JPH0954742A - Disk device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To shorten the time needed for reading the initial data in a power-on start mode or reset start mode of a disk device. SOLUTION: The disk device is provided with a disk medium 7 which has a system using area and a control means 1 which performs the read/write operations to the medium 7 by the instructions of a host 8. The means 1 stores command and address issued by the host 8 in the system using area in a power-on start mode or a reset start mode. In such a constitution, the command and address are read out of the system using area and a seeking operation, etc., is previously carried out in a power-on or reset start mode of the disk device. Thus, the response to the host can be made faster.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a host such as a personal computer including a portable information terminal, a mobile communication device, a workstation, various OA devices (facsimile, printer, financial terminal, etc.) and a disk array device for some mainframes. The present invention relates to a disk device connected to a computer (hereinafter referred to as "host").

In the above-mentioned system, the operating system (OS) used in each system is almost fixed, and it is often used only in a few kinds of specific OS. The disk device is widely used as a main device of these auxiliary storage devices for system storage. However, if a function specialized for a specific OS is installed, it is expected to improve the performance of the startup time at system startup (time reduction). it can.

[0003]

2. Description of the Related Art Conventionally, a disk device has a function of performing self-diagnosis and self-adjustment at power-on startup and reset startup, but seek / positioning is performed after a read command is issued from the host for data. And data read.

FIG. 10 is a processing flowchart of a conventional example. FIG. 10 shows the processing and interrupt processing routine at the time of power-on (reset) of the hard disk device connected to the host by SCSI (small computer system interface).

(A) Explanation of Power-on Startup and Reset Startup S121: At the time of power-on startup and reset startup of the hard disk device, the processor performs self-diagnosis of hardware such as RAM (random access memory), and processing S122. Move on to.

S122: The processor makes initial adjustments such as clearing RAM and setting initial values of registers of the control unit,
The process moves to step S123. S123: The processor enables the SCSI interrupt, that is, enables the SCSI protocol processing with the host and the static command processing that can be processed without activating the spindle motor, and proceeds to processing S124.

S124: The processor activates the spindle motor and proceeds to processing S125. S125: The processor performs drive mechanism initial adjustment, such as head calibration, and performs processing S12.
Go to 6.

S126: The processor downloads the system file for storing the control program in the RAM from the disk medium, and proceeds to the processing S127. S127: The drive is ready (idle state), and the seek-type command processing of reading / writing data by positioning the head on the disk medium, which is dynamic processing, becomes possible.

(B) Description of Interrupt Processing Routine After the SCSI interrupt is enabled in step S123,
When there is an interrupt from the host, the SCSI bus used in the arbitration phase (ARBITRATION PHASE) between the host and the hard disk drive is acquired according to the SCSI protocol, and the selection phase (SE
The SCSI bus is selected in LECTION PHASE), the information transfer method is determined in the message phase (MESSAGE PHASE), and the command is transferred in the command phase (COMMAND PHASE).

S131: The processor of the hard disk device analyzes the command received from the host, and proceeds to processing S132. S132: The processor determines whether the analyzed command is a static command. If it is a static command by this judgment, execute that command,
If it is not a static command, the process proceeds to step S133.

S133: The processor determines whether or not the drive is ready. If it is determined that the drive is ready, the command is executed. If the drive is not ready, the command is notified to the host that the drive cannot be used (Not Ready).

[0012]

SUMMARY OF THE INVENTION The above-mentioned conventional device has the following problems. The hard disk drive is
Seek, positioning, and data read are performed to a specific boot sector every time for checking the memory when the hard disk drive is started, and the system startup time is long and the total system performance is deteriorated.

It is an object of the present invention to solve such a conventional problem and to shorten the time required for initial data read at the time of power-on startup or reset startup of the disk device.

[0014]

FIG. 1 is a diagram for explaining the principle of the present invention. In FIG. 1, 1 is a control means, 2a is a head, 7 is a disk medium, 8 is a host, 11a is a control unit, and 13a is The memory 15a is a non-volatile memory.

The present invention is configured as follows to solve the above-mentioned conventional problems. A disk medium 7 having a system use area and a control means 1 for reading / writing data from / to the disk medium 7 according to an instruction from the host 8 are provided, and the control means 1 is provided at the time of power-on start-up or reset start-up. The command and address issued from the host 8 are stored in the system use area. As a result, at the time of power-on start-up or reset start-up of the disk device, the command and address are read from the system use area and a seek operation or the like is performed in advance, so that the host 8
The response to can be quick.

Further, a control means 1 for reading / writing data from / to the disk medium 7 according to an instruction from the host 8,
The control means 1 is provided with a non-volatile memory 15a, and the control means 1 stores the command and address issued from the host 8 in the non-volatile memory 15a at power-on start-up or reset start-up. As a result, at the time of power-on start-up or reset start-up of the disk device, commands and addresses are read from the non-volatile memory 15a that does not require mechanical operation and seek operation or the like is performed in advance, so that the response to the host 8 can be made faster. can do.

Further, a disk medium 7 having a system use area and a control means 1 for reading / writing data from / to the disk medium 7 according to an instruction from the host 8 are provided, and the control means 1 logs commands. When a command is received from the host 8, the command and address issued by the host 8 are stored in the system use area.
As a result, when the same operation is always performed after a certain operation, the command and address are read from the system use area in advance, stored in memory, etc., and the read address is sought to speed up the response to the host. It can be carried out. Further, by collecting frequently used commands and address statistics and storing them in the system use area, it is possible to easily take countermeasures against failures by checking frequently used addresses.

Further, a control means 1 for reading / writing data from / to the disk medium 7 according to an instruction from the host 8,
The control means 1 is provided with a non-volatile memory 15a, and the control means 1 sends a command for logging commands to the host 8
When received from, the command and address issued from the host 8 are stored in the non-volatile memory 15a. As a result, when the same operation is always performed after a certain operation, the command and address are read from the non-volatile memory 15a in advance and stored in the memory 13a or the like, and the read address is sought or the like to respond to the host. Can be done faster. By storing frequently used commands and address statistics in the non-volatile memory 15a and checking frequently used addresses, it is possible to easily take measures against a failure, and at the same time, save the disk device. Even when the mechanical unit does not move, information can be obtained from the non-volatile memory without the mechanical unit.

Further, there is provided a disk medium 7 in which commands and addresses are stored in the system use area, and a control means 1 having a memory 13a for reading / writing data from / to the disk medium 7 in response to an instruction from the host 8. The control means 1 buffers the read data in the memory 13a in advance at the power-on start-up or the reset start-up according to the command and the address stored in the system use area. As a result, the read data used at the power-on start-up and the reset start-up can be read from the memory 13a, so that the response to the host 8 can be speeded up.

Further, a control means 1 for reading / writing data from / to the disk medium 7 according to an instruction from the host 8,
The control means 1 comprises a memory 13a and a non-volatile memory 15a storing commands and addresses.
Stores the read data in the memory 13a in advance at the power-on start-up or the reset start-up according to the command and address stored in the nonvolatile memory 15a. As a result, according to the command and address read from the non-volatile memory without the mechanical unit,
Since the read data used at power-on startup and reset startup can be buffered in the memory 13a, the response to the host 8 can be made faster.

Further, there is provided a disk medium 7 in which commands and addresses are stored in a system use area, and a control means 1 having a memory 13a for reading / writing data from / to the disk medium 7 according to an instruction from a host 8. The control means 1 buffers the read data in the memory 13a according to the command and address stored in the system use area when receiving the command for buffering the data read according to the logged command and address from the host. As a result, read data that is frequently accessed can be buffered in the memory 13a in advance and the response to the host 8 can be made faster even when the power-on or the reset is not started.

Further, a control means 1 for reading / writing data from / to the disk medium 7 according to an instruction from the host 8,
The control means 1 comprises a memory 13a and a non-volatile memory 15a storing commands and addresses.
When receiving a command for buffering read data according to the logged command and address from the host 8, the read data is buffered in the memory 13a according to the command and address stored in the nonvolatile memory 15a. As a result, read data with high access frequency read according to a command and an address from the non-volatile memory 15a having no mechanism unit is buffered in the memory 13a even when the power is turned on or the reset is not started. The response to can be faster.

Further, a disk medium 7 storing read data to be used at power-on startup and reset startup in a system use area, and a memory 13a for reading / writing data from / to the disk medium 7 according to an instruction from the host 8 are provided. And a control means 1 having the
At power-on startup or reset startup, the read data stored in the system use area is buffered in advance in the memory 13a. As a result, at the time of power-on activation or reset activation, the data to be used each time is checked in the memory 13a and reported to the host 8, so that the response to the hit data to the host 8 can be speeded up.

Also, a control means 1 for reading / writing data from / to the disk medium 7 according to an instruction from the host 8,
The control means 1 is provided with a memory 13a and a non-volatile memory 15a storing read data used at the time of power-on start-up and reset start-up, and the control means 1 has the non-volatile memory 15 at power-on start-up or reset start-up.
The read data stored in a is stored in advance in the memory 1
Buffer in 3a. As a result, the data to be used each time at power-on startup or reset startup is buffered in advance from the non-volatile memory 15a having no mechanism unit to the memory 13a, so that the response to the hit data to the host 8 can be made faster. can do.

Further, there is provided a disk medium 7 in which read data is stored in the system use area, and a control means 1 having a memory 13a for reading / writing data from / to the disk medium 7 in response to an instruction from the host 8. The means 1 buffers the read data stored in the system use area in the memory 13a when receiving a command for logging read data from the host. As a result, read data that is frequently accessed can be buffered in the memory 13a in advance and the response to the host 8 can be made faster even when the power-on or the reset is not started.

Further, a control means 1 for reading / writing data from / to the disk medium 7 according to an instruction from the host 8,
The control means 1 comprises a memory 13a and a non-volatile memory 15a storing read data, and the control means 1 receives the read data stored in the non-volatile memory 15a when a command for logging read data is received from the host 8. The data is buffered in the memory 13a. As a result, read data having a high access frequency read from the non-volatile memory 15a having no mechanical unit is buffered in the memory 13a even when the power-on or the reset is not started, so that the response to the host 8 is faster. can do.

[0027]

2 to 9 are views showing an embodiment of the present invention, and an embodiment of the present invention will be described below with reference to the drawings.

1): Description of Disk Device FIG. 2 is an overall block diagram of the disk device in the embodiment. The disk device includes a control means 1, heads 2, 3,
4, 5, a spindle motor (SPM) 6 and a disk medium 7 are provided.

The control means 1 performs a seek for positioning the heads 2 to 5 on a designated track in accordance with a data read / write instruction from the host, and reads / writes to the disk medium 7. The spindle motor 6 is controlled by the control means 1 and drives the disk medium 7.

2): Description of control means FIG. 3 is a circuit block diagram of the control means in the embodiment. The control means 1 includes a disk control circuit 10 and an MPU.
A processor 11, such as a (microprocessor unit),
ROM (read only memory) 12, data buffer RAM (random access memory) 13, program memory RAM 14, non-volatile memory EEPR
OM (Electrically Erasable Programmable Read Only)
A memory) 15, a writing / reading circuit 16, and a head IC circuit 17 are provided.

The disk control circuit 10 includes a processor 1
1, the RAM 13, the write / read circuit 16 and the host 8 are connected by SCSI. The disk control circuit 10 controls the write / read circuit 16 in accordance with a data read / write instruction from the host. To do.

The processor 11 is a disk control circuit 1
0, ROM 12, RAM 14, and EEPROM 15 are connected to control the entire control means 1. The ROM 12 is a read-only memory that stores firmware for starting up this device.

The RAM 13 is a memory for buffering data. The RAM 14 is a program memory that stores a control program and the like. EEPRO
M15 is a non-volatile memory in which information can be written and erased by electrical means (memory in which information is not erased even when power is off).

The write / read circuit 16 is provided between the disk control circuit 10 and the head IC circuit 17, and performs read / write processing (demodulation, modulation, etc.) of data to the disk medium. The head IC circuit 17 performs read / write data amplification processing between the write / read circuit 16 and the heads 2-5.

In the above embodiment, the RAM 13 and RA
The case where M14 is provided separately has been described, but one RAM
May be used. In addition, the EEPROM 15
Can also use other non-volatile memory.

3): Description of command and LBA (logical block address) logging at power-on start and reset start LBA (logical) which is a command and address issued from the host at power-on start and reset start・
When using the function of storing (block address), the CLOG FLAG (for example, this control bit is allocated to the RAM 14) which is a flag for logging (history) commands in the conventional sequence (see FIG. 10) is turned on, Furthermore, a function for allocating a storage buffer area for commands and LBAs to the RAM 13 or RAM 14 is added. This CLOG FLAG is a status flag indicating that this function is turned on.

At the time of command processing, after confirming whether the CLOG FLAG is on or not, if the CLOG FLAG is on, the command and the LBA are stored. The command and LBA are stored by executing the SCSI command (= command descriptor block (CDB)) issued by the host in a command storage buffer area other than the data buffer of the RAM 13.

Therefore, prior to the storage of commands and LBAs, the work of allocating storage buffer areas for commands and LBAs is performed in advance. In the SCSI command interrupt processing routine where the command is issued from the host,
If CLOG FLAG is set, it recognizes that this function is used and starts storing the command and LBA in the buffer area. When the capacity of the buffer area becomes full, the contents are written into the EEPROM 15 which is a system use area or a non-volatile memory provided on the inner circumference side of the disk medium 7, for example.

FIG. 4 is a flowchart of a command and LBA logging process at startup in the embodiment. The process at the time of power-on startup and reset startup of the disk device and the SCSI command interrupt processing routine will be described below with reference to FIG. The process enclosed by a square in the process flowchart indicates a new process in the embodiment (the same applies to the following embodiments).

(A) Explanation of Power-on Startup and Reset Startup S1: At power-on startup and reset startup of the disk device, the processor 11 performs self-diagnosis of the hardware such as the RAMs 13 and 14, and shifts to processing S2. .

S2: Processor 11 is RAM 13, 1
Initial adjustment such as clearing of No. 4 and setting of the initial value of the register is performed, and the process proceeds to step S3. S3: The processor 11 turns on the CLOG FLAG, and proceeds to processing S4.

S4: The processor 11 allocates a command and LBA storage buffer area to the RAM 13 or RAM 14, and proceeds to processing S5. S5: The processor 11 enables the SCSI interrupt, that is, enables the SCSI protocol processing with the host 8 and the static command processing that can be performed without activating the spindle motor 6, and proceeds to processing S6.

S6: The processor 11 activates the spindle motor 6 and proceeds to processing S7. S7: The processor 11 performs drive mechanism initial adjustment such as calibration of the heads 2 to 5, and then proceeds to processing S8.

S8: The processor 11 uses the disk medium 7
To download a system file for storing the control program in the RAM 14 and move to processing S9. S9: Drive ready (idle state) is achieved, and seek-related command processing for reading / writing data by positioning the heads 2-5 on the disk medium 7 which is dynamic processing becomes possible.

(B) Description of Interrupt Processing Routine If an interrupt is received from the host 8 after the SCSI interrupt is enabled in the process S5, the SCSI used in the arbitration phase between the host 8 and the disk device according to the SCSI protocol. The bus is acquired and the SCSI bus is selected in the selection phase, the information transfer method is determined in the message phase, the command is transferred in the command phase, and the process proceeds to step S11.

S11: The processor 11 uses the CLOG F
It is determined whether the LAG is on (standing). If CLOG FLAG is set in this determination, the process proceeds to step S12, and if CLOG FLAG is not set,
The process moves to step S13.

S12: The processor 11 sends the command and L
The storage of BA in the buffer area is started, and the process proceeds to step S13. S13: The processor 11 determines whether or not the capacity of the buffer area for storing the command and the LBA has become full. When the capacity of the buffer area for storing the command and the LBA is full in this determination, the process proceeds to step S14. If the buffer area is not full, the process proceeds to step S15.

S14: The processor 11 sends the command and L
The contents of BA are the system use area (for example, provided on the inner circumference side) of the disk medium 7 or EEP which is a nonvolatile memory
The data is written in the ROM 15, and the process proceeds to step S15.

S15: The processor 11 analyzes the command received from the host 8 and moves to processing S16. S16: The processor 11 determines whether the analyzed command is a static command. If the command is a static command in this determination, the command is executed. If it is not a static command, the process proceeds to step S17.

S17: The processor 11 determines whether or not the drive is ready. If it is determined that the drive is ready, the command is executed. If the drive is not ready, the host 8 is notified that the drive cannot be used (Not Ready).

In this way, at the time of power-on start-up and reset start-up, a command for seeking to a specific boot sector and the LBA are stored each time. Therefore, in particular, at the time of power-up of the notebook computer, startup of the disk array system using the fault tolerance function and the hot standby function at the time of startup of the disk array system, etc., the command stored before the command and the LBA come from the host. And LBA can be used for seeking and positioning, and the response to the host 8 can be speeded up.

4): Description of using command logging command A command (COMMAND LOGGING COMMAND) is provided for executing the command history function not only at power-on startup and reset startup, but also at normal idle. In this case, CLOG B which is a bit for logging the command and LBA in the CDB which is a SCSI command
Create IT and issue a COMMAND LOGGING command.

Then, when the drive (disk device) receives and recognizes this command, first, CLOG BI
Check T, and if it is on, CLOG FLAG will be set up,
Allocates a command and LBA storage buffer area. C
If the LOG BIT is off, the CLOG FLAG is dropped to cancel the allocation of the command and LBA storage buffer areas. This allows COMMAND LOGGI
The NG command has a function of canceling the command history function.

FIG. 5 shows the COMMAND L in the embodiment.
It is an OGGING command processing routine. In S21: COMMAND PHASE, when the command is received from the host 8, the processor 11 analyzes the command and moves to the process S22.

S22: The processor 11 determines whether the received command is a COMMAND LOGGING command. With this judgment COMMAND LOGGIN
If the command is a G command, the process proceeds to step S23, and if C
If the command is not the OMMAND LOGGING command, other command processing is performed.

S23: The processor 11 determines whether the CLOG BIT of the command is on (ON). If CLOG BIT is turned on in this determination, the process S
24, and if CLOG BIT is not on, the process proceeds to step S26.

S24: The processor 11 uses the CLOG F
The LAG is turned on, and the process proceeds to step S25. S25: The processor 11 allocates a command and LBA storage buffer area (RAM 13 or 14).

S26: The processor 11 executes the CLOG F
The LAG is turned off (OFF), and the process proceeds to S27. S27: The processor 11 cancels the allocation of the command and LBA storage buffer areas.

In this way, if the same operation is to be performed after a certain operation, the response to the host can be made faster by seeking to the previously logged address (for example, when going to the same database). . Also,
By taking statistics of frequently used commands and addresses (LBA) and storing them in the system use area or in non-volatile memory, the frequently used addresses can be checked to take measures against failures (from the most frequently used addresses Check) can be facilitated. Further, when the information is stored in the non-volatile memory, the information can be obtained from the non-volatile memory having no mechanism even if the mechanism of the disk device does not move.

5): Description of holding initial data At power-on and at reset, data read is performed according to the read command and address stored in the system use area or the non-volatile memory, and the read data is stored in the data buffer. Then, the LOG FLAG provided in the RAM 13 or 14 is set up.

FIG. 6 is a processing flowchart for holding the initial data. The process and SCS at the time of power-on startup and reset startup of the disk device will be described below with reference to FIG.
The I command interrupt processing routine will be described.

(A) Explanation of Power-on Startup and Reset Startup S31: At power-on startup and reset startup of the disk device, the processor 11 performs self-diagnosis of hardware such as the RAMs 13 and 14, and shifts to processing S32. .

S32: The processor 11 is the RAM 13,
Initial adjustment such as clearing of 14 and setting of initial values of registers is performed, and the process proceeds to step S33. S33: The processor 11 reads the data according to the read command and the address stored in the system use area or the non-volatile memory at the time of power-on and reset, and is a bit that is a LOAD BIT (RAM 13 or 14).
(Allocated to) is turned on, and the process proceeds to step S34.

S34: The processor 11 is the RAM 13
A buffer storage area for read data is allocated to, and the process proceeds to step S35. S35: The processor 11 loads the read command and address stored in the system use area or the non-volatile memory, and moves to the processing S36.

S36: The processor 11 reads the data from the loaded address and moves to the processing S37. S37: The processor 11 stores the read read data in the buffer storage area allocated in the process S34, and proceeds to the process S38.

S38: The processor 11 indicates L indicating that the read data is stored in the buffer storage area.
The OG FLAG is turned on, and the process proceeds to step S39. S39: The processor 11 enables the SCSI interrupt,
That is, the SCSI protocol process and the static command process with the host 8 are enabled, and the process proceeds to step S40.

S40: The processor 11 activates the spindle motor 6 and moves to the processing S41. S41: The processor 11 performs drive mechanism initial adjustment such as calibration of the heads 2 to 5, and then proceeds to processing S42.

S42: The processor 11 downloads the system file for storing the control program in the RAM 14 from the disk medium 7, and proceeds to the processing S43. S43: The drive ready (idle state) is set, and the seek-related command processing for reading / writing data by positioning the heads 2-5 on the disk medium 7 which is dynamic processing becomes possible.

(B) Description of Interrupt Processing Routine When an interrupt is received from the host 8 after the SCSI interrupt is permitted in the processing S39, the SCSI used in the arbitration phase between the host 8 and the disk device according to the SCSI protocol. The bus is selected and the SCSI bus is selected in the selection phase, the information transfer method is determined in the message phase, the command is transferred in the command phase, and the process proceeds to step S51.

S51: The processor 11 uses the CLOG F
It is determined whether the LAG is on (standing). If the CLOG FLAG is on in this determination, the process proceeds to step S52. If the CLOG FLAG is not on, the process proceeds to step S53.

S52: The processor 11 sends the command and L
The BA is stored in the system use area of the disk medium or the EEPROM 15 which is a non-volatile memory, and the process proceeds to step S53. S53: The processor 11 analyzes the command interrupted by the host 8 and moves to the processing S54.

S54: The processor 11 sends the LOG FL
Determine if AG is on. LOG F
If LAG is on, the process proceeds to step S55, and if LOG
If FLAG is not on, normal command processing is performed.

S55: The processor 11 sends the command and L
BA determines if there is a hit. If the command and LBA are hit in this determination, the process proceeds to step S56. If the command and LBA are not hit, normal command processing is performed.

S56: The processor 11 reads the hit data from the buffer storage area, and the process S57
Move on to. S57: The processor 11 transfers the read data to the host 8.

As described above, since the read data is stored in the buffer storage area in advance based on the storage of the command and the LBA issued at the power-on start and the reset start, the response to the host 8 can be speeded up. .

6): Description of using load logging data command A command (LOAD LOGGING DATA command) for executing the command history function not only at power-on startup and reset startup but also at normal idle Set up. In this case, LOAD is set in CDB which is a SCSI command.
Create a BIT and issue a LOAD LOGGING DATA command.

When the drive (disk device) receives and recognizes this command, first, LOAD BI
Check T, if it is on, set LOAD FLAG,
Allocates the read data buffer storage area. LOA
If D BIT is off, the LOAD FLAG is dropped to cancel the allocation of the read data buffer storage area. By this, this LOAD LOGGING DA
The TA command has a function of canceling the command history function.

FIG. 7 shows LOAD LOGG in the embodiment.
This is an ING DATA command processing routine. In S61: COMMAND PHASE, when the command is received from the host 8, the processor 11 analyzes the command and moves to the process S62.

S62: The processor 11 determines whether the received command is a LOAD LOGGING DATA command. This judgment is LOAD LOGGING
If the command is a DATA command, the process proceeds to step S63. If it is not a LOAD LOGGING DATA command, another command process is performed.

S63: The processor 11 determines whether the LOAD BIT of the command is ON. If LOAD BIT is on in this determination, the process S
The procedure moves to 64, and if the LOAD BIT is not on, the procedure moves to the process S69.

S64: Processor 11 loads LOAD F
The LAG is turned on, and the process proceeds to step S65. S65: The processor 11 allocates a read data storage buffer area (RAM 13) and moves to the processing S66.

S66: The processor 11 reads the read command and the address (LBA) stored in the EEPROM 15 which is the system use area or the non-volatile memory, and moves to the processing S67.

S67: The processor 11 reads the data from the read address and moves to the processing S68. S68: The processor 11 stores the read read data in the buffer storage area allocated in the processing S65. If this buffer storage area becomes full, then data will be transferred to the system use area or EEPROM.
It can be stored in M15.

S69: Processor 11 loads LOAD F
The LAG is turned off (OFF), and the process proceeds to step S70. S70: The processor 11 cancels the allocation of the read data buffer storage area.

As described above, by storing the frequently accessed data in the buffer in advance, it is possible to speed up the response to the host. 7): Description of logging initial data At power-on and reset, a bit newly provided in the RAM 13, DLOG BIT, is set to ON to set this function mode. Next, the read data buffer storage area is allocated. This buffer storage area is
This area is different from the data buffer used for normal commands and is an area dedicated to this function.

After that, when the normal power-on sequence is executed and the download of the control firmware is completed, the disk medium 7 is automatically sent before the drive ready is issued.
System use area or EEP which is non-volatile memory
Initial read data is loaded from the ROM 15 and stored in a buffer, and DLOG FLAG is set.

This DLOG FLAG is a status flag indicating the buffering status of the initial read data.
When this flag is on, it indicates that the data is buffered.

Command issued from host 8, SC
In the SI command interrupt processing routine, the command type and the value of DLOG BIT are checked, and the read data is stored in the buffer when it is a read command and when DLOG BIT is on.

When the read data to be stored becomes buffer full, this data is saved (written) in the system use area or the non-volatile memory. FIG. 8 is a processing flowchart for logging initial data in the embodiment. The processing at the time of power-on startup and reset startup of the disk device and the SCSI command interrupt processing routine will be described below with reference to FIG.

(A) Explanation of Power-on Startup and Reset Startup S81: At the time of power-on startup and reset startup of the disk device, the processor 11 performs self-diagnosis of the hardware such as the RAMs 13 and 14, and shifts to processing S82. .

S82: The processor 11 is the RAM 13,
Initial adjustment such as clearing of 14 and setting of initial values of registers is performed, and the process proceeds to step S83. S83: The processor 11 turns on the DLOG BIT, and proceeds to processing S84.

S84: The processor 11 is the RAM 13
To the buffer storage area for the initial read data, and the process proceeds to step S85. S85: The processor 11 enables the SCSI interrupt,
That is, the SCSI protocol processing with the host 8 and the static command processing that can be performed without activating the spindle motor 6 are enabled, and the process proceeds to step S86.

S86: The processor 11 activates the spindle motor 6 and moves to the processing S87. S87: The processor 11 performs drive mechanism initial adjustment such as calibration of the heads 2 to 5, and then proceeds to processing S88.

S88: The processor 11 downloads a system file for storing the control firmware in the RAM 14 from the disk medium 7, and proceeds to the processing S89. S89: The processor 11 loads the initial read data from the system use area of the disk medium 7 or the buffer storage area allocated from the EEPROM 15 and allocated in the processing S84, and proceeds to processing S90.

S90: The processor 11 turns on the DLOG FLAG assigned to the RAM 13, and the process S
Move to 91. S91: The processor 11 is set to drive ready (idle state), and it is possible to perform seek-type command processing, which is dynamic processing, in which the heads 2 to 5 are positioned on the disk medium 7 to read / write data.

(B) Description of Interrupt Processing Routine If an interrupt is received from the host 8 after the SCSI interrupt is enabled in step S85, the SCSI used in the arbitration phase between the host 8 and the disk device according to the SCSI protocol. The bus is acquired, the SCSI bus is selected in the selection phase, the information transfer method is determined in the message phase, and the command is transferred in the command phase. Then, the process proceeds to step S101.

S101: The processor 11 executes the DLOG
Determine if BIT is on (standing). If the DLOG BIT is on in this determination, the process proceeds to step S102. If the DLOG BIT is not on, the process proceeds to step S104.

S102: The processor 11 determines whether the received command is a read command. If it is a read command in this determination, the process proceeds to step S103, and if
If it is not a read command, other commands are processed.

S103: The processor 11 stores the read data read by the read command in the buffer storage area, and proceeds to the processing S104. S104: The processor 11 determines whether or not the capacity of the buffer storage area for storing the read data has become full. If the capacity of the buffer storage area is full according to this determination, the process proceeds to step S105. If the capacity of the buffer storage region is not full, the process proceeds to step S106.

S105: The processor 11 writes the read data into the system use area of the disk medium 7 or the EEPROM 15 which is a non-volatile memory, and the process S1.
Move to 06.

S106: The processor 11 analyzes the command received from the host 8 and moves to processing S107. S107: The processor 11 determines whether the analyzed command is a static command. If it is determined that the command is a static command, the command is executed. If it is not a static command, the process proceeds to step S108.

S108: The processor 11 determines whether or not the drive is ready. If it is determined that the drive is ready, the command is executed. If the drive is not ready, the host 8 is notified that the drive cannot be used (Not Ready).

As described above, since the data to be used each time is loaded in advance at the power-on start-up and the reset start-up, the response to the host 8 can be speeded up. 6): Explanation of using data logging command A command (DATA LOGGING command) is provided to execute the command history function not only at power-on startup and reset startup, but also at normal idle. In this case, in the CDB which is a SCSI command, DLOAD
Create a BIT and issue a DATA LOGGING command.

When the drive receives and recognizes this command, it first confirms DLOADBIT, and if ON, allocates a read data storage buffer area. Then, the read data is loaded from the system use area of the disk medium 7 or the EEPROM 15 which is a non-volatile memory and stored in the buffer, and the DLOG FLA is stored.
Set G.

DLO BIT is off, DLO
AD FLAG is dropped to cancel the allocation of the read data buffer storage area. Therefore, this DATA
The LOGGING command has a function of canceling the command history function.

FIG. 9 shows DATA LOGG in the embodiment.
It is an ING command processing routine. In S111: COMMAND PHASE, when the command is received from the host 8, the processor 11 analyzes the command and moves to the processing S112.

S112: The processor 11 determines whether the received command is a DATA LOGGING command. If it is the DATA LOGGING command in this determination, the process proceeds to step S113, and if the DATA
If it is not a LOGGING command, other command processing is performed.

S113: The processor 11 determines whether or not the command DLOAD BIT is ON. If the DLOAD BIT is on in this determination, the process proceeds to step S114, and if the DLOAD BIT is not on, the process proceeds to step S118.

S114: The processor 11 allocates the read data buffer storage area (RAM 13) and moves to the processing S115. S115: The processor 11 uses the system use area or E
The data stored in the EPROM 15 is read out and the process S is performed.
Move to 116.

S116: The processor 11 stores the read read data in the buffer storage area allocated in the processing S114, and shifts to the processing S117. S117: The processor 11 indicates DLOAD indicating that the read data is stored in the buffer storage area.
Turn on FLAG.

S118: The processor 11 executes DLOAD
The FLAG is turned off (OFF), and the process proceeds to step S119. S119: The processor 11 cancels the allocation of the read data buffer storage area.

As in this embodiment, by storing frequently accessed data in the buffer, it is possible to speed up the response to the host. As described above, it is possible to shorten the time required to read specific data each time such as at power-on startup and reset startup, and improve the command response. With this function, especially laptop computers that use a lot of power save, in which the contents of RAM are retained,
This is effective for a disk array system that uses a fault tolerance function and a hot standby function, a system that installs a specified OS for a specific OEM host at the time of factory shipment, an internal hard disk device for a specific system, and the like.

[0113]

As described above, the present invention has the following effects. (1) The control means stores the command and address issued from the host in the system use area at power-on start-up or reset start-up. Therefore, the control means stores the command and address at power-on start-up or reset start-up. A seek operation or the like can be performed in advance by reading from the system use area, and the response to the host can be speeded up.

(2) Since the control means stores the command and address issued from the host in the non-volatile memory at the time of power-on start-up or reset start-up, the command and address are stored at the time of power-on start-up or reset start-up of the disk device. The address can be read in advance from the non-volatile memory that does not require a mechanical operation and the seek operation can be performed in advance, so that the response to the host can be made faster.

(3) The control means stores the command and address issued by the host in the system use area when the command for logging the command is received from the host. By reading commands and addresses from the system use area in advance and storing them in memory, etc., and seeking to those addresses, the response to the host can be made faster, and frequently used commands and addresses can be collected to obtain a system. By storing the address in the used area, it is possible to easily take countermeasures in case of failure by checking the frequently used addresses.

(4) Since the control means stores the command and address issued from the host in the non-volatile memory when the command for logging the command is received from the host, if the same operation is always performed after a certain operation, the command is executed. Also, the address can be read out from the non-volatile memory in advance, stored in the memory or the like, and the address can be sought to speed up the response to the host. In addition, by collecting frequently used commands and address statistics and storing them in a non-volatile memory, it is possible to easily take countermeasures against failures by checking frequently used addresses, and also the mechanism of the disk device. Even when the unit does not move, information can be obtained from the non-volatile memory having no mechanical unit.

(5) The control means buffers the read data in the memory in advance at the power-on start-up or the reset start-up according to the command and the address stored in the system use area. Since the read data to be used can be reported from the memory at the time of startup, the response to the host 8 can be speeded up.

(6) In accordance with the command and address stored in the non-volatile memory, the control means buffers the read data in the memory in advance at power-on startup or reset startup. At start-up, the read data to be used can be buffered in the memory according to the command and address from the non-volatile memory having no mechanism unit in advance, and the response to the host can be made faster.

(7) The control means buffers the read data in the memory according to the command and address stored in the system use area when receiving the command for buffering the data read according to the logged command and address from the host. Therefore, read data that is frequently accessed is buffered in advance in memory even when the power is turned on or the reset is not started.
The response to the host can be made faster.

(8) The control means buffers the read data in the memory according to the command and address stored in the non-volatile memory when receiving the command for buffering the data read according to the logged command and address from the host. Therefore, read data that is frequently accessed according to commands and addresses from a non-volatile memory without a mechanism unit can be buffered in the memory even when the power is turned on or the reset is not started, and the response to the host can be made faster. can do.

(9) The control means buffers the read data stored in the system use area in the memory in advance at power-on start-up or reset start-up, and therefore, at each power-on start-up or reset start-up. Since the data used can be checked in the memory and reported to the host, the response to the hit data to the host can be made faster.

(10) The control means buffers the read data stored in the non-volatile memory in the memory in advance at power-on startup or reset startup. The data to be used can be buffered in the memory from the non-volatile memory without the mechanical unit in advance, and the response to the hit data to the host can be made faster.

(11) Since the control means buffers the read data stored in the system use area in the memory when the command for logging the read data is received from the host, the control means does not perform the power-on start-up or the reset start-up. However, read data that is frequently accessed can be buffered in the memory in advance to speed up the response to the host.

(12) Since the control means buffers the read data stored in the non-volatile memory in the memory when receiving the command for logging the read data from the host, the control means does not perform the power-on startup or the reset startup. However, read data that is frequently accessed and read from a non-volatile memory without a mechanical unit can be buffered in the memory and the response to the host can be made faster.

[Brief description of drawings]

FIG. 1 is a diagram illustrating the principle of the present invention.

FIG. 2 is an overall block diagram of a disk device according to an embodiment.

FIG. 3 is a circuit block diagram of control means in the embodiment.

FIG. 4 is a flowchart of a command and LBA logging process at startup in the embodiment.

FIG. 5: COMMAND LOGGIN in the embodiment
It is a G command processing routine.

FIG. 6 is a processing flowchart for holding initial data in the embodiment.

FIG. 7: LOAD LOGGING D in Example
It is an ATA command processing routine.

FIG. 8 is a processing flowchart for logging initial data in the embodiment.

FIG. 9 is a DATA LOGGING command processing routine in the embodiment.

FIG. 10 is a processing flowchart of a conventional example.

[Explanation of symbols]

 1 control means 2a head 7 disk medium 8 host 11a control unit 13a memory 15a non-volatile memory

Claims (12)

[Claims] 1. A disk medium having a system use area, and a control means for reading / writing data from / to the disk medium according to an instruction from a host, wherein the control means is configured to perform a power-on startup or a reset startup. A disk device characterized in that a command and an address issued from a host are stored in the system use area. 2. A control means for reading / writing data from / to a disk medium according to an instruction from the host, and a nonvolatile memory in the control means, wherein the control means is operated by the host at power-on startup or reset startup. A disk device characterized in that the issued command and address are stored in the non-volatile memory. 3. A disk medium having a system use area and a control means for reading / writing data from / to the disk medium according to an instruction from a host, wherein the control means receives a command for logging a command from the host. A disk device, wherein a command and an address issued by a host are stored in the system use area at times. 4. A control means for reading / writing data from / to a disk medium according to an instruction from the host, and a nonvolatile memory in the control means, wherein the control means receives a command logging command from the host. A disk device characterized in that a command and an address issued from a host are stored in the non-volatile memory. 5. A disk medium in which a command and an address are stored in a system use area, and a control unit having a memory for reading / writing data from / to the disk medium according to an instruction from a host, the control unit comprising: A disk device, wherein read data is buffered in advance in the memory at power-on startup or reset startup in accordance with a command and an address stored in the system use area. 6. A control means for reading / writing data from / to a disk medium according to an instruction from a host, and the control means includes a memory and a non-volatile memory storing a command and an address. Disk device, wherein read data is buffered in advance in the memory at power-on startup or reset startup in accordance with a command and an address stored in the memory. 7. A disk medium in which a command and an address are stored in a system use area, and a control unit having a memory for reading / writing data from / to the disk medium according to an instruction from a host, the control unit comprising: A disk device, wherein when a command for buffering read data according to a logged command and address is received from a host, read data is buffered in the memory according to the command and address stored in the system use area. 8. A control means for reading / writing data from / to a disk medium according to an instruction from a host, and a non-volatile memory storing a memory and a command and an address in the control means, wherein the control means performs logging. A disk device, wherein when a command for buffering read data according to a command and an address is received from a host, the read data is buffered in the memory according to the command and the address stored in the nonvolatile memory. 9. A control means having a disk medium storing read data to be used at power-on start-up and reset start-up in a system use area, and a memory for reading / writing data on the disk medium according to an instruction from a host. The disk device, wherein the control means buffers the read data stored in the system use area in the memory in advance at power-on startup or reset startup. 10. A control means for reading / writing data from / to a disk medium according to an instruction from a host, and the control means includes a memory and a non-volatile memory storing read data used at power-on startup and reset startup. The disk device, wherein the control means buffers the read data stored in the non-volatile memory in the memory in advance at power-on startup or reset startup. 11. A disk medium having read data stored in a system use area, and a control unit having a memory for reading / writing data from / to the disk medium in response to an instruction from a host, the control unit being a read unit. A disk device, wherein when the command for logging data is received from a host, the read data stored in the system use area is buffered in the memory. 12. A control means for reading / writing data from / to a disk medium according to an instruction from a host; and a non-volatile memory storing the memory and the read data in the control means, wherein the control means stores the read data. A disk device, wherein when the command for logging is received from a host, the read data stored in the non-volatile memory is buffered in the memory.