host device) 10 connected via the bus 7.

The buffer 6 has a storage capacity that can hold, for example, data for hundreds or more of recording units (sectors) on the magnetic disk 1.

The control unit 8 includes a processor for executing a control program, a control program, a memory holding data such as a defect map indicating a unit of a bad sector, and the like. As a memory, RAM and ROM are provided, and a control program is stored in ROM, for example. Although the operation of the ROM is high speed, the capacity of the ROM is limited in terms of practicality or cost, so that a control program that cannot enter the ROM is stored in a predetermined area on the magnetic disk 1, and the power is supplied (POR (power). read it into RAM and execute it.

The controller 8 controls the operation of the entire disk drive apparatus by executing a control program (micro program), and reads the magnetic disk 1 on the basis of commands and data supplied from the host apparatus via the bus 7. Control recording and the like.

The processor of the controller 8 is capable of executing a plurality of processes in parallel, one of which is a command between the host device 10 and the control of input / output of data, and the other is a buffer ( 6) is a write cache process for recording the record data retained in 6) onto the magnetic disk 1, and the other is a pre-reading from the magnetic disk 1 of the data instructed to be read from the host device or the data to be read. It is a read cache process that looks ahead and holds in the buffer 6. In addition, processes for servo control and error recovery processing are executed in parallel.

In the input / output control process for the host device 10, the control unit 8 includes a command for requesting data recording from the host device 10 and an address indicating a recording destination (sector on the magnetic disk 1) (for example, When the logical block address (LBA, etc.) and write data are supplied, the host device 10 is released from the recording process after the supplied write data is held in the buffer 2. When the command for requesting reading of the data and the address LBA in which the data to be read are stored are supplied from the host device 10, the data requested for reading is read from the magnetic disk 1 and the buffer 2 ), It is read and supplied to the host apparatus 10. If it is not held in the buffer 2, the read / write system 5 is instructed to read data that has been requested to read, and the read data is read from the host apparatus 10. ).

In the write cache process, the control unit 8 is supplied from the host device 10 as described above, sequentially reads the record data held in the buffer 2, and supplies the read / write system 5 to the magnetic disk 1. ) To the record.

In the read cache process, the control unit 8 instructs the read / write system 5 to read data that is expected to have a read request from the host device 10 and holds the read data in the buffer 2.

By the way, the processor of the controller 8 detects whether the read of the data by the read / write system 5 is a read error due to noise on the bus 7 in addition to the above-described process, and based on the detection result The error recovery process is performed.

A characteristic of the read error due to noise on the bus 7 is that errors frequently occur during data transfer, and data is recovered in one reread. For this reason, in the error recovery process, the frequency of occurrence of a read error is first checked, and the number of rereads is checked again to distinguish an error due to noise on the bus 7 from another error. In addition, only one of these may be distinguished from other errors.

If a read error due to noise on the bus 7 occurs, data transmission to the host device 10 and reading of data from the magnetic disk 1 are stopped at the same time, and only data reading is performed. This reduces bus noise and reduces the occurrence of read errors due to bus noise.

When data is read in this state, the data read from the magnetic disk 1 are held in the buffer 6 sequentially. When one of the following conditions is satisfied, data transfer from the magnetic disk 1 and held in the buffer 6 is started to the host.

(a) Drive Catchup (buffer (6) full of read data)

In this state, the next data cannot be read unless the data held in the buffer 6 is transferred to the host device 10. For this reason, reading continues after data transfer.

(b) Buffer Ready (the state which has been read in advance by Look Ahead and the data required for the command from the host device 10 has already been read in the buffer 6)

For example, this corresponds to a case where a read process for a plurality of commands from the host device 10 is executed, and the reading of data for any of the commands is completed. In this state, it is necessary to continue reading data for other commands.

(c) Command completion (the state in which reading according to the read command is completed while the look-ahead is not performed)

In this state, since all data for the command is already held in the buffer 6, no further reading is necessary and the reading ends.

(d) Read hard error (unreadable)

In this state, it is impossible to read data to the command. For this reason, the host device 10 is notified of the effect and the reading is finished.

2 and 3 are flowcharts showing the error recovery processing as described above. First, when a read error is detected in step 110, it is determined in the following steps 112 and 114 whether the read error is due to bus noise.

In step 112, a frequency of occurrence of a read error is detected. In general, access to the magnetic disk 1 is often made for a close LBA in a short time. Therefore, when reading errors occur continuously in the adjacent LBA, a plurality of reading errors are generated in a short time, and it is considered that the occurrence frequency of reading errors is high. Therefore, in this step 112, the difference between the logical block address (LBA) in which the current read error occurred and the LBA in which the previous read error was issued is obtained, and the difference is read by comparing a predetermined threshold value (for example, 16). It is indirectly determined whether an error occurrence frequency is equal to or greater than a predetermined threshold value.

At present, when the difference between the logical block address LBA where a read error occurs and the LBA where a previous read error occurs is smaller than the threshold value 16, the occurrence frequency of the read error is high and there is a possibility that the read error is caused by bus noise. For this reason, the process proceeds from step 112 to step 114, and continues the determination as to whether or not it is a read error due to bus noise. If the difference is greater than or equal to the threshold value 16, the occurrence frequency of the read error is not so high, so the possibility of the read error due to the bus noise is low. For this reason, the process proceeds from step 112 to step 128 in FIG. 3 and continues reading data. On the other hand, the setting of the threshold is not limited to sixteen. Alternatively, detection of the occurrence frequency of the read error may be performed directly by counting the number of occurrences of the read error at predetermined time intervals. In this case, it is necessary to count the number of read errors for each predetermined time, which increases the processing load than the processing of this step 112.

In step 114, it is determined whether the number of rereads of the LBA when the last read error has occurred is equal to or greater than a predetermined threshold (for example, 1).

If the number of rereads is equal to the threshold value 1, the process after step 116 is executed because there is a high possibility of a read error due to bus noise. If the number of rereads is larger than the threshold value 1, it is unlikely that it is a read error due to a cause other than bus noise, and the flow proceeds to step 128 in Fig. 3 to continue reading data. On the other hand, the setting of the threshold value is not limited to 1, and may be, for example, two or more numbers or may be changed depending on the elapsed LBA from the LBA in which the last read error occurred.

If it is determined in step 112 and 114 that the read error is due to bus noise, the process proceeds to step 116 where all valid data on the buffer 6 is transferred to the host device 10 to transfer the buffer 6. Empty it once. After that, in the next step 118, the concurrent I / O mode, which executes the data transfer to the host device 10 and the data read from the magnetic disk 1, is performed in a disabled state. . This eliminates bus noise due to data transfer and resumes reading in step 120. By reading data in such a state, it is possible to reduce a read error due to bus noise.

After resuming reading of data from the magnetic disk 1, the conditions (c), (d), (a), and (b) described above in steps 122, 123, 130, and 131 The determination is made.

If the condition (a) (buffer is full) or (b) (buffer ready) is satisfied, the process proceeds from step 122 or step 123 to step 124 in FIG. After enabling the simultaneous input / output mode, the data held in the buffer 6 is transferred to the host device 10 in the next step 126. Thereafter, the reading of the remaining data is started in step 128 and the process returns to step 110 in FIG. 2 to monitor the occurrence of the read error.

If condition (c) (command completion) or (d) (read hard error) is satisfied, the process proceeds from step 130 or step 131 to step 132 in FIG. After the simultaneous input / output mode is enabled, the next step 134 transfers the data retained in the buffer 6 to the host device 10 and then terminates.

As described above, in this embodiment, when a read error due to bus noise is detected, the bus noise is caused by reading from the magnetic disk 1 in a state in which data transmission to the host device 10 is stopped. One reading error can be reduced.

In addition, in this embodiment, a read error caused by bus noise is not required by hardware (such as a micro program) that is executed by a processor of the control unit 8 without requiring measures by hardware such as strengthening of a shield as in the prior art. Can be reduced. For this reason, it can contribute to the cost reduction of the apparatus using the said memory apparatus.

In addition, this invention is not limited to embodiment mentioned above, Various changes are possible within the scope of the technical idea of this invention. For example, in the above-described embodiment, the case where the present invention is applied to the disk drive device has been described. However, if the storage device is connected to the host device via a bus, the present invention can also be applied to, for example, a memory card and other storage devices. .

According to the present invention, it is possible to stop performing data transfer to the host device and reading data from the storage device at the same time in the control step based on the error detected in the error detection step. The readout software error rate due to bus noise can be reduced.

Further, according to the present invention, since the reduction of the read error can be realized only by the software of the storage device, no countermeasures by hardware such as strengthening of the shield are required. For this reason, it can contribute to the reduction of the cost of the apparatus using the said memory apparatus.

Claims (10)

A method for reading data from a storage device having a bus, the method comprising: A reading step of reading data from the storage device; Detecting occurrence of a read error due to bus noise by determining a reread count for a previous read error and determining that the read error is not due to bus noise when the read count is equal to or greater than a predetermined threshold. step; And In response to the detecting step, stopping data transmission to the host device to reduce bus noise and then rereading the designated data from the storage device. Data reading method comprising a. The method of claim 1, wherein the detecting step determines whether a read error occurrence frequency for a previous read error is equal to or greater than a predetermined threshold and wherein the read error occurrence frequency is equal to or greater than the predetermined threshold. And determining that the read error is not due to bus noise when abnormal, and temporarily storing data read from the storage device in a buffer, and when the buffer is full, transfer data to the host device. The method further comprises the step of resuming. The method of claim 1, further comprising resuming data transmission to a host device when the reading of the designated data is completed without error. CLAIMS 1. A method for reading data from a storage device having a bus, comprising: reading specified data from the storage device; Detecting the occurrence of a read error due to bus noise by determining that the read error is not due to bus noise when the error occurrence frequency for the previous read error is equal to or greater than a predetermined threshold; And responsive to the detecting step, stopping data transfer to a host device to reduce bus noise, and then rereading the designated data from the storage device. A data storage device having a bus and a control unit for generating bus activity during data transfer to a host, comprising: a storage medium; A head for reading data stored in the storage medium; When rereading data receiving a bus noise error by distinguishing occurrences of read errors due to bus noise, the data transmission to the host is stopped, and the difference between the address of the previous data unit and the address of the current unit where the bus noise error occurred is detected. And a control section for judging the magnitude and discriminating occurrence of a read error not caused by bus noise when the magnitude of the difference is equal to or greater than a predetermined threshold value. 6. The data storage device according to claim 5, wherein said control section stores the number of rereads for a previous read error and distinguishes occurrence of read errors not caused by bus noise when the number of rereads is greater than or equal to a predetermined threshold. 6. The apparatus of claim 5, wherein the controller determines the magnitude of the difference between the address of the previous data unit and the address of the current unit where the bus noise error has occurred, and if the magnitude of the difference is greater than or equal to a predetermined threshold, reading is not caused by bus noise. And a buffer for discriminating occurrence of an error and temporarily storing data read from the storage device, wherein the control unit resumes data transmission to a host when the buffer is full of data. 8. The data storage device according to claim 7, wherein said control section resumes data transfer to a host device when reading of said data is completed without error. A data storage device comprising: a storage medium containing data; A bus on which activity occurs when data is sent to a host; Means for detecting a read error when a data unit is read from the storage medium; Means for rereading the data unit when a read error is detected; Means for recording a first address and a reread count of the first designated data unit in which a read error occurred; And data to the host before rereading the second designated data unit when the difference between the first address and the second address of the second designated data unit is equal to or less than a predetermined threshold and the number of rereads is equal to or less than a predetermined value. Means for processing said second specified data unit at which a read error occurs by stopping the transfer. 10. The apparatus according to claim 9, further comprising a buffer for temporarily storing data read from the storage device, wherein the second designated data unit processing means includes an error when the buffer is full of data or the second designated data unit is in error. Data storage that resumes data transfer to the host when complete without.

Images