JP2000067515A - Lsi for disk control, memory device and information processing system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To increase the data transfer speed between disk devices for which a parallel recording system is adopted. SOLUTION: A disk controller 8 of this information processing system receives recording data in parallel from a host computer 1, forms an ECC code in an LSI 13 for disk control and sends the recording data and the ECC code in parallel to a disk drive device 9. The disk drive device 9 writes respective pieces of the data received in parallel into plural sectors. The LSI 13 for disk control forms the data for synchronization, etc., respectively in each sector.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a storage device using a storage disk such as a magnetic disk, an optical disk and a magneto-optical disk, and more particularly to an LSI for controlling the disk.

[0002]

2. Description of the Related Art FIG. 7 shows a configuration of a conventional storage device using a storage disk (hereinafter referred to as a disk device).

[0003] As shown in the figure, a disk device 10 comprises a disk control device 8 and a disk drive device.

[0004] The disk drive device includes a code decoding circuit 6 for encoding data into a storage code on the disk and decoding the storage code, and a recording and reproduction circuit 7 for recording and reproduction on the storage disk. The device 8 includes a host interface circuit 2 serving as an interface with the host computer, a disk control LSI 3 controlling transfer of data accessed by the host computer, a buffer memory 4 holding data to be transmitted / received to / from the host computer,
A microprocessor 5 for controlling the entire system is provided.

In the above-described disk device, data transmission / reception between the disk control device and the disk drive device is performed by two serial data of read and write.

Therefore, the disk control LSI used in the disk control device has one serial data interface for each of read and write, and handles each data as serial data.

[0007]

However, in recent years, disk devices employing a parallel recording system have been developed to increase the data transfer speed.

In this disk drive, if the control is to be realized with only one conventional disk control LSI, the data between the disk control LSI and the disk drive must be serial data, However, the data transfer speed could not be increased.

Further, if an attempt is made to use a plurality of disk control LSIs, the processing of the microprocessor for controlling the disk control LSIs becomes excessively large, and the hardware becomes large-scale, so that the disk drive becomes small and large. There has been a problem that it is difficult to increase the capacity and the cost increases.

SUMMARY OF THE INVENTION It is an object of the present invention to provide a disk control LSI for a disk device adopting a parallel recording method, which can increase the data transfer speed of the disk device with one LSI.

It is another object of the present invention to provide a disk device capable of increasing the data transfer speed, and an information processing system using the disk device.

[0012]

In order to achieve the above object, the present invention provides a host interface for inputting / outputting data accessed by a host device, and a disk for inputting / outputting storage data to / from a storage disk in parallel. There is provided a first disk control LSI, comprising: an interface; and means for controlling data transfer between the two interfaces.

In order to achieve the above object, a host interface for inputting / outputting data accessed by a host device in parallel, a disk interface for inputting / outputting storage data to / from a storage disk in parallel, Means for controlling the parallel transfer of data, and a code for detecting an error in the transfer data is obtained in units of parallel data, and is parallelized to the same number of bits as the transfer parallel data, and the parallel data transferred from the host interface to the disk interface. And a means for detecting an error in the data of the transfer data from a code for detecting an error in the transfer data added to the parallel data transferred from the disk interface to the host interface. Second disc system To provide a use LSI.

In particular, the second disk control L
In a SI, a disk control LSI is provided, wherein a Reed-Solomon code is used as a code for detecting an error in the transfer data. Further, it is preferable that each of the disk control LSIs includes means for outputting, from the disk interface, the sector management information of the storage disk to be written to the storage disk in parallel with the number of bits of the transfer parallel data.

In addition, the present invention provides a first disk device comprising a storage disk and the disk control LSI for controlling the parallel transfer of data to and from the storage disk. .

[0016] To achieve the above object, the present invention provides:
A buffer memory for holding parallel data accessed by the host device, a disk control circuit for controlling parallel transfer of data between the buffer memory and the storage disk, encoding of storage data of the storage disk into recording codes on the storage disk, and A second disk device includes an encoding / decoding circuit for decoding a recording code and a storage disk for storing storage data in parallel.

Further, the present invention also provides an information processing system comprising the disk device and a host device accessing the disk device.

As described above, in the present invention, the data lines between the disk control LSI and the disk drive device are parallelized. Also, if there is a circuit that handles data serially even at one place in the LSI, all of these internal circuits have a parallel configuration with respect to data lines in order to impair high-speed operation.

In addition, if the characteristics of the hardware configuration of the disk device are described, regardless of the LSI, the code (recording code) to be written to the disk drive device, or the decoding process when reading the data (host). This is equivalent to realizing all of the data lines for the portion between the interface between the computer and the disk device with a circuit having a parallel configuration.

[0020]

According to the first disk control LSI of the present invention, the input / output of the storage data of the storage disk is performed as parallel data by the disk interface. Data transfer between the host interface and the disk interface is performed.

According to the second disk control LSI of the present invention, data accessed by the host device is input / output in parallel at the host interface, and storage parallel data of the storage disk is input / output in parallel at the disk interface. The data between the two interfaces is transferred in parallel. On the other hand, a code for detecting an error in the transfer data is obtained in units of parallel data, parallelized to the same number of bits as the transfer parallel data, and added to the transfer parallel data from the host interface to the disk interface.

Further, an error in the transfer data is detected from a code for detecting an error in the transfer data added to the parallel data transferred from the disk interface to the host interface.

In the second disk control LSI, it is preferable to use a Reed-Solomon code as a code for detecting an error in the transfer data.

It is preferable that each of the disk control LSIs outputs the sector management information of the storage disk to be written to the storage disk in parallel with the number of bits of the transfer parallel data from the disk interface.

Further, according to the first disk device of the present invention, parallel transfer of data to and from a storage disk is controlled by the disk control LSI.

According to the second disk device of the present invention, the buffer memory holds the parallel data accessed by the host device. Further, the data transfer between the buffer memory and the storage disk is performed as parallel data by the disk control circuit. The encoding / decoding circuit encodes data stored in the storage disk into a recording code on the storage disk and decodes the data from the recording code.

According to the information processing system of the present invention, the host device accesses the disk device and performs the processing.

[0028]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below using an information processing system having a magnetic disk storage device as an example.

First, the storage format of the disk device according to the present embodiment will be described.

First, FIG. 8 shows a storage format of a conventional disk device for recording data serially.

As shown in the figure, the disk device records data in units of sectors 17 on a magnetic disk, and each sector 17 includes data (DATA FIELD) 29 sent from the host computer 1 and data 29. In addition to the error correction code (DATA CHECK BYTES) 30, information for managing the sector 17 must be written. Such information includes an ISG (intersector gap) 20 inserted between each sector, an address 18 of each sector, a synchronization pattern 28, and the like. Writing this information is called formatting.

When data is serially recorded as described above, a sector having a certain sector number exists as a continuous area on a disk. In this embodiment, data is recorded in parallel. Therefore, a plurality of sectors having the same sector number exist on the disk at a distance. Since these sectors must be accessed simultaneously as one sector for management, this must be taken into account when formatting.

FIG. 5 shows a storage format of a disk device for recording data in parallel according to this embodiment.

In this embodiment, as shown in FIG.
The TA FEELD 29 and the DATA CHECK BYTES 30 are written in a plurality of sectors 17a, 17b, and 17c having the same sector number.
Is a signal used for synchronizing data reading at the time of recording / reproducing, and if it is written separately, data cannot be read. Further, ISGs (intersector gaps) 20a to 20c are required for each sector according to the sector length. Therefore, these pieces of information are independently added to each sector on the disk.

With the above configuration, data can be correctly reproduced even in parallel storage of data.

Although only the data field 29 and the ECC byte 30 are written separately to the sectors 17a, 17b and 17c in the configuration shown in FIG. 5, the address field in the address area 23,
The address check byte 24 (see FIG. 8) may also be written separately to the sectors 17a, 17b, 17c.

Next, the configuration of the disk device according to this embodiment will be described.

FIG. 1 shows the configuration of a disk drive according to this embodiment.

The disk device 10 includes a disk controller 8
And a disk drive device.

The disk drive device includes a code decoding circuit 6 for encoding data into a storage code on the disk and decoding the storage code, and recording / reproduction circuits 7a, 7b to 7h for recording / reproduction on / from a magnetic disk. Have. A magnetic head is connected to each recording / reproducing circuit, and each magnetic head accesses the same or different magnetic disks in parallel.

The disk control device 8 includes a host interface circuit 2 for interfacing with a host computer, a disk control LSI 3 described later, a buffer memory 4 for storing data transmitted to and received from the host computer, and a microprocessor 5 for controlling the entire system. ing.

At the time of normal data writing, the disk drive device 9 writes a plurality of sectors 17a, b, and c (sectors having the same sector number on the formatted disk) of the parallel data received from the disk control LSI 3 on the formatted disk, as shown in FIG. Record in parallel).

At this time, it is necessary to encode / decode the data into the recording code for recording on the disk. In this embodiment, the encoding / decoding circuit 6 converts the data from the disk control LSI into a parallel code. As it is, and performs encoding / decoding processing in parallel (actually, 2-7 RLL)
Many recording codes, such as codes and 1-7 RLL codes, are simpler in circuit configuration when converting parallel data).

However, the coding / decoding circuit 6 may include coding / decoding circuits for the number of output data lines of the disk control LSI.

As described above, the disk device according to the present embodiment stores data in parallel on the magnetic disk, but the configuration is different from that of the conventional disk device 8 (see FIG. 6) which stores data serially. In order to transmit and receive data in parallel, a bus is connected between the code decoding circuit 6 and the control LSI 3.

Internal Configuration of Disk Control LSI Next, a disk control LSI according to this embodiment will be described.

FIG. 2 shows a disk control L according to this embodiment.
2 shows a configuration of an SI.

The disk control LSI includes a CPU interface unit 12, a buffer controller unit 13, a host interface unit 11, a drive interface unit 15,
It comprises a format control unit and an ECC unit 14.

The CPU interface section 12 is an interface circuit between the external microprocessor 5 for controlling the entire disk controller 8 and the present LSI, and controls the transfer of data and addresses between the microprocessor 5 and the present LSI.

The buffer controller 13 is a circuit for controlling a buffer memory provided to absorb a difference in data transfer speed between the disk drive 3 and the host computer 1. All data transfer between the disk drive 3 and the host computer 1 is performed via the buffer memory 4. Therefore, the data transfer controlled by the buffer controller 13 is performed by the disk drive 13 -the buffer memory 4, the buffer memory 4
-At least two systems of the host computer 1 are required, and the microprocessor 5 and the buffer memory 4 may be present in some cases.

The host interface control section 11 is a circuit for controlling data transfer between the disk control LSI 3 and the host interface control circuit 2, and handles signals between the host interface control circuit 2 and the disk control LSI.

When recording and reproducing the data sent from the host computer 1 to the disk device, the ECC unit generates and decodes an error correction code for correcting an error of the data caused by noise or a defect of the recording medium. It is a circuit to perform. Conventionally, a fire code is usually used as an error correction code in a hard disk device. However, in the case of a fire code, when a decoding circuit is realized by hard logic, it is necessary to store the data as serial data, so that a parallel configuration cannot be used. Therefore, in the present invention, data is handled in a parallel configuration using, for example, a Reed-Solomon code described in JP-A-63-91418 as an error correction method.

The drive control section 15 is a circuit for controlling the transfer between the disk control LSI 3 and the encoding / decoding circuit 6, and has a format control section 50.

The format control unit 50 is a circuit for controlling a format for writing sector management information necessary for writing and reading data, and for writing a synchronization signal used in a recording and reproducing circuit when reading data. When the disk device is used for the first time, this format must be performed. Since the data to be written at this time is not data sent from the host computer, the microprocessor 5 and the format control unit 50 perform the above-described operation for each sector. The data is generated and written in accordance with the storage format described above (see FIG. 5).

When writing the data sent from the host computer 1 to the disk device, the data 29 sent from the host computer is added to the error correction code 30 generated by the ECC unit and the other data area 19.
It is necessary to add and write the information (26, 27, 28, 31). The format control unit 50 also adds these pieces of information.

Next, in comparison with a conventional disk control LSI which stores data serially, FIG. 7 showing the features of the configuration of the disk control LSI according to the present embodiment is shown in FIG. Show.

As shown in the figure, the data is conventionally handled in parallel in the CPU interface unit 12 and the buffer control unit 13. However, since the ECC unit is a circuit configured to handle serial data, the data is converted into serial data by the parallel / serial conversion unit 16, and then the ECC unit generates an error correction code and sends it to the drive control unit. .

On the other hand, the disk control LSI according to the present embodiment employs the Reed-Solomon code in the ECC section and handles data in parallel.

The drive control section is also configured in a parallel configuration, and the entire LSI handles data in parallel.

The ECC unit has a conventional configuration, in which a parallel-to-serial conversion circuit is provided in the drive control unit, and the output of the drive control unit achieves a constant speed-up even if the data is parallel. I can do it. With regard to data transfer inside the LSI, a certain degree of speed can be obtained by increasing the transfer speed even by serial transfer, but it is difficult to increase the transfer speed with respect to data transfer outside the LSI. is there. That is, it is highly probable that data transfer with the outside of the LSI will be the rate-determining step of the process, and therefore, by parallelizing the data transfer with the outside of the LSI, the overall process can be speeded up.

Next, the internal structure of the format control unit 50 (see FIG. 2) will be described.

FIG. 3 shows the structure of the format control unit 50.

In the figure, reference numeral 51 denotes a microprocessor 5 (first
A sequencer 53 decodes a command from the system and controls each unit. Reference numeral 53 denotes a register for storing format ISG data 20a to 20c, various synchronization patterns 28a to 28c, and the like.
Reference numeral 4 denotes a drive interface unit which performs parallel / serial conversion and output of data based on an instruction from the sequencer 51.

FIG. 4 shows the drive interface unit 54.
Is shown.

In the figure, 55a and 55b are registers for storing write data, 56 is a selector for performing parallel / serial conversion of write data, and 57 is a buffer for outputting data selected by the selector 56.

The operation of the format control unit 50 during formatting will be described below.

First, the microprocessor 5 instructs the sequencer to that effect, and sends the format ISG data 2
0a-c, addresses, various synchronization patterns 28a-c, etc. are stored in the register 53. Next, the microprocessor 5
Instructs the sequencer 51 to execute the format.

In response to this, the sequencer 51 sequentially transfers the ISG data 20a to 20c, addresses, various synchronization patterns 28a to 28c, etc. from the register to the drive interface unit.

The drive interface unit outputs the ISG data 20a to the disk drive.
As described above, it is necessary to add each of the synchronization patterns 28a to 28c to each sector as described above.
Serialized data and output.

By the above operation, ISG data 20a-
c and various synchronization patterns 28a to 28c can be added to each sector. Also, normal storage data and the like can be stored in parallel by outputting without serialization.

Note that, instead of the parallel / serial conversion,
The microprocessor 5 may store a plurality of data in advance in the register 53 so that the bit string of each digit forms desired data.

Hereinafter, the normal operation of the disk device will be described.

At the time of data writing, the host computer 1 sends a SCSI (Small Computer Sys
A write command is sent to the host interface control circuit 2 from the tem interface).

The disk drive 9 positions the magnetic head at the target sector 17 in accordance with the command received via the host interface control circuit 2.

Thereafter, the host computer sends the stored data. This data is parallel data of 1 byte width. The stored data is transferred from the host interface control circuit 2 to the buffer memory 4 by the buffer control unit 13 of the disk control LSI, and further transferred to the disk control LSI.

The ECC unit 14 in the disk control LSI
Then, a 12-byte Reed-Solomon code, which is an error correction code, is added to this data, and the data is transferred to the code decoding circuit 6 via the drive control unit 15 as 1-byte width parallel data.

The code decoding circuit 6 has a recording code 1-7.
The data is converted into a code or the like, and each bit is written to the sectors 17a, b, and c corresponding to the address specified by the write command on the disk by the recording / reproducing circuits 7a, b having a plurality of systems.

When reading data The host computer 1 sends a read command to the host interface control circuit 2 from SCSI or the like. The disk drive 9 positions the magnetic head (s) at the sector (s) 17a, b, c corresponding to the address of the data specified by this command received via the host interface control circuit 2.

The recording / reproducing circuit 7 corresponding to each of these magnetic heads simultaneously outputs a plurality of sectors 17a,
Data is read from b and c, and the codec 6 corrects the data skew between the recording and reproducing circuits and decodes the data into parallel NRZ signals. The disk control LSI 3 receives this data, performs error detection in the ECC unit 14, and if there is an error, the microprocessor 5 controls the ECC unit 14
Error correction based on the error information from If there is no error, the data is transferred to the buffer memory 4 and further transferred to the host computer 1 via the host interface control circuit 2.

As described above, according to the present embodiment, in the disk device, the data lines of the entire disk control device can be configured in parallel, so that the data line in principle is parallel as compared with the case where there is a portion that has a serial configuration. The speed can be increased by the width.

Further, according to the present embodiment, only one disk control LSI is required, which has great advantages in terms of cost and mounting area, and does not complicate microprocessor processing.

In the above embodiment, an information processing system having a magnetic disk storage device has been described as an example. However, the present invention is not limited to a magnetic disk storage device, but an information processing system having an optical disk or a magneto-optical disk storage device may be used. The same can be realized in a processing system.

[0083]

As described above, according to the present invention, the disk control LS of the disk device adopting the parallel recording system, which can increase the data transfer speed with one LSI.
I can be provided.

Further, according to the present invention, it is possible to provide a disk device capable of increasing the data transfer speed, and an information processing system using the disk device.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a disk device according to an embodiment of the present invention.

FIG. 2 is a block diagram showing a configuration of a disk control LSI.

FIG. 3 is a block diagram showing a configuration of a drive control unit.

FIG. 4 is a block diagram showing a configuration of a drive interface unit.

FIG. 5 is an explanatory diagram showing a storage format on a disk.

FIG. 6 is a block diagram showing a configuration of a disk device according to a conventional technique.

FIG. 7 is a block diagram showing a configuration of a disk control LSI according to a conventional technique.

FIG. 8 is an explanatory diagram showing a storage format on a disk according to a conventional technique.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Host computer, 2 ... Host interface control circuit, 3 ... Disk control LSI, 4 ... Buffer memory, 5 ... Microprocessor, 6 ... Code decoding circuit, 7
... Recording / reproducing circuit, 8 ... Disk control device, 9 ... Disk drive device, 10 ... Disk device, 11 ... Host interface control unit, 12 ... CPU interface unit, 13 ... Buffer control unit, 14 ... ECC
Unit, 15: drive control unit, 16: parallel-serial conversion unit, 17: sector, 18: address area,
19: data area, 20: ISG (inter-sector gap), 27: pro sync byte, 28: sync pattern byte, 29: data field, 31: data pad, 50: format control unit, 51: sequencer, 5
3 ... register, 54 ... drive interface unit.

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Motoyasu Tsunoda 292 Yoshida-cho, Totsuka-ku, Yokohama-shi, Kanagawa Prefecture Inside the Microelectronics Equipment Development Laboratory, Hitachi, Ltd. (72) Inventor Toketsu Karasawa Yokohama 292 Yoshida-cho, Totsuka-ku, Hitachi Video Engineering Co., Ltd. 292 Yoshidacho Inside Microelectronics Device Development Laboratory, Hitachi, Ltd. (72) Inventor Yoshio Yukawa 2880 Kofu, Odawara-shi, Kanagawa Prefecture Inside Odawara Plant, Hitachi, Ltd.

Claims (7)

[Claims] A host interface for inputting / outputting data accessed by a host device; a disk interface for inputting / outputting storage data to / from a storage disk in parallel; and means for controlling data transfer between the two interfaces. And a disk control LSI. 2. A host interface for inputting / outputting data accessed by a host device in parallel, a disk interface for inputting / outputting storage data to / from a storage disk in parallel, and a parallel transfer of data between both interfaces. Means for controlling, and a code for detecting an error in the transfer data is obtained in units of parallel data, and is parallelized to the same number of bits as the transfer parallel data;
Means for appending to the parallel data transferred from the host interface to the disk interface, and a code for detecting an error in the transfer data appended to the parallel data transferred from the disk interface to the host interface, the error of the data of the transfer data being detected. A disk control LSI, comprising: means for detecting. 3. The disk control LSI according to claim 2, wherein a Reed-Solomon code is used as a code for detecting an error in the transfer data. 4. The disk control according to claim 1, 2 or 3, wherein sector management information of the storage disk to be written to the storage disk is output in parallel from the disk interface by the number of bits of the transfer parallel data. A disk control LSI characterized by comprising means. 5. The method according to claim 1, wherein parallel transfer of data between the storage disk and the storage disk is controlled.
A storage device comprising the disk control LSI described above. 6. A buffer memory for holding parallel data accessed by a host device, a disk control circuit for controlling parallel transfer of data between the buffer memory and the storage disk, and a recording code on the storage disk for data stored on the storage disk. A storage device, comprising: a coding / decoding circuit for performing coding on a recording code and decoding from a recording code; and a storage disk for storing storage data in parallel. 7. An information processing system comprising: the storage device according to claim 5; and a host device that accesses the storage device.