JPH0354775A - Disk device and data write/read method for the same - Google Patents

Abstract

PURPOSE:To improve a rate occupied by stored data to storage capacity by writing and reading more than two data blocks to one sector equipped with one ID field on a disk. CONSTITUTION:A sector 12 is composed of a GAP 1 part 6, ID field 7, PAD 8, DATA field 9, PAD 8 and GAP 2 part 10. In the ID field 7, the cylinder, head and sector numbers of the sector 12 and the check flag of this area are written. Accordingly, the more the number of heads to simultaneously write/read the data to a disk 14 is increased, the more an effect to provide an n-piece of data blocks 5 in one sector 12 is increased. Thus, efficiency for using the disk can be improved.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to writing or reading data in a disk device. [Prior Art] As described in Japanese Unexamined Patent Publication No. 65-298868, a conventional device uses four magnetic heads out of j magnetic heads to read data divided into pieces. I was using tx to write to a magnetic disk. Furthermore, out of the j magnetic heads, l magnetic heads are used to simultaneously read data from the magnetic disk, synthesize it, and output it. This enabled high-speed data writing/reading.

[Problem to be solved by the invention]

The above conventional technologies focused on high-speed data writing/reading. However, in simultaneous writing/reading by multiple heads, if the logical data block length sent from the host is the same, the length of the DATA field that stores data decreases as the number of heads writing/reading simultaneously increases. .

Even in the above case, the sector management information, such as the number representing the location of the sector on the disk device, decreases by -1. Therefore, when one data is one sector, the ratio of the length of the sector management information is larger than the data length divided into the number of heads that write/read simultaneously in the disk device, and the storage of the disk device is There was a problem in that the proportion of stored data was extremely low compared to the capacity.

An object of the present invention is to significantly increase the ratio of stored data to storage capacity, particularly in a disk device that performs simultaneous writing/reading using multiple heads. [Means for Solving the Problems] In order to achieve the above object, in the present invention, R data blocks (R = 2 or more) are collected in one sector.

The above data block consists of data obtained by dividing a logical data block, which is the basic unit of data sent from the host, into the number of heads that can be simultaneously written or read by the disk device, and data obtained by dividing the logical data block, which is the basic unit of data sent from the host, into the number of heads that can be written or read simultaneously by the disk device. Error correction code (error correction code) or +16.

Also, if one sector has 1 data blocks and there is a bad data block in that sector, write/! ! In order for the controller to know the number of usable data blocks, information indicating the number of usable data blocks is recorded in the management information area of the sector.

Here, the error correction code in the data block is used to independently correct data in the core data block in order to identify a defective data block.

Furthermore, since defective data blocks can be identified, it is possible to combine data blocks that effectively utilize the storage capacity of the disk device according to the number of heads that can simultaneously write/continue data.

Further, by providing a head number designation circuit and designating the number of heads according to the number of defective data blocks, the storage capacity of the disk device can be effectively utilized.

[Effect]

By collecting multiple data blocks (3 is 1 or more) in one sector, the sector management information required for each sector can be reduced by the number of sector management information per sector x (a-1), and the disk device It is possible to improve the ratio of storage data to the storage capacity of .

Further, the error correction code for error correction of each data block can independently correct the data of the data block, and is used to specify which data block is defective.

When one sector has one data block and a bad data block exists in the sector, the controller can write/continue to record information indicating the number of usable data blocks in the management information area of the sector. The number of data blocks in a sector can be known and the storage capacity of the disk device can be prevented from decreasing by -1. Furthermore, by providing a head number specification circuit that allows the user or host to specify the number of heads that can write/read at the same time, the controller can identify bad data blocks and adjust the number of heads according to the number of heads that can write/read at the same time. You can select a combination of data blocks that makes effective use of the storage capacity of the disk device. [Embodiment] An embodiment of the present invention will be described below using a magnetic disk device as shown in FIG. Figure 2, Figure 3, Figure 4. Figures 5 and 6. 7th
This will be explained with reference to FIGS.

FIG. 1 is an example of a data arrangement diagram when the present invention is applied to a magnetic disk device. In this embodiment, the data sent from the host is divided into eight pieces in advance and simultaneously recorded on the disk 14 by eight magnetic hexadecimal units 36 from head number 0 to head number 7. This is a case where data simultaneously read from the disk 14 by eight magnetic heads 36 can be combined and sent to the host. Sectors 12 are created in advance on the disk 14 as units for managing data. This is created by formatting the disk device. The sector 12 is . GAP1 part 6. ID field 7. P.A.D.
8, DATA field 9, PAD8 and GAP2
Consisting of 10 parts. GAP i section 6 and GAP 2 section 10 are
This indicates the boundary between sectors 12. Especially GAP 2 part 1
0 is provided to prevent the possibility of destroying the next sector 12 due to (I) rotational fluctuations of the spindle motor 16 during a write operation to the ATA field 9;

The ID field 7 indicates the cylinder, head,
The sector number and check flag in this area are praised. PAD8 indicates the boundary of the DATA field 9, and when writing to the DATA field 9, the read state and write state are switched here.VXY line 5.

DA TA field 9 is the part where data is written, and is the part where data is written. Address mark 2 parts 2. Warning from data block 5. The sink 2 unit 1 uses the controller's VFO (
VariabLa FrgquaI%cy OzoiL
This is a memorial area for synchronizing with Latar). A special pattern is written in the address mark 2 part 2, and by detecting this pattern, the controller 2 can know that the next data block 5 starts.

Data block 5 may also include data 3 and error correction code 4. The error correction code 4 is an error correction code for preventing erroneous data from being read due to a defect on the disk 14 or the like. Index 11 is
The GAPS section, which is a signal that indicates the beginning of the 7'J:<,} rakukuku, occurs only once per revolution per track, is intended to cope with speed fluctuations of the spindle motor 16 when the index 11 comes. Here, b pieces of the above data block 5 are placed in the data field 9 (2 or more outside).
GAP required in sector 12 by having
1 part 6, ID field 7, PADB, sink 2 part 1l
Address mark 2 part 2 and GAP 2 part 10 are considered saving. In particular, as the number of heads that simultaneously write on the disk 14 and thread out increases,
The data 3 corresponding to 6 is small < tx, and the effect of having 1 data blocks 5 in one sector 12 becomes large. Of course, the data arrangement on the disk 14 can be arranged in other ways than in this embodiment.

Figure 2 is an example of the conventional sector arrangement and size of a magnetic disk device. The role of each area is the same as in Figure 1. Assuming that one data block is in one sector as in the past, when writing/reading is performed simultaneously by eight heads, one data management unit is 512 Byta x 8 = AO96 Byta. In disk devices, data is written or read in data management units, but data is written or read in less than data management units. ! When data is removed, an unused area is created, resulting in poor disk capacity efficiency. The ratio of the data 3 area per track when the same data management unit is written or output using one conventional head and eight heads is shown below.

(l) Writing or reading by one head [1 data block/sector] 1 data block length: 51 BByts/sector (including 6 Bytm ECC) Number of sectors/track = 51 sectors/track sector length
: 587 BytgGAP S length :
Ratio of 303 Byte data area/track 512 X I X 51 / 50240 = 0.
86 {2} Writing or reading with 8 heads [1
Data block/sector] 1 data block length: 70 Byta/sector (including 68 yta ECC) Number of sectors/track: 217 sectors/track Sector length: 159 Byte Data area ratio/track 64 X I X 217 / 3024 o = 0.4
6 (3) Writing or reading by 8 heads [8 data blocks/sector] 1 data block length: 70 Bytg/sector (including 6Eyta ECC) Number of sectors/track: 48 sectors/track sector length
: 629 ByteGAP3 length :
Ratio of 48 Bytg data area/track 64 X a
The ratio of data area in data block/sector is α
46, while 0.46 for 8 data blocks/sector.
81, which greatly improves disk usage efficiency. Since the disk usage efficiency when writing/reading is performed by one head is 0.86, it can be seen that this is almost equivalent to the case of 8 data blocks/sector.

Of course, in this embodiment, the usage efficiency of the disk changes depending on the number of heads that write/read data simultaneously and the number of data blocks that fit into one sector.

Figure 5 is an example of an ID field data layout diagram. The ID field 7 has a sink 1s17, an address mark 1 part 18. Header 19. CRC (Cycle
icRadundasay Chuck) Starts from 20. The sink 1 section 17 is connected to the next succeeding header 19 .
Controller VF for reading CRC 20
This is an area for synchronizing with O. address mark 1
In the section 18, a special pattern is written, and by detecting this no turn, the controller can know that the next header 19 will start. Header 19 is
Cylinder number 21, head number 22. Sector number 23.
From flag 24. The controller has the cylinder number 21. By reading head number 22 and sector number 23, it is confirmed whether the sector is to be processed by the controller. Furthermore, by storing the state of the sector in the flag 24, the controller can know the state of the sector. The CRC 20 is a cyclic code for checking whether there are any errors in the read data of the header 19. In the case of conventional writing or reading from one sector, if the DATA field 9 has a defect and becomes a bad sector, the controller can detect the bad sector by recording a predetermined code in the header 19.

However, as shown in Figure 1, in the case of simultaneous writing or reading by 8 heads and 8 data blocks/sector, with the conventional method, even if one data block is defective, the above sector may be considered a bad sector, or 7 normal data blocks may be Blocks were sometimes wasted. Of course, this embodiment is just an example, and ID field data arrangements other than this embodiment can be adopted.

Figure 4 is an example of a flag diagram. This is an example of the allocation when the flag 24 is set to 1 byte as shown in FIG. 6, and the number of bytes may be allocated according to the disk device κ as necessary. The flag 24 is a defective track flag and a replacement track flag. Bad sector flag, replacement sector flag.
From the bad data block number. The bad track flag indicates that the bad sector is on a bad track when this bit is set. The alternate track flag indicates that the sector is on the alternate track when this bit is set. The bad sector flag indicates that the sector in question is a bad sector when this bit is set. The alternate sector flag indicates that the sector in question is an alternate sector when this bit is set.

The defective data block number indicates the number when there is a defective part in the sector. This is an example of a situation in which the controller can know the number of usable data blocks in the sector; other information may also be used. For example, in a particular track 15, the information of the above-mentioned ID field 7 of all parts in which defects exist is recorded. When the a1 disk device is turned on, it reads out the track.
A Jf (R a atLλccazz NonO
ry).〈●When a magnetic disk device receives a command from the host, if necessary, it stores it in the RAM.
By reading out the information on defective tracks and bad sectors stored in the target sector 12, the same effect as directly reading the In field 7 of the target sector 12 and knowing the defective state can be obtained. In this case, the 7 lag 24 of the ID field 7 contains
Possible options include recording how many data blocks 5 are used or eliminating the flag 24. If there are several bad data blocks in one sector,
Record the first defective data block number. This is because if a defective point exists, there is a high possibility that the VFO will become out of synchronization and data will no longer be able to be read. By recording the state of the sector in the flag 24 as described above, the controller can know the state of the core sector.

Figure 5 shows a data block layout diagram.
This is a case where writing or reading can be performed simultaneously with eight heads. In addition, in the case of a normal disk device, data blocks can be provided on both sides of the disk 140.
For the sake of explanation, I decided to have a data block on only one side. Each data block will have its own error correction code for error correction. By doing this, you can identify which data block is bad.
As shown in Figure 5, DATA 21 and DATA 51
is the bad data location. When writing or reading is performed simultaneously using eight heads, these eight data blocks become defective locations. If the user of the disk drive does not require a data transfer rate high enough to write or read simultaneously with eight heads, write or read simultaneously with eight heads.
5! The controller is equipped with a circuit for specifying the number of heads to be output. This allows the user to write/read rows 5 and 5 at the same time.
Determine the number of heads by magnetic decoder. This can be specified by providing a switch on the disk device. Alternatively, a method of specifying the number of heads for simultaneous writing/reading by sending a command from the host may also be considered.

This allows the controller to identify which data block is defective when writing or reading with two heads at the same time.I) ATAO1 and DATA11
．． DATA 31 and DATA 41. DATA
61 and DATA 71 combination to write or! !
It is possible to take out. If an error correction code is created for each data management unit that is written or read at the same time, it will be difficult to identify bad data blocks. When there is a defective data block as shown in Fig. 5, and it is possible to write or read simultaneously with two heads, the I
)ATAO1 and DATA 11, DATA61 and D
Writing or reading is possible in combination with ATA710, and normal data blocks DATA 31, DATA
41 is wasted 6 FIG. 6 is an example of a circuit configuration diagram of a magnetic disk device when simultaneous writing or reading is performed using eight heads. This embodiment uses the standard fK. SCSI (SrsaLL Censputar Sy)
ztatx Interface) is taken as an example. The operation of the circuit will be explained using an example in which a Write command is sent from the host. CDB (CammancL I) azcripto of the IFritg command with the protocol according to the SCSI convention from the host.
rBlock) and data are sent to the disk device via the SCSI bus. At this time, the SCSI*J control section 25 controls the scsr protocol. The microprocessor 26 interprets the command sent from the host and determines that it is a write command. Also,
Data sent from the host passes through the data bus to the data buffer 2 under the control of the data buffer controller 27 activated by the microprocessor 26.
8 is saved. Next, the microprocessor 26
) C (Hard Di z▲CantroNar
) The HDC control circuit 29 of 34 is turned off after the write settings are made. The HDC 34 retrieves data from the host from the data buffer 28 via the data buffer controller 27 according to the settings. The data taken out from the data buffer 28 is transferred to the head number designation circuit 31.
Transfer the data to the number of heads according to the data portion II/synthesizing circuit 3
Separate at 0. The separated data is transferred to the bus width conversion circuit 5.
An error correction code is added through the read/write circuit 35 to the head 36 to the disk 1.
Data is recorded in 4. As in this case, defective data blocks can be identified by separating the data sent from the host in advance using a data division II/combination circuit. Processing a read command from the host is the opposite of writing processing. The HDC 34, which has been set to read by the microprocessor 26, receives the target data from the disk 14 through the head 36 and the read/write circuit 35.

For the data passing through the bus width conversion circuit 33, an error correction code is created in the error detection/correction circuit 32 and compared with the error correction code read from the disk 14. If an error is detected, correction is performed, and if the error cannot be corrected, the microprocessor 26 is notified from the HDC control circuit 29. The data that has been processed normally is 111 m/synthesizing circuit 30, The data read from the disk device is returned to the state before separation, and the data read from the disk device is sent from the scsr control unit 25 to the host via the SCSI path via the data bus sofa 28.The above circuit S command is installed in the magnetic disk device. Accordingly, the controller determines the number of heads for simultaneous writing/reading designated by the user using the head number designation circuit 31, and performs writing/reading at the time of rifJ using the number of heads designated by the user.

As a result, it is possible to realize a magnetic disk device that can minimize the data transfer speed difference between the host and the magnetic disk device and set the efficiency of the entire system optimally.

FIG. 7 is an example of a circuit configuration of a magnetic disk device when simultaneous writing/reading is performed by eight heads having a head switching circuit 57. The function of this embodiment is the same as that shown in FIG. Here, the head switching circuit 37 associates the read/write circuit 35 with the head 36. For example, a case will be explained in which a processing request for a Write command is received from the host. As in the case explained in Figure 6, the host sends the command CDB and data pL SCS using the protocol according to the SCSI conventions.
It is sent to the disk device through the I / { screen. The command sent from the host is sent to the microprocessor 26.
It interprets the information and instructs the settings and startup of controllers such as the HDC 34. The data stored in the data buffer 28 is separated into the number of heads according to the number of heads designation circuit 31 and sent to the read/write circuit 35 . The data passed through the read/write circuit is recorded on the disk 14 from the head 36 designated by the head switching circuit 37. At this time, the read/write circuit 35 and the head 36 are associated with each other based on the defect information recorded in the head switching circuit 3 and the 7 lags 24 of the ID field 7. For example, it can be determined from the flag information in the ID field 7 of this sector that the microprocessor 2
When it is determined that the data is to be transferred to the disk 14, the microprocessor 26 connects the read/write circuit 0 to the head switching circuit 37 and the head 8 in which there is a spare spare sector. can be recorded. Furthermore, by providing the head switching circuit 67, data can be recorded/read even if the head 36 is located above the read/write circuit 35. As a result, read/write circuits 55 for the number of heads, error detection/
Correction circuit 52. There is no need to provide the cross width conversion circuit 35, and the cost of the circuit can be reduced. By installing the above circuit structure in a magnetic disk device, the read/write circuit 35 and head 36 can be combined in an optimal fK based on the defect information recorded in the flag 24 of the ID field 7. Since the microblock sexer 26 can be specified to
Since data can be recorded/read quickly as it reaches the disk 14, an efficient magnetic disk device can be created. FIG. 8 is an example of an operation procedure during data writing when eight heads each having a head switching circuit 37 perform simultaneous writing. In this case, the number of circuits that can be written or read simultaneously is nine. First, when the head 36 reaches the target sector, the ID field 7 is read in the head number 0 or 809 sector. Next, the shilling number 21 of the ID field 7 read. After confirming that the head number 22 and sector number 23 are the values of the target sector, the contents of the flag 24 are checked. If there is no defect information in this flag, the microprocessor 26 instructs the head switching circuit 37 to connect the read/write circuit 55 and the head 56 in numerical order. Then, data corresponding to the number of data blocks in the sector 12 is recorded on the disk 14. If there is still data to be written next, the same process described above is repeated until the ID field of the sector in which the next data is to be recorded is reached. Furthermore, if the data to be written is fK, then the writing has been completed and the writing process is ended. If flag 2
If defective information is found in the content check in step 4, check whether it is possible to write to it simultaneously with 8 heads out of 9 heads. If possible, the microprocessor 26 instructs the head switching circuit 37 to use the read/write circuit 35 and the head 36 in such a way as to disable the defective data block. Then, data corresponding to the number of data blocks that can be written is recorded on the disk. and,
The process then moves on to checking whether there is any data to be written. If it is not possible to write simultaneously with eight heads, the process moves to waiting until the next sector ID field arrives.

As described above, since data can be stored on the disk 14 and recorded, an efficient magnetic disk device can be constructed. It goes without saying that the present invention is applicable not only to magnetic disks but also to general disk devices such as optical disks.

〔Effect of the invention〕

According to the present invention, by inserting several Oatα fields in one sector that are small due to simultaneous writing/reading by multiple heads, the sector management information is prevented from becoming too large for the data, and the disk device is Capacity usage efficiency can be greatly improved.

Also available in the sector ID field is 1. By recording information indicating the number of data blocks, the data blocks from the beginning of the sector to the defective data block in the Dαtα field can be used for data recording and reproduction, thereby preventing a reduction in the capacity of the disk device.

Furthermore, by attaching an error correction code to each data block that can independently detect and correct errors, defective data blocks can be identified. This makes it possible to minimize the number of defective and defective data transfers. By making it possible to specify the number of heads that can write or read at the same time, the controller can select combinations of data blocks equal to the number of heads that can write or read at the same time, excluding defective data blocks. The number of data blocks required can be kept to a minimum.

[Brief explanation of drawings]

Fig. 1 is a data arrangement diagram of an embodiment of the present invention, Fig. 2 is a conventional sector arrangement diagram, Fig. 3 is an ID field data arrangement diagram, and Fig. 4 is a 72g diagram of an embodiment of the invention. FIG. 5 is a data block layout diagram of one embodiment of the present invention, FIG. 6 is a circuit diagram of a magnetic disk device of one embodiment of the present invention, and FIG. 7 is a magnetic disk device of another embodiment of the present invention. The circuit configuration diagram, FIG. 8, is a diagram showing the operation procedure during data writing according to the present invention. 5... Data block 7...1...
......ID field 9...
・DA TA field 30 ・・・・・・・・・ Data portion lllll/combination circuit 31 ・・・・・・・・・Head number designation circuit ￥51 Figure T7 YL Figure 5 rD7E-/ LFf'-7m Otsu 11Xl (!p
Place: 8゛I F) 17 Shi〉Cl 沓P +5.・Header +6
-7Ft, 17-71u 20-CF. CZ1
Solk 4r now 2Z head 4lrJ! 25 Deceased number z4-flag spear 4 figure 7 love figure man 5 figure table 71"f: L/7 layout diagram figure 6

Claims (1)

[Claims] 1. A disk device for writing or reading data on a disk, characterized in that two or more data blocks are written to and read from one sector having one ID field on the disk. disk device. 2. A disk device having two or more heads that simultaneously writes or reads data on a disk, characterized by writing and reading two or more data blocks in one sector with one ID field on the disk. disk device. 3. It is equipped with one or more disks, has two or more surfaces for writing or reading data, and has two or more heads for writing or reading data at the same time, and the head is connected to the disk surface on which the data is written or read. 1. A disk device, each having at least one disk device, in which two or more data blocks are written to and read from one sector on the disk having one ID field. 4. The disk device according to claim 1, 2, or 3, wherein each of the two or more data blocks is provided with an ECC (error correction code) for independently correcting data in the data block, and the data block is A disk device characterized by writing and reading. 5. In the disk device according to claim 1, 2, or 3, information regarding a defective data block (data block management information) is written in an ID field in the sector;
A disk device characterized by reading and reading data. 6. The disk device according to claim 1, 2, or 3, wherein the number of defective data blocks in the sector is written into and read from an ID field in the sector. 7. The disk device according to claim 1, 2, or 3, wherein each of the two or more data blocks in the sector is numbered, and the number of the defective data block is written in an ID field in the sector; A disk device characterized by reading. 8. The disk device according to claim 1, 2, or 3, wherein each of the two or more data blocks in the sector is numbered, and the ID field in the sector is filled with the number of the first defective data block in the sector. A disk device characterized by writing and reading the number. 9. The disk device according to claim 5, wherein a storage means is provided in addition to the disk device, and the data block management information written in the ID field is stored in the storage means other than the disk. Device. 10. The disk device according to claim 5, 6, 7, or 8, wherein all data block management information is collectively stored in a part of the disk. 11. The disk device according to claim 2 or 3,
A disk device characterized in that data is written and read by designating heads that simultaneously write or read data based on information regarding defective data blocks. 12. The disk device according to claim 2 or 3,
A disk device that writes and reads data by specifying the number of heads that simultaneously write and read data. 13. The disk device according to claim 2 or 3,
head number designating means for storing information regarding defective data blocks in the disk or a storage means other than the disk, and for designating the number of heads for simultaneously writing or reading data; and a means for storing information, based on the information regarding the defective data block and the information on the number of heads,
A disk device characterized in that data is written and read by specifying the above-mentioned heads for simultaneous writing and reading. 14. In a disk device having two or more heads that simultaneously write data on a disk, one head on the disk
A disk device characterized in that two or more data blocks are written in one sector having one ID field. 15. The disk device according to claim 14, wherein two or more of the heads to write data simultaneously are designated to write the data based on information regarding the defective data block. 16. The disk device according to claim 14, wherein the above-mentioned 2.
A disk device characterized in that data is written by specifying the number of heads that simultaneously write data among the above heads. 17. In a disk device having two or more heads that simultaneously read data on a disk, one head on the disk
A disk device characterized in that two or more data blocks written in one sector having one ID field are simultaneously read by two or more heads. 18. The disk device according to claim 17, wherein two or more of the heads that simultaneously read data are designated to read the data based on information regarding the defective data block. 19. The disk device according to claim 17, wherein the above-mentioned 2.
A disk device characterized in that data is read by specifying the number of heads from which data is to be read simultaneously among the above heads. 20. In a data writing/reading method for a disk device that writes or reads data on a disk using two or more heads at the same time, two or more data blocks are written and read in one sector with one ID field on the disk. Disk device data writing/reading method. 21. A method for writing and reading data in a disk device according to claim 20, characterized in that the heads that simultaneously write or read data are specified to write and read data based on information regarding defective data blocks. Data write/read method. 22. The method for writing and reading data in a disk device according to claim 20, wherein the data is written and read by specifying the number of heads that simultaneously write or read data.