JPS63124124A - Method for reading out data from disk device - Google Patents

Abstract

PURPOSE:To ensure a data reading method with which a fixed reading time is always assured by adding a process where data are read out at and after a sector emerged first and transferred and a process where data are read out at and after the first sector and transferred. CONSTITUTION:A disk controller 2 starts reading with the data on the n-th sector that emerged first after a data read instruction is delivered and the read data equivalent to a following track are transferred to a data buffer in the first process. In the second process the reading job is started with the data equivalent to the first sector in the data buffer and the buffer data corresponding to a following track are transferred to outside. In such a reading method, a fixed processing time is maintained with maximum efficiency in a disk access mode and furthermore it is not required at all to change the processing method of software at the host processor side.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a data reading method for a disk device, and particularly to a method for reading data from a disk device.
The present invention relates to a data reading method for a disk device having a data buffer for tracks.

[Prior Art] A conventional method for reading data in this type of disk device is to detect an index mark and then start reading data. FIG. 4(a) is a timing chart showing a conventional data read operation, and FIG. 4(b) is a diagram showing a conventional track buffer access mode. Therefore, as shown in FIG. 4(a), for example, if the first sector that appears after timing ① when a read command is generated on a rotating magnetic disk is not the first sector, the read operation will not be performed. The motor was in a waiting state until the index input signal was detected after one rotation. Therefore, for example, one rotation takes approximately 17ms.
When using a hard disk drive, the read time in the worst case is approximately 17 ms of rotational waiting time + 17 m5 of data read time + 3 ms of data transfer time to the host device, for a total of 37 m5. On the other hand, in the ideal case, that is, when the first sector immediately after the index mark is first detected and read, the rotation waiting time is approximately Oms + the data read time is 17ms + the data transfer time is 3m = 20ms, In the very worst case, 37
This requires 1.9 times the reading time (÷20), which has the drawback of significantly lowering the efficiency of the system.

[Problems to be Solved by the Invention] The present invention eliminates the above-mentioned drawbacks of the prior art,
The purpose is to propose a data read method for a disk device that can always guarantee a constant data read time.

[Means for Solving the Problems] In order to achieve the above object, the data read method for a disk device of the present invention starts reading data from the n-th sector that appears first after a data read command is generated. a first step of transferring one track's worth of read data to the data buffer; and a first step of starting reading from data corresponding to the first sector in the data buffer and transferring the following one track's worth of buffer data to an external device. A second step of transferring the data to

[Operation] In the first step, when the occurrence of a data read command from the host device is detected, reading starts from the data in the n-th sector that appears first, and the read data for the following one track is read out. Transfer to buffer. In the second step, reading is started from data corresponding to the first sector in the data buffer, and the following one track's worth of buffer data is transferred to the outside.

[Description of Embodiments] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

FIG. 1 is a block diagram of a disk device to which an embodiment is applied. In the figure, 1 is an external host processor that accesses the disk device of the embodiment, 2 and 3 constitute the disk device of the embodiment, 2 is a disk controller, and 3 is a magnetic recording medium. Further, in the disk controller 2, 2-1 performs data read/write and head operations, for example, according to the flowchart shown in FIG.
It is a control unit that performs seek, etc., and 2-2 is a track buffer (TB) that stores one track worth of data including the first sector to the Nth sector.

Furthermore, the control unit 2-1 has a memory address register (M
AR) 21 and a sector counter (S) that counts the number of sectors of data to be transferred for the track buffer 2-2.
C) 22, and 1 is a signal line for exchanging commands and data between the host processor 1 and the disk controller 2, and 2 is a signal line for exchanging commands and data between the control unit 2-1 and the track buffer 2-2. A signal line for exchanging signals, dan 3 is a magnetic recording medium 3 and a disk controller 2
This is a signal line that exchanges data reading and writing signals between the two.

In such a configuration, when an access command is issued by the host processor 1, the control unit 2-1 analyzes the command on the signal edge 1, and, for example, when executing a read command, reads 3-statement 3→ from the magnetic recording medium 3. The read data is transferred through a path 2-2 and stored in the track buffer 2-2. Next, the buffer data is 2-2 → sentence 2 → 2-1 → intersection 1
→ Transferred to host processor 1 via path 1.

FIG. 2 is a flowchart showing the data read process of the embodiment, FIG. 3(a) is a timing chart showing the data read operation of the embodiment, and FIG. 3(b) is a diagram showing the track buffer access mode of the embodiment. It is. In FIG. 2, when a data succession command is received from the host processor 1, it is input to step s1. In step S2, MAR is set to address 0, and in step S3, SC is set to 0. Step S4 waits for the first sector mark to appear. The sector mark may be one that is actually read from the recording medium 3, or may be electrically generated by dividing the period of the index signal. When a sector mark is detected, the process proceeds to step S5, and the sector number n is held in the pointer register (PR). In step S6, the read data of sector number n is transferred to TB2-2. At the same time, MAR21 is updated (+1) every time one byte is transferred. When data transfer for one sector is completed, the contents of 5C22 are incremented by 1 in step S7.

In step S8, it is determined whether the content of 5C22 is N or not. N
If not, the process advances to step S9 and waits for the next sector mark.

When a sector mark is detected, the process proceeds to step S6, where data for the next one sector is transferred.When 5C=N is determined in step S8, it means that data for one track has been transferred. That is, TB2-2 stores one track worth of read data in the sequence shown in FIG. 3(b). This causes the flow to proceed to step S10, where the MAR
21 is set to the content (N-(PR)+1). Here, in this example, since the content of PR (PR) is n, MAR21 ultimately points to data corresponding to the first sector in TB2-2. In process Sll, SC is set to O.
Make it. In step S12, buffer data corresponding to sector number 1 of TB2-2 is transferred to host processor 1.

At the same time, MAR21 is updated every time 1 byte is transferred (+
1), and the contents of MAR21 operate to return to address O when it exceeds the maximum address of TB2-2. When data transfer for one sector is completed, step S1
3 adds +1 to the contents of 5C22. In step 514, 5C2
It is determined whether the content of 2 is N or not. If not N, step 31
2, the next one sector worth of data is transferred, and when 5C-N is determined in step 314, the process proceeds to step 515.
Return. In this way, data for one track has been transferred to the host processor 1.

According to FIG. 3(a), a data read command is generated at timing 2 in the middle of sector 2 after index signal timing 2. In this case, data reading can be started from sector 3, so the data storage time in TB2-2 ends at sector 2 after the next index signal timing 3, and data can then be transferred to host processor 1. Comparing this with the case of FIG. 4(a), it can be seen that the waiting time for approximately one rotation is omitted.

Further, according to the process shown in FIG. 2 described above, read data is transferred to TB2-2 in the manner shown in FIG. 3(b). That is, in the first step, starting from the first address of TB2-2, read data for one track is stored,
In the second step, starting from the address corresponding to the (N-n+1)th sector of TB2-2, one track's worth of buffer data is output to the outside.

On the other hand, by changing the control of MAR21 etc., T
B2-2 reading control may also be used.

That is, in the first step, read data for one track is stored starting from the address corresponding to the n-th sector of TB2-2, and in the second step, read data for one track is stored starting from the first address of TB2-2. It outputs buffered data for tracks to the outside.

In the description of the embodiment, data read control has been described, but data write control can also be easily performed using the same idea.

[Effects of the Invention] As explained above, according to the present invention, reading is started from the data in the n-th sector that appears first after a data read command is generated, and the read data for one subsequent track is stored in the data buffer. Since data is transferred, there is no waiting time, and the processing time during disk access can be maintained at a constant and maximum efficiency.

Furthermore, data is transferred from the disk device to the track buffer starting from arbitrary sector n, but data is always transferred from sector 1 to the host processor, so there is no need to change the software processing method on the host processor side. There is.

[Brief explanation of the drawing]

1 is a block configuration diagram of a disk device to which the embodiment is applied; FIG. 2 is a flowchart showing data read processing in the embodiment; FIG. 3(a) is a timing chart showing data read operation in the embodiment; FIG. 4(b) is a diagram showing the access mode of the track buffer of the embodiment. FIG. 4(a) is a timing chart showing the conventional data read operation. FIG. 4(b) is a diagram showing the access mode of the conventional track buffer. It is a diagram. In the figure, 1... host processor, 2... disk controller, 3... magnetic recording medium, 2-1... control section, 2-2... track buffer.

Claims (3)

[Claims] (1) In a data reading method for a disk device equipped with a data buffer for one track including sectors 1 to N, data is read from the data in the n-th sector that appears first after a data read command is generated. a first step of starting and transferring one track's worth of read data to the data buffer; and starting reading from data corresponding to the first sector in the data buffer and continuing one track's worth of buffered data. 1. A method for reading data from a disk device, comprising a second step of transferring the data to the outside. (2) A patent claim characterized in that in the first step, the transfer is started from the first address of the data buffer, and in the second step, the transfer is started from the address corresponding to the N-n+1th sector of the data buffer. A method for reading data from a disk device according to item 1. (3) In the first step, the transfer is started from an address corresponding to the n-th sector of the data buffer, and in the second step, the transfer is started from the first address of the data buffer. 2. A method for reading data from a disk device according to item 1.