JPS60182063A - Magnetic disc device and its error detecting system - Google Patents

Abstract

PURPOSE:To attain error detecting processing in high speed by saving an error detecting code into a reliable storage means at all times and reading the error detecting code only twice at the start of use and end of use of a disc. CONSTITUTION:When a data read instruction is given from a host computer via an information cable 10, a CPU4 allows a drive section 2 to track a magnetic head and the data byte read by the drive section 2 under the control of a direct memory access control circuit 8 is transferred and stored in a data buffer 7. Each data byte is read at the same time with an adder 3. When the read of one sector is finished, the CPU4 reads the result of addition of the adder 3, and it is compared with an added data corresponding to an added value save memory 6. When coincidence is obtained, it shown that a normal read without error is attained. It is not necessary to read a CRC code at each sector access, and time used for error detection among the time required for one access of sector is neglected.

Description

[Detailed Description of the Invention] [Technical Field] The present invention relates to a magnetic disk device and its error detection method, and particularly to a magnetic disk device and its error detection method that have high error detection reliability and can perform error detection processing at high speed. It is related to.

[Prior Art] Conventionally, error detection in magnetic disk devices has been performed using, for example, a cyclic check character (CRC) provided at the rear of the data portion of a sector. That is, when reading a sector, a predetermined cyclic operation is performed on the data read out one after another, and when the CRC code is finally read and the operation is performed, it is considered good if the operation result becomes 0. If the value does not become O, it is determined that an error has occurred and the sector is reread. In order to implement such an error detection method, a CRC code must be written in the data frame of the sector in advance. Therefore, when writing a sector, it is necessary to perform a predetermined cyclic operation in the same manner as described above, and to write the CRC code of the operation result after the last data write. In this way, since the error detection code is read and written in both data read and write cases, the entire data access time is not used for the original purpose of accessing only data. . Therefore, the efficiency of data access is correspondingly low, which causes a reduction in the processing efficiency of the entire system. In addition, in conventional error detection methods such as CRC, data and error detection codes are read together from the same recording medium under the same conditions, so it can be said that there is a limit to ensuring reliability. In other words, even if the data part is read normally, if there is an error in reading the CRC code, it will end up being an error detection, and if there is an error in reading both the data part and the CRC code, it may not be detected as an error. Ta.

[Purpose] The present invention has been made in view of the two drawbacks of the prior art, and its purpose is to provide a magnetic magnetic field that can perform error detection processing at high speed and has high reliability in error detection. The purpose is to provide disk devices.

Another object of the present invention is to provide an error detection method for a magnetic disk drive that can perform error detection processing at high speed and has high reliability in error detection.

[Embodiment] Hereinafter, a magnetic disk device according to an embodiment of the present invention will be described in detail with reference to the accompanying drawings. FIG. 1 is a block diagram showing the configuration of a flexible magnetic disk device according to an embodiment. In the figure, 1 is a flexible disk that is a magnetic recording medium, 2 is a drive unit that controls the rotation of the flexible disk 1 and data reading/writing control, and 3 is a drive unit that controls the rotation of the flexible disk 1 and the reading and writing of data on the flexible disk 1.
an adder that performs addition key performance on data read from the
4 is a central processing unit (CPU) that is in charge of the main control of the disk device, 5 is a program memory that stores the program executed by CPU 4, and 6 is an added value save that holds data added values for each sector, which will be described later. Memory, 7 is a data buffer that holds data for reading and writing to/from the flexible disk 1, and 8 is a tie left memory access that controls automatic transfer of data exchanged between the flexible disk 1 and the data buffer 7. A control circuit (DMAC), 9 an internal common path of the CPU 4, 10 an information cable connecting the magnetic disk device and an external host computer (not shown), 11 a connection between the magnetic disk device and the host computer Interface circuit for exchanging necessary information, 12
13 is an operation switch for locking the device after loading the flexible disk 1 into the device, or unlocking the device when removing the flexible disk 1 from the device; An interrupt signal line 14 is used to exchange signals between the CPU 4 and the adder 3. A signal line 15 is used to exchange signals between the CPU 4 and the DMAC 8. Signal line, 16 is DMAC8
is a signal line for supplying address signals and the like to the data/long buffer 7.

FIG. 2(A) shows the recording format of the flexible disk 1 used by the apparatus of this embodiment.

In the figure, 30 to 32 are sectors that are access units for data reading and writing, 20 is a track consisting of a set of sectors,
21 is the innermost track among the plurality of tracks existing on the recording surface "2". FIG. 2(B) shows the storage structure of the added value save memory 6. (In A, 60-62
is a save area for each addition value configured in bytes. For example, the result of adding all the data contained in sector 30 is saved in the save area 60. Similarly, a save area 61 is provided corresponding to the addition result of sector 31, and a save area 62 is provided corresponding to the addition result of sector 32. In this way, a save area exists at each corresponding address for every sector. FIG. 2(C) shows the recording format of l sector. In the figure, an l sector consists of an indicator (ID) section containing address information of the sector and a take section. Here, the frame of the take section does not include information for error detection such as a cyclic check character (CRC). Everything is good with useful takes.

The operation of the above configuration will be explained below. FIG. 5 shows the lock interrupt processing when the flexible disk (2) is loaded in the disk device and the lock operation is performed. When the flexible disk 1 is loaded into the apparatus, a disk loading interrupt is generated in the CPU 14 via the interrupt signal line 13. In step 351, the disk drive is put into a locked state via the drive section 2. In other words, the motor is rotated to make the disc readable. In step S52, each added value data for all sectors stored in advance in the innermost track 21 of the flexible disk 1 is read out and stored in the added value save memory 6. When the flexible disk 1 is used for the first time, nothing is written on the path recording surface. Therefore, no useful data is obtained in the process of step S52. However, as will be described later, when a useful take is written to a certain sector of the flexible disk 1, the useful addition η value is
i1 is calculated and the addition result is saved at the corresponding address in the added value save memory 6. When the disk 1 is removed after use, the entire contents of the added value save memory 6 are dumped onto the innermost track 21 of the disk 1. Therefore, when the flexible disk 1 is used for the second time or later, meaningful data is loaded into the charge value save memory 6 by the process of step S52.

FIG. 3 shows data and cleaning processing for one device of this embodiment. When a data read command is issued from the host computer via the information cable IO, the CPU 4 receives the command and necessary address information via the interface circuit 11. The CPU 4 controls the drive unit 2 according to this address information.
A reading operation is started in which the magnetic head is tracked and the indicator part of each sector is checked. Eventually, when the sector is found and its data portion is read, it is input to the data reading process shown in FIG. In step 33, the adder 3 is initialized. In step S32, the DMAC 8 is activated. Thereafter, the data bytes read by the drive section 2 under fl'd'J of the DMAC 8 are transferred to the data buffer 7 via the internal common path 9 and stored therein.

At the same time, each data high 1 on the internal common bus 9 is added by the adder 3. The addition is performed every time new data is read into the internal common bus 9L. The addition is performed in such a way that the value that overflows the most significant bit of adder 3 is ignored. In step S33, the CPU waits for completion of reading of l sector. After reading l sectors, step S
34, the addition 1'J of the adder 3 is read through the Crt 14 or the signal line 14, and the result is compared with the corresponding addition (+Ci data) in the added value save memory 6. In step S35, it is determined whether or not a match is obtained in the comparison. If a match is found, it indicates that normal readout without errors has been performed, and the flow exits.If a match is not found, the process advances to step 336 and an error flag is set.Later, the main routine (see Fig. (shown) recognizes this error flag and performs a retry operation to read the sector.In this way, for sectors in which valid data has been written in advance on the flexible disk, addition of that data frame is performed. The value is written in the corresponding location of the innermost sector, rank 21.The added value save memory 6, which is loaded with this when the flexible disk l is loaded, holds the added result of the sector at the corresponding address. Therefore, when the sector is read, an error can be detected immediately by simply comparing the addition result of the adder 3 and the addition value of the addition value save memory 6. The example device does not need to perform error detection using the cyclic check character (CRC) method that was conventionally used.
This means that there is no need to read out the RC code. Therefore, the time used for error detection within the time required to access sector 1 can be ignored, and data access efficiency is greatly improved.

FIG. 4 shows the data writing process. When a data write command is issued from the host computer, the CPU 4 searches for the relevant sector on the flexible disk and at the same time stores the write data from the host computer in the data buffer 7. Eventually, when the sector is detected and the data portion thereof is written, input is made to the data writing process shown in FIG. 4. In step S41, the adder 3 is initialized. In step S42, the DMAC 8 is activated via the signal line 15. The data byte read from the data patcher 7 is transferred to the internal common bus 9.
is sent to the drive unit 2 via. At the same time, the adder 3 adds each data byte output to the internal common/hess 9.

This addition is performed every time new data appears in the internal common /< space 9. The addition is performed in such a way that the value that overflows the most significant hit of the adder 3 is ignored. In step S43, it waits for the writing of one sector to be completed. Eventually, when writing of one sector is completed, the addition result of the adder 3 is read through the signal line 14 in step S44, and is saved in the address corresponding to the write sector in the added value save memory 6. In this way, the contents of the save memory 6 are updated with a new addition value each time new data is written to a sector. When writing, the error detection code of the cyclic check character (cRc)-i is not written.The data frame of 1 sector consists only of data.Therefore, the writing time is saved each time it is accessed. As a result, data writing efficiency increases.

Generally, when writing data to a sector is completed, the disk waits for rotation and a data read check is performed. This is a so-called read-after-write check. In this case, a process corresponding to the data read process shown in FIG. 3 described above is executed.

In other words, error detection processing is performed, but data writing processing to the data buffer 7 is not performed. In this way, even when performing a door after write check, the contents of the additional value safe memory 6 have already been updated 5 when data was written one revolution before the disk, so a match can be obtained without any problem.

FIG. 6 shows the unlock interrupt process when removing the flexible disk l. When a series of data access processing by the host computer is completed, the flexible disk 1 is removed from the apparatus. At this point, when the operator presses the operation switch 12 again, a disk unlock interrupt is generated in the CPU 4 via the interrupt signal line 3. In step S61, the added value saved in the added value save memory 6 is tamped onto the innermost track 21 of the flexible disk l. In step S62, the locked state of the disk device is released. At this point it is possible to remove the flexible disk 1. The removed flexible disc can be used again.

Incidentally, in the explanation of the above-mentioned embodiment, the calculation for error detection was performed by addition by the addition means, but the invention is not limited to this.

It is clear that conventional CRC calculation means may be used in place of the addition means.

Furthermore, in the description of the above-mentioned embodiments, the data part of the sector has no error detection code at all. However, the error detection method according to the present invention does not perform the conventional CR
It can be used together with the C error detection method. Conventional error detection methods such as CRC have limitations in ensuring reliability because both data and error detection codes are read together from the same recording medium under the same conditions. In other words, even if the data section is read normally, if there is an error in reading the CRC code, an error will be detected eventually, and if there is an error in both the data section and the CRC code, error detection may not be performed. Therefore, if the error detection method according to the present invention is used in 4Jr, such disadvantages can be avoided since the error detection code is stored in a reliable storage means.

[Effects] As described above, according to the present invention, the reliability and efficiency of data access in a magnetic disk device are extremely high.

Conventionally, error detection codes were read and written every time data was accessed. However, according to the present invention, it only needs to be done twice: at the start of use of the disk and at the end of use. Otherwise, no matter how large the amount of data is accessed, there is no time required to read the error detection code.

, the aj-judge detection code is always saved in reliable storage means and updated as necessary. Therefore, the more a large amount of data is accessed, the more processing efficiency and reliability will be improved.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the configuration of a flexible magnetic disk device according to an embodiment of the present invention, FIG. 2(A) is a diagram showing the recording format of a flexible disk used by the device of this embodiment, (B) is a diagram showing the storage structure of the added value save memory, 2nd (Δ(C) is a diagram showing the recording format of l sector, 3rd is a flowchart showing the data read process of the device of this embodiment, 4th Figures 1 and 5 are flowcharts showing the data writing process of the device of this embodiment, Figure 5 is a flowchart showing the lock aging process when loading a flexible disk, and Figure 6 is a flowchart showing the unlocking process when removing the flexible disk. This is a flowchart 1 showing the processing of processing.Here, 1...Flexible disk, 2...Drive unit, 3...Adder, 4...Central processing unit (CPU), 5... ...Program memory, 6
...Additional value save memory, 7...Data harassment 2a,
8...Direct memory access control circuit (DMAC)
), 9... Internal common path, 10... Information cable,
11...Interface circuit, 12...Operation switch, 13...Interrupt signal line, 14-16...13 line. Figure 2 (A) (B) -111111-i'Nuk rotation direction Figure 3

Claims (3)

[Claims] (1) In magnetic disk devices used for information processing,
creating means for creating an error detection code for the written data when writing data; storage means for storing the error detection code created by the creating means at a predetermined address; A magnetic disk drive characterized by comprising error detection means for creating an error detection code and comparing it with an error detection code read from a predetermined address of the storage means to determine whether or not there is a match. (2) In an error detection method of a magnetic disk device used for information processing, when writing data, an error detection code is created for the write take, and the error detection code is stored at a predetermined address of a storage means; A magnetic disk device characterized in that when reading data, an error detection code is created for the read data, and this is compared with an error detection code read from a predetermined address of the storage means to determine whether or not there is a match. Error detection method. (3) When the use of the magnetic disk device is finished, the error detection code stored in the storage means is dumped into a predetermined area of the magnetic recording medium, and when the use of the magnetic disk device is started, the error detection code stored in the storage means is dumped into a predetermined area on the magnetic recording medium. 3. An error detection method for a magnetic TiSF device according to claim 2, characterized in that an error detection code in the rear is loaded into said storage means.