JPH1050004A - Magnetic disk control device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To prevent a throughput of a system from deteriorating in retrial processing at the time of detecting a data error. SOLUTION: The device is equipped with a CRC(cycle redundant check) code checking circuit A15 for checking a data and a CRC code from a magnetic disk device 20 at the time of reading out the data, a CRC code checking circuit B16 for inverting and checking the CRC code while the data is kept unchanged as it is and an error discriminating circuit 17 for discriminating whether the generated data error is a true error or a data error caused by a countermeasure to incomplete writing as a result of checking operation of the CRC code checking circuit A15 and the CRC code checking circuit B15. Consequently, only a report to a host device is required without necessity of extra retrial processing depending on the contents of a data error, thus preventing the throughput of the system from deteriorating.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a magnetic disk controller, and more particularly, to a data error detecting means of the magnetic disk controller.

[0002]

2. Description of the Related Art Conventionally, a magnetic disk control device has an error-correcting code (hereinafter referred to as "ECC") for generating a syndrome indicating an abnormality during data reading and a circuit for generating a syndrome indicating an abnormality in a transfer system. Redundancy Check (hereinafter Cyclic Redundancy Check: CR
ECC for a variable-length block containing a code
After performing error correction using the CRC code, error detection using the CRC code is performed. When a data error is detected, a correction pattern and a correction position based on the ECC detected by the magnetic disk device and a syndrome where the CRC code occurs are used. Then, the syndrome of the data after the error correction is regenerated, and the data is rechecked from this (for example, JP-A-4-355279).

[0003]

In the above-described conventional magnetic disk control device, if an ECC error does not occur, C
Since error correction of RC code is impossible, CRC
When a data that causes a code error and a CRC code are written in advance and the data is read, an ECC error is not detected and only a CRC code error is detected. In this case, although it is natural that the error is not recovered even if the retry (retry) is performed, the retry process is performed a predetermined number of times, so that there is a disadvantage that the system throughput is reduced. .

An object of the present invention is to consciously write data that causes a CRC code error in a magnetic disk drive,
When the data is read, it is determined whether or not the data is a CRC code error generated artificially (it is a true CRC code error). It is another object of the present invention to provide a magnetic disk control device that can only report to the system and prevent a decrease in system throughput.

[0005]

SUMMARY OF THE INVENTION According to the present invention, when data is not sent due to a failure of a higher-level device when writing data to a magnetic disk device, dummy data is transferred and a CRC code added to the transferred data is inverted. In a magnetic disk control device having a function of preventing incomplete writing to be transmitted, a first CR for directly checking data and a CRC code when reading data from the magnetic disk device.
A C code check circuit, a second CRC code check circuit for checking the data by inverting the CRC code while keeping the data as it is, and a check result of the first CRC code check circuit and the second CRC code check circuit. When detecting a data error, an error discriminating circuit for discriminating whether the data error is a true data error or a data error due to the incomplete writing countermeasure is provided.

The error discriminating circuit transfers the data read from the magnetic disk device to a host device regardless of the result of the second CRC code checking circuit when there is no data error in the first CRC code checking circuit. The first CRC code check circuit determines that there is a data error and the second CR
If there is no data error in the C code check circuit, the data is determined to be data written by the countermeasure against incomplete writing, and the first CRC code check circuit and the second CRC code are used.
When both the code check circuits detect a data error, it is determined that there is a data error due to some failure.

[0007]

Next, the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram showing an embodiment of the present invention. Referring to FIG. 1, a magnetic disk controller 10 of the present invention is connected to a host device (not shown) via a write data line 31, a read data line 32, an inoperable display line 33, and a transfer result report line 34. The write data line 35 and the read data line 36
Is connected to the magnetic disk drive 20 via the.

The magnetic disk controller 10 includes an incomplete write countermeasure instruction circuit 11 for instructing an incomplete write when a signal sent from the host device via the inoperable display line 33 is active. A dummy data generation circuit 14 for generating dummy data according to an instruction from the countermeasure instruction circuit 11, and a CRC code added to the end of a signal transmitted from a higher-level device via the write data line 31 or data output from the dummy data generation circuit 14. The CRC code generation circuit 12 and the data normally sent from the CRC code generation circuit 12 and the CRC code are directly written to the magnetic disk drive 20 via the write data line 35.
The CRC code is inverted only when instructed by the write countermeasure instructing circuit 11 and the magnetic disk device 20
A CRC code inverting circuit 13, a dummy data generating circuit 14 for generating dummy data according to an instruction from the incomplete writing countermeasure instruction circuit 11, and a read data line 36.
Code check circuit A for checking the CRC code and the data sent from the magnetic disk device 20 via the
15, a CRC code check circuit B16 for inverting the CRC code sent from the magnetic disk device 20 via the read data line 36 to check data,
An error discriminating circuit 17 is provided for discriminating the presence or absence of a data error or the reading of dummy data based on the check results by the C code check circuit A15 and the CRC code check circuit B16.

Next, the processing operation of the magnetic disk control device of the present invention configured as described above will be described.

First, a CRC code generation circuit 12 performs a CRC code operation on data to be written to the magnetic disk device 20, adds the data to the end of the data, and transfers the data to the magnetic disk device via a CRC code inversion circuit 13. . And
The incomplete writing countermeasure instructing circuit 11 constantly monitors the inoperable display line 33 which is an interface signal with the host device, and when the signal of the inoperable display line 33 becomes active during writing to the magnetic disk device 20, It instructs the CRC sign inversion circuit 13 and the dummy data generation circuit 14 to take measures against incomplete writing.

Here, the countermeasure against incomplete writing means that when data cannot be sent out due to some kind of failure of the host device during data writing to the magnetic disk device 20.
Dummy data is to be transferred for the remainder of the predetermined number of data transfers. In that case, CR
By inverting the C code, a data error is detected because the data is not normal when read later.

The CRC sign inverting circuit 13 normally has
The data to which the CRC code generated by the CRC code generation circuit 12 is added is transferred to the magnetic disk device 20 via the write data line 35 as it is. Then, the CRC code generated by the CRC code generation circuit 12 is inverted and transferred to the magnetic disk device 20 as data with an inverted CRC code. When the incomplete write countermeasure instruction circuit 11 gives an instruction for the incomplete write countermeasure, the dummy data generation circuit 14 generates the dummy data to generate the CRC code generation circuit 12 and the CRC code inversion circuit 13.
Is transferred to the magnetic disk device 20 via the.

Next, the CRC code check circuit A15
When reading data from the magnetic disk device 20 via the read data line 36, the data read from the magnetic disk device 20 and the CRC code are checked, and the check result is reported to the error determination circuit 17. on the other hand,
When the CRC code check circuit B16 reads data from the magnetic disk device 20 via the read data line 36, the data from the magnetic disk device 20 and C
A check is performed by inverting the RC code, and the check result is reported to the error determination circuit 17.

Here, the error determination circuit 17 comprises a CRC code check circuit A15 and a CRC code check circuit B
An error is determined based on the 16 check results. If there is no data error in the CRC code check circuit A15, the data read from the magnetic disk device 20 by the CRC code check circuit A15 is read out from the magnetic disk device 20 via the read data line 32 regardless of the result of the CRC code check circuit B16. And reports the normal end to the higher-level device via the transfer result report line 34.

The error discriminating circuit 17 reads out data previously written by the countermeasure against incomplete writing when there is a data error in the CRC code checking circuit A15 and when there is no data error in the CRC code checking circuit B16. Is reported to the host device via the transfer result report line 34.

Further, the error determination circuit 17 comprises a CRC code check circuit A15 and a CRC code check circuit B16.
If both are detected as a data error, retry processing is performed because the data error is due to some kind of failure, and if recovery is not performed after performing the specified number of retries, the transfer result report line 34 indicates that a data error has been detected. To the host device via

[0018]

As described above, the magnetic disk control device of the present invention, when detecting a data error when reading data from the magnetic disk device, consciously detects the error.
Is it a data error caused by inverting the RC code?
Because you can determine whether it is due to a real data error,
There is an effect that it is not necessary to perform extra rerun processing, and a decrease in system throughput can be prevented.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an embodiment of the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 10 Magnetic disk controller 11 Incomplete write countermeasure instruction circuit 12 CRC code generation circuit 13 CRC code inversion circuit 14 Dummy data generation circuit 15 CRC code check circuit A 16 CRC code check circuit B 17 Error discrimination circuit 20 Magnetic disk devices 31, 35 Write data line 32, 36 Read data line 33 Inoperable display line 34 Transfer result report line

Claims (2)

[Claims] 1. When data is not sent due to a failure of a host device when writing data to a magnetic disk device,
Dummy data transfer and CRC added to transfer data
In a magnetic disk controller having an incomplete writing countermeasure function of inverting and transmitting a code, a first CRC code check circuit for directly checking data and a CRC code when reading data from the magnetic disk device, A second CRC code check circuit for inverting and checking the CRC code;
When a data error is detected from the check results of the CRC code check circuit and the second CRC code check circuit, it is determined whether the data error is a true data error or a data error due to the incomplete writing countermeasure. A magnetic disk control device comprising: an error determination circuit. 2. The method according to claim 1, wherein the error determining circuit includes a first CR.
If there is no data error in the C code check circuit, the data read from the magnetic disk device is transferred to the host device regardless of the result of the second CRC code check circuit, and it is determined that the data has been normally terminated. If there is a data error in the code check circuit and no data error in the second CRC code check circuit, it is determined that the data has been written by the countermeasures against incomplete writing, and the first
2. The magnetic disk control device according to claim 1, wherein when both the CRC code check circuit and the second CRC code check circuit detect a data error, it is determined that there is a data error due to some failure.