JP2003330798A - Storage device, recording and/or reproduction device, information storage system, storage method, program and recording medium - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To improve the cache responsiveness by an efficient flash processing. <P>SOLUTION: According to the attribute of information stored in a cache memory (temporary storage means), the operation of reading information temporarily stored in the temporary storage means and writing it in the storage medium, that is, flash operation is controlled. Particularly, flash processing is started preferentially from data having a shorter flash processing time, or in order of continuous data, noncontinuous data, and blocking necessary data. Many data-writable empty spaces are formed in the cache memory in its early stages as much as possible by such a control to improve the cache responsiveness rate. <P>COPYRIGHT: (C)2004,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a storage device, a recording and / or reproducing device having the storage device, an information storage system including the storage device, a storage method, and the storage device or the storage method. The present invention relates to a program to be realized and a recording medium storing the program.

[0002]

2. Description of the Related Art The development of data storage devices that use media such as optical disks and magneto-optical disks has been diversified, and the data intended for data storage has also been diversified. Various storage devices have been proposed. For example, in a device using a medium such as an optical disk and a magneto-optical disk, it is used as a so-called AV (audio / visual) device for storing (recording / reproducing) music data and video data, and as a storage device as a personal computer peripheral device. A device for storing various data (text data, audio data, video data, various program files, application data, etc.) used in a personal computer has been put into practical use.

In recent years, a disk medium compatible with both AV equipment and personal computer peripheral equipment has been developed. As a recording / reproducing apparatus, for example, it functions as an AV machine by itself and is connected to a personal computer or a network. Therefore, a device that can function as a storage device similar to a PC peripheral device has also been developed.

For the sake of explanation, data having temporal continuity such as music contents and video contents is called "AV data", and various data such as control data, program data and text data handled by a network or a personal computer is called. It is called "PC data".

[0005]

By the way, when considering an AV device that can be connected to a network or a personal computer, versatility and high speed are required. Conventionally, A
With V storage devices, there is no problem unless recording / playback of AV data that is temporally continuous is interrupted, and there has been a demand for power consumption and other convenience rather than for higher speeds. However, when there is a demand for a method of connecting a device to a network and downloading data, while reading and writing various settings such as PC data, AV data can be transmitted several tens to several thousand times faster. It requires high speed and versatility such as reading and writing.

Considering a general data path, the data recorded / reproduced on / from a recording medium such as a disc is
In many cases, the cache memory is temporarily passed. About this cache memory, PC
The purpose of use was different between data and AV data.

The general idea of a cache in a storage device for PC data is that data read from a disk or the like in the past is held in the cache, and if a data read request from the disk occurs, the cache is stored. By hitting the above data, the read / write of the disk in the storage device becomes unnecessary, and the role of speeding up the response to the request is great. And when there is a cache hit, the response is fast, but when there is no cache hit, it is much slower than when there is a cache hit, so increasing the probability of a cache hit is important for improving the responsiveness of the device. is there. For this purpose, a hash algorithm that minimizes collisions and speeds up search by randomly distributing the data to be secured in the cache as ideally as possible is used. This hash algorithm, as a result, promotes fragmentation of the data in the cache memory, but the searchability is good because the address in the cache can be specified by the predetermined algorithm for the requested data.

On the other hand, the general idea of the cache in the recording / reproducing device for AV data is to prefetch as much as possible and store the data to absorb the delay of the processing in the storage device.
It plays a large role as a buffer for saving power by turning off the power source of the motor, etc. AV continuous in time series
Most of the cache is used as a read-ahead area because the data has a very high read-ahead validity. Further, since AV data has a larger data size and higher continuity than PC data, a ring buffer method is often adopted in order to keep a cache area. In the case of such AV data use, the hash algorithm in which the data fragmentation progresses on the cache as described above is not optimal.

Further, in particular, considering a storage device using a disk as a recording medium, if a plurality of read / write processes can be made to have consecutive addresses as much as possible and can be processed in one read / write operation, the speed and power consumption are reduced. Can be achieved. In other words, if tracks can be recorded and reproduced by continuous scanning (light beam tracing), for example, on a spirally formed track on a disk, frequent track jumps, access, disk rotation waiting time, etc. can be eliminated. Because you can. However, the fragmentation of data on the cache by the above hash algorithm makes it difficult to record / reproduce with continuous addresses in this way, and therefore, in the case of use for recording / reproducing AV data on a disc, The hash algorithm is not optimal.

As described above, due to the difference in the purpose of caching PC data and AV data, when a storage device that supports both PC data and AV data is considered, it is difficult to optimize cache processing. That is, it has been difficult to realize high-speed response and low power consumption by efficiently reading and writing PC data while realizing a high-speed recording / reproducing function of AV data.

Considering that the data temporarily stored in the cache memory is to be written to the disk, efficient processing can be performed due to a write operation (for example, access) to the disk and an operation convenience such as a write data unit. Although it may not be possible, it is difficult to realize efficient processing, especially when considering both AV data and PC data.

[0012]

In view of the above, the present invention takes into consideration a storage device adapted to a network / personal computer and a device that also functions as an AV device, and has a high-speed AV data recording / reproducing function and PC data efficiency. It is an object of the present invention to provide a cache control method that ensures high speed response, low power consumption, and low cost, and also makes the operation of writing cache accumulated data into a recording medium such as a disk efficient.

The storage device of the present invention discriminates the temporary storage means for temporarily storing the information read from the recording medium and / or the information to be written to the recording medium, and the attribute of the information temporarily stored in the temporary storage means. An attribute discriminating unit and a control unit for controlling the information stored in the temporary storage unit to be read out and written in the recording medium according to the discriminating result of the attribute discriminating unit. Also,
The attribute discriminating means is configured such that, regarding the information temporarily stored in the temporary storage means, continuous information continuously written in the recording medium, and discontinuously written in the recording medium discontinuously as compared with the continuous information. The control means controls to read the continuous information preferentially and write it to the recording medium. Further, the attribute determining means determines, for the information temporarily stored in the temporary storage means, blocking necessary information that requires blocking when writing to the recording medium, and the control means determines information other than the blocking necessary information. Is read out and written to the recording medium. In this case, the control means writes the necessary blocking information in the recording medium after performing a blocking process of forming a predetermined write data unit using the information read from the recording medium. To control. Further, the attribute determination means determines the hit status of the information temporarily stored in the temporary storage means,
The control means controls so that the information indicating the hit status having a low value is preferentially read and written in the recording medium.

The recording and / or reproducing apparatus of the present invention includes a writing / reading means for writing and / or reading information to / from a recording medium, and information and / or information read from the recording medium by the writing / reading means. And / or the temporary storage means for temporarily storing the information written to the recording medium by the writing / reading means, the attribute determination means for determining the attribute of the information temporarily stored in the temporary storage means, and the attribute determination means. And a control unit that controls the information temporarily stored in the temporary storage unit to be read and written in the recording medium according to the determination result.

An information storage system according to the present invention, which comprises, for example, a storage device and an information processing device, is a temporary storage means for temporarily storing information read from a recording medium and / or information to be written to the recording medium, Attribute discriminating means for discriminating the attribute of information temporarily stored in the temporary storage means;
Control means for controlling the information stored in the temporary storage means to be read and written in the recording medium in accordance with the result of the determination by the attribute determining means, and for writing and / or reading the information in the recording medium. To this end, a writing / reading requesting means for notifying the control means of the address indicating the access position to the recording medium is provided.

According to the storage method of the present invention, when the information read from the recording medium and / or the information to be written to the recording medium is temporarily stored in the temporary storage means, the attribute of the information temporarily stored in the temporary storage means is determined. Then, the information temporarily stored in the temporary storage means is read out and written in the recording medium according to the result of the discrimination of the attribute.

The program of the present invention, when the information read from the recording medium and / or the information to be written to the recording medium is temporarily stored in the temporary storage means, determines the attribute of the information temporarily stored in the temporary storage means. The program is a program for executing a process of controlling to read the information temporarily stored in the temporary storage unit and write the information in the recording medium according to the result of the determination of the attribute.

According to the recording medium of the present invention, when the information read from the recording medium and / or the information to be written to the recording medium is temporarily stored in the temporary storage means, the attribute of the information temporarily stored in the temporary storage means is determined. The recording medium stores a program for executing a process of controlling the information stored in the temporary storage unit to be read out and written in the recording medium according to the result of the determination of the attribute.

That is, according to the present invention, the operation (flash operation) of reading the information temporarily stored in the temporary storage means and writing the information in the recording medium is controlled according to the attribute of the information stored in the cache memory (temporary storage means). To do. Particularly, the flash processing is preferentially performed from the data such as continuous data having a short flash processing time. In addition, the flash process is preferentially performed from the data having a low hit state value. As a result, it becomes possible to increase the free area in which data can be written in the cache memory (temporary storage means) at the earliest possible stage.

[0020]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below. The description will be given in the following order. 1. 1. Configuration of recording / reproducing apparatus Configuration of disk and storage unit 3. Basic cache control method 4. Write / read operation 5. Cache search process 6. Flash processing example <1> 7. Flash processing example <2>

1. Configuration of Recording / Reproducing Apparatus The recording / reproducing apparatus as an embodiment is, for example, a recording / reproducing apparatus for a mini disk (MD) type disk which is a magneto-optical disk on which data is recorded by a magnetic field modulation method. However, the recording / reproducing apparatus is not limited to the mini-discs that are already in widespread use, but is also applicable to a high-density disc that enables higher-density recording and can be used for storage of various data for computer use.

The configuration of the recording / reproducing apparatus of this example will be described with reference to FIG. In FIG. 1, a recording / reproducing apparatus 1 of this example is, for example, a personal computer (or network) 1
00 is shown as a device capable of data communication with an external device. For example, the recording / reproducing apparatus 1 includes a personal computer 100 and a transmission line 101 such as a USB cable.
By connecting with the personal computer 100, the personal computer 100 can function as an external storage device. Also, music or various data is downloaded via the personal computer 100 or by being connected to the network by having a function of directly connecting to the network, and the music and various data are downloaded to a disc loaded in the storage unit 2 in the recording / reproducing apparatus 1. It can also be saved.

On the other hand, the recording / reproducing apparatus 1 functions as an audio device, for example, without being connected to the personal computer 100 or the like. For example, music data input from another audio device or the like can be recorded on the disc, and music data recorded on the disc can be reproduced and output. That is, the recording / reproducing apparatus 1 of this example can be used as a general-purpose data storage device by being connected to the personal computer 100 or the like, and can also be used as an audio recording / reproducing device by itself.

The recording / reproducing apparatus 1 includes a storage unit 2, a cache memory 3, a USB interface 4, an input / output processing unit 5, a display unit 6, an operation unit 7, a system controller 8, a ROM 9, a RAM 10, a cache management memory 1.
1, NV-RAM 12 is provided.

The storage unit 2 performs recording / reproduction on the loaded disc. The so-called mini disk type disk used in this example and the configuration of the storage unit 2 corresponding thereto will be described later.

The cache memory 3 is a cache memory for buffering the data recorded on the disk by the storage unit 2 or the data read from the disk by the storage unit 2. For example D-R
It is composed of AM2. Writing / reading of data to / from the cache memory is controlled by a task activated in the system controller (CPU) 8.

The USB interface 4 performs processing for data transmission when it is connected to the personal computer 100 via the transmission path 101 as a USB cable, for example. The input / output processing unit 5 performs processing for inputting / outputting recording / reproducing data when the recording / reproducing apparatus 1 functions as an audio device by itself, for example.

The system controller 8 controls the entire recording / reproducing apparatus 1 and controls communication with the connected personal computer 100. ROM9
The operation program of the system controller 8, fixed parameters, and the like are stored in. The RAM 10 is used as a work area by the system controller 8 and also as an area for storing various necessary information. For example, the storage unit 2 stores various management information and special information read from the disk. For example, P-TOC data, U-TOC data, playlist data, FAT data, Unique I
D, hash value, etc. are stored. P-TOC data, U-
The TOC data is management information such as music tracks recorded on the mini disc. Further, as will be described later, the high-density disc conforming to the mini-disc system, which the recording / reproducing apparatus 1 of the present example can support, is a management format by P-TOC and U-TOC, and a FAT file system is constructed. The playlist is A for high-density discs.
This is information for managing addresses and the like of music data in the TRAC system and is recorded as one file on the FAT system. When a high-density disc is loaded, the FAT and play list information is also read. The unique ID, hash value, and the like are information used for authentication processing and encryption / decryption during data transmission with the personal computer 100 or the like.

The cache management memory 11 is, for example, S-
It is composed of a RAM and stores information for managing the state of the cache memory 3. The system controller 8 controls the data cache process while referring to the cache management memory 11. The information in the cache management memory 11 will be described later. NV-RAM (nonvolatile RAM) 12
Is used as a data storage area that is not lost even when the power is turned off.

The display section 6 displays various information to be presented to the user under the control of the system controller 8. For example, operation status, mode status, name information of data such as music, track number, time information, and other information are displayed. Operation keys, a jog dial, and the like are formed on the operation unit 7 as various operators for user operations. The user operates the operation unit 7 to instruct necessary operations for recording / reproduction and data communication. The system controller 8 performs a predetermined control process based on the operation information input by the operation unit 7.

The control by the system controller 8 when the personal computer 100 or the like is connected is as follows, for example. The system controller 8 is communicable with the personal computer 100 connected via the USB interface 4, and receives commands such as a write request and a read request and transmits status information and other necessary information. The system controller 8 reads the management information and the like from the disk when the disk is loaded in the storage section 2, for example.
To the RA and load it via the cache memory 3
Store in M10. By reading the management information of P-TOC and U-TOC, the system controller 8 can grasp the track recording state of the disc. Further, by reading the CAT, the high density data cluster structure in the data track can be grasped, and the personal computer 1
The access request from 00 to the data track is ready. Further, the unique ID or the hash value can be used to perform disk authentication or other processing, or to send these values to the personal computer 100 for processing.

When there is a request for reading certain data from the personal computer 100, the system controller 8 causes the storage section 2 to execute the reading of the data. The read data is written in the cache memory 3. However, if the requested data has already been stored in the cache memory 3, the storage unit 2
Is not required to be read. This is a so-called cache hit. Then, the system controller 8 causes the data written in the cache memory 3 to be read out, and the personal computer 100 is accessed via the USB interface 4.
Control to send to.

When there is a request for writing certain data from the personal computer 100, the system controller 8 causes the cache memory 3 to store the transmitted data. Then, the storage unit 2 records the data stored in the cache memory 3 on the disk. It should be noted that data recording on the disc is performed with a cluster unit as a minimum unit. For example, the cluster is 32 FAT sectors. If the amount of data requested to be recorded by the personal computer 100 or the like is several sectors and the amount is less than one cluster, a process called blocking is performed. That is, the system controller 8 first causes the storage unit 2 to read the cluster including the FAT sector. The read cluster data is written in the cache memory 3. The system controller 8 is the personal computer 10
FAT sector data (record data) from 0 to US
The data is supplied to the cache memory 3 via the B interface 4, and the stored cluster data is rewritten to the data of the corresponding FAT sector. Then, the system controller 8 transfers the cluster data stored in the cache memory 3 in a state where the necessary FAT sector is rewritten, to the storage unit 2 as recording data. The storage unit 2 writes the cluster unit data to the disk.

The above is the control for recording / reproducing data accompanying the transmission with the personal computer 100, for example, the data transfer at the time of recording / reproduction of audio data of the mini disk system is performed by the input / output processing unit 5. Done through. The input / output processing unit 5 includes, for example, an analog audio signal input unit such as a line input circuit / microphone input circuit, an A / D converter, and a digital audio data input unit as an input system. It also comprises an ATRAC compression encoder / decoder. Further, as an output system, a digital audio data output section and an analog audio signal output section such as a D / A converter and a line output circuit / headphone output circuit are provided.

Audio data is recorded on the disc as a process via the input / output processing unit 5, for example, the input T
This is a case where digital audio data (or analog audio signal) is input to the input / output processing unit 5 as IN. The input linear PCM digital audio data or the linear PCM audio data input by the analog audio signal and converted by the A / D converter is ATRA.
It is C-compressed and encoded and stored in the cache memory 3. Then, it is read from the cache memory 3 and transferred to the storage unit 2 at a predetermined timing (a data unit corresponding to an ADIP cluster). The storage unit 2 modulates the transferred compressed data by a predetermined modulation method and records it on the disk.

When the mini-disc audio data is reproduced from the disc, the storage unit 2 demodulates the reproduced data into an ATRAC compressed data state and transfers it to the cache memory 3. Then, it is read from the cache memory 3 and transferred to the input / output processing unit 5. The input / output processing unit 5 sets A for the supplied compressed audio data.
TRAC compression decoding is performed to obtain linear PCM audio data, which is output from the digital audio data output unit. Alternatively, the D / A converter performs line output / headphone output as an analog audio signal.

The configuration of the recording / reproducing apparatus 1 of FIG. 1 is an example, and for example, the input / output processing unit 5 may be provided with an input / output processing system for not only audio data but also video data. . Further, the connection with the personal computer 100 is not limited to USB, and other external interface such as IEEE 1394 may be used.

2. Configuration of Disk and Storage Unit The disk used as a recording medium in the recording / reproducing apparatus 1 of the present embodiment is, for example, a mini disk type disk. In particular, it supports not only conventional mini discs for music, but also high-density discs that can record various data for computer use.

FIG. 2 shows a comparison between the standards of the audio mini disc (and MD-DATA) and the high density disc. As a format of the mini disk (and MD-DATA), the track pitch is 1.6 μm and the bit length is 0.59 μm / bit. Further, the laser wavelength λ = 780 nm, and the numerical aperture NA of the optical head =
It is set to 0.45. As a recording method, a groove recording method is adopted. That is, the groove (groove on the disc surface) is used as a track for recording and reproduction. As an address method, a method is used in which a groove (track) is formed by a single spiral and then a wobbled groove in which wobbles are formed as address information on both sides of this groove is used. In this specification, the absolute address recorded by wobbling is referred to as ADIP (Address in Pregroov
Also called e).

The recording data modulation method is EFM (8
-14 conversion) method is adopted. As an error correction method, ACIRC (Advanced Cross Interleave Reed-
Solomon Code) is adopted, and convolutional type is adopted for data interleaving. Data redundancy is 46.3
%.

The data detection method is a bit-by-bit method. CLV (Const
ant linear velocity) is adopted, and the linear velocity of CLV is 1.2 m / s. The standard data rate during recording and reproduction is 133 kB / s, and the recording capacity is 164 MB (140 MB for MD-DATA). Also, the data unit called cluster is the minimum rewriting unit of data.
It consists of 36 main sectors with 32 main sectors and 4 link sectors.

On the other hand, the high density disc has a track pitch of 1.6 μm and a bit length of 0.44 μm / bit. The laser wavelength λ = 780 nm, the aperture ratio NA of the optical head = 0.45, the recording method is a groove recording method, and the address method is a groove (track) formed by a single spiral. A method of using a wobbled groove in which a wobble as information is formed is adopted. In other words, these points are the same as those of the audio mini-disc, and the configuration of the optical system in the disc drive device, the ADIP address reading method,
Since the servo processing is the same, compatibility is maintained.

As a recording data modulation method, the RLL (1,7) PP method (RL
L; Run Length Limited, PP: Parity preserve / Proh
ibitrmtr (repeated minimum transition runlength))
RS-L with BIS (Burst Indicator Subcode), which has a higher correction capability, is adopted as the error correction method.
The DC (Reed Solomon-Long Distance Code) method is used. A block complete type is adopted for data interleaving. The data redundancy is set to 20.50%. The data detection method is the partial response PR (1,
2, 1) Viterbi decoding method using ML. R
Regarding the LL (1-7) modulation and the RS-LDC error correction method, for example, "Japanese Patent Laid-Open No. 11-346154" is used.
Alternatively, the technology is disclosed in International Patent Publication WO 00/07300.

The disk drive system is CLV (Constant
Linear Verocity) or ZCAV (Zone Constant Angu)
lar Verocity) and its standard linear velocity is 2.7 m /
s, and the data rate at the standard linear velocity during recording and reproduction is 4.9 Mbit / s. As a recording capacity, 297 MB can be obtained. Modulation method is E
By adopting the FM to RLL (1,7) PP method, the window margin is changed from 0.5 to 0.666, and in this respect, the density can be increased by 1.33 times. In addition, from the CIRC system as the physical format, RS- with BIS
The data efficiency is 53.7% to 7 due to the difference between the LDC method and the sector structure and the method using Viterbi decoding.
This is 9.5%, and a 1.48 times higher density can be realized in this respect. Overall, 1.97 times (about 2 times)
The data capacity of is realized. In other words, recording capacity is 297M
It can be double the size of a mini disc for audio called B. The minimum data rewriting unit is 64 KB.

The area structure on the disc is schematically shown in FIG. As shown in FIG. 3, the innermost side of the disc is PT
It is an OC (pre-mastered TOC) area, which is a pre-mastered area as a physical structure. That is,
This is an area in which reproduction-only data is recorded by embossed pits, and P is the management information as the reproduction-only data.
-TOC is recorded.

The outer periphery of the premastered area is a recordable area (a magneto-optically recordable area), and is a recordable / reproducible area in which a groove as a guide groove for a recording track is formed. The innermost side of the recordable area is the U-TOC area. In the U-TOC area, a buffer area with the pre-mastered area and a power calibration area used for adjusting the output power of the laser light are provided, and in the section of specific 3 clusters in the U-TOC area. U-TOC data is recorded repeatedly three times. Although details of the contents of the U-TOC are omitted, the address of each track recorded in the program area, the address of the free area, etc. are recorded, and information such as the track name and recording date and time attached to each track is recorded. The U-TOC sector is defined so that it can be done.

In the recordable area, an alert track and a data area are formed on the outer peripheral side of the U-TOC area. The alert track is a warning track in which a warning sound indicating that this disc is a high density disc and cannot be reproduced by a conventional mini disc player is recorded. The data area is an area used for recording audio data and computer use data. That is, in the case of a normal music mini disk,
One or more music tracks are recorded in the data area. In the case of a high density disc, the data area is used as a storage area for various data.

The management structure of the disk will be described with reference to FIG. FIG. 4B shows a management structure in the music mini disk system. In this case, the management information is P
-TOC and U-TOC are recorded. P-TOC is recorded by pits as non-rewritable information. This P
-TOC includes basic management information of the disc, such as the total capacity of the disc, the U-TOC position in the U-TOC area, the position of the power calibration area, the start position of the data area, and the end position of the data area (readout). (Position) etc. are recorded.

On the other hand, U recorded in the recordable area
-TOC is management information that is rewritten according to recording and erasing of tracks (audio tracks / data tracks), and each track (parts that make up the track)
Manage start position, end position, and mode. It also manages a free area in which no track is recorded in the data area, that is, a part as a writable area. For example, in the data area as shown,
When three audio tracks are recorded as ATRAC data, the U-TOC manages the start address, end address, mode information, recording date and name information of each of the three tracks.

In the case of a high density disc, it becomes as shown in FIG. Even in the case of high density disc, U-TOC and P-
The TOC is recorded by a method that complies with the conventional mini disk system. And in the data area of FIG. 4 (a),
One high density data track is shown, U-TO
In C, the high-density data track is managed as one track. In other words, the U-TOC manages the position of one or more parts of the data track as a whole on the disc. The position of the alert track described above is also managed by the U-TOC.

The high-density data track is a track formed by high-density data modulated by RS-LDC and RLL (1-7) PP system. A cluster attribute table (CAT) that manages a high-density data cluster included in the data track is recorded in the high-density data track. C
The AT manages attributes (open / not open, normal / defective, etc.) for each of the high-density data clusters forming the data track.

A unique ID (UID) unique to the disc, which is used for copyright protection and the like, by using a high-density data cluster that cannot be disclosed in this data track.
Also, a hash value (hash) for checking data tampering is recorded. Of course, other than this, various kinds of non-public information may be recorded. In the area that cannot be published,
Only specially authorized equipment shall have limited access.

An area (exportable area) formed by a high-density data cluster that can be disclosed in the data track is accessed by an external computer or the like via a general-purpose data interface such as USB or SCSI and is used as a recording area. It is considered as a usable area. For example, in the case of FIG. 4A, in the exportable area, the FAT and the FAT management data file
The state where the AT file system is constructed is shown. That is, the data recorded in the exportable area is not managed by the U-TOC but is managed by general-purpose management information such as FAT, and can be recognized by an external computer that does not comply with the mini disk system. It becomes data.

Music data such as ATRAC system is FAT
When recorded on the system, a playlist is recorded as management information for managing files such as music data. This playlist is created as one file on the FAT system, and the recording / reproducing apparatus 1
The system controller of (1) grasps the address and the like from the playlist when reproducing the music data file.

In some cases, a plurality of data tracks having such a structure are recorded on the disc. In this case, each data track is managed by the U-TOC as one track, and the data in the exportable area in each data track is F
Managed by AT, etc. For example, each data track has its own FAT file system. Alternatively, one F can be used across multiple data tracks.
The AT file system may be recorded.

Further, the information such as the unique ID has been described as an example in which the data is not FAT managed in the data track. However, if the information is not FAT managed, it may be recorded in any logical form. . For example U-TOC
A track for private information should be provided as a track directly managed by the U-TOC.
It may be recorded in / P-TOC. Furthermore, in the U-TOC area, by using a part not used for U-TOC,
A recording area for private information may be provided.

For the sake of explanation, the conventional mini disc is called a "music mini disc".
The mini disc known as is also in the category of this music mini disc. Further, it is naturally possible that music data is recorded as a data file on the high density disc. For example, when downloading a music file from a network or a personal computer.

The storage unit 2 shown in FIG. 1 is a disc drive unit capable of supporting the above-mentioned music mini disc and high-density disc as a general-purpose data recording medium. FIG. 5 shows an example of the configuration of this storage unit 2.

The illustrated disc 90 is the above-mentioned music mini disc or high-density disc. In the storage unit 2, the loaded disk 90 is rotationally driven by the spindle motor 29 by the CLV method. The optical head 19 is used for recording / reproducing with respect to the disk 90.
Is irradiated with laser light. The optical head 19 outputs a high level laser output for heating the recording track to the Curie temperature during recording, and a relatively low level laser output for detecting data from the reflected light by the magnetic Kerr effect during reproduction. . Therefore, although not shown in detail here, the optical head 19 is equipped with a laser diode as a laser output means, an optical system including a polarization beam splitter and an objective lens, and a detector for detecting reflected light. There is. Optical head 19
As the objective lens provided in the disk, for example, it is held by a biaxial mechanism so as to be displaceable in the disk radial direction and the direction in which the disk comes in and out of contact with the disk.

Further, the optical head 1 with the disk 90 sandwiched therebetween.
A magnetic head 18 is arranged at a position opposed to 9. The magnetic head 18 operates to apply a magnetic field modulated by the recording data to the disk 90. Although not shown, a sled motor and a sled mechanism are provided for moving the entire optical head 19 and the magnetic head 18 in the radial direction of the disk.

In this storage unit 2, the optical head 1
9. In addition to the recording / reproducing head system using the magnetic head 18 and the disk rotation driving system using the spindle motor 29, a recording processing system, a reproducing processing system, a servo system, and the like are provided. In the recording processing system, modulation of the first modulation method (EFM modulation / ACIRC encoding) is performed at the time of recording on a mini disc for music.
And a part for performing modulation of the second modulation method (RLL (1-7) PP modulation, RS-LDC encoding) at the time of recording on a high density disc. In the playback processing system, music mini discs (and high-density discs U-
When reproducing TOC, demodulation (EF) for the first modulation method
M demodulation / ACIRC decoding) and demodulation (partial response PR (1,2,1) for the second modulation method when reproducing a high density disc and RLL (1-7 based on data detection using Viterbi decoding) ) Demodulation, RS-
A part for performing LDC decoding) is provided.

Information (photocurrent obtained by detecting the laser reflected light by the photodetector) detected as the reflected light by the laser irradiation of the optical head 19 onto the disk 90 is supplied to the RF amplifier 21. RF amplifier 2
In 1, the input detection information is subjected to current-voltage conversion, amplification, matrix calculation, etc. to reproduce R as reproduction information.
An F signal, a tracking error signal TE, a focus error signal FE, groove information (ADIP information recorded on the disk 90 by track wobbling) and the like are extracted.

At the time of reproducing the mini disk for music, the reproduced RF signal obtained by the RF amplifier is supplied to the EFM demodulation unit 24 and A.
It is processed by the CIRC decoder 25. That is, the reproduced RF signal is binarized by the EFM demodulation unit 24 to form an EFM signal sequence, then EFM demodulated, and further, the ACIRC decoder 2
In step 5, error correction and deinterleave processing is performed. That is, at this point, the ATRAC compressed data state is set. Then, at the time of reproducing the mini disc for music, the B contact side is selected in the selector 26, and the demodulated ATRAC compressed data is output as the reproduction data from the disc 90.
In this case, the compressed data is supplied to the cache memory 3 of FIG.

On the other hand, when reproducing a high density disc, the reproduced RF signal obtained by the RF amplifier is RLL (1-7) PP.
It is processed by the demodulation unit 22 and the RS-LDC decoder 25. That is, the reproduced RF signal is reproduced in the RLL (1-7) PP demodulation unit 22 by data detection using PR (1,2,1) and Viterbi decoding to obtain reproduced data as an RLL (1-7) code string. RLL (1-7) demodulation processing is performed on this RLL (1-7) code string. And further RS
-Error correction and deinterleave processing are performed by the LDC decoder 23. During reproduction of the high density disc, the A contact side of the selector 26 is selected and the demodulated data is output as reproduction data from the disc 90. In this case, the demodulated data will be supplied to the cache memory 3 of FIG.

The tracking error signal TE and the focus error signal FE output from the RF amplifier 21 are supplied to the servo circuit 27, and the groove information is supplied to the ADIP decoder 30.

The ADIP decoder 30 extracts a wobble component by band-limiting the groove information with a bandpass filter, and then performs FM demodulation and bi-phase demodulation to extract an ADIP address. The extracted ADIP address, which is the absolute address information on the disc, is supplied to the system controller 8 shown in FIG. The system controller 8 executes a required control process based on the ADIP address. The groove information is also supplied to the servo circuit 27 for spindle servo control.

The servo circuit 27 reproduces a clock (PLL system clock at the time of decoding) for the groove information, for example.
C based on the error signal obtained by integrating the phase error between
A spindle error signal for LV servo control is generated. Further, the servo circuit 27 sends a spindle error signal,
As described above, various servo control signals (tracking control signal, focus control signal, sled control) based on the tracking error signal, the focus error signal supplied from the RF amplifier 21, or the track jump command, the access command, etc. from the system controller 8. Signal, spindle control signal, etc.) and output to the motor driver 28. That is, various servo control signals are generated by performing necessary processing such as phase compensation processing, gain processing, and target value setting processing on the servo error signals and commands.

In the motor driver 28, the servo circuit 27
A required servo drive signal is generated based on the servo control signal supplied from. The servo drive signal here is a biaxial drive signal (two types of focus direction and tracking direction) for driving the biaxial mechanism, a sled motor drive signal for driving the sled mechanism, and a spindle motor drive signal for driving the spindle motor 29. Becomes With such a servo drive signal, focus control, tracking control for the disc 90 and CLV control for the spindle motor 29 are performed.

When the recording operation is executed on the disk 90, the data is supplied from the cache memory 3. When recording a mini disc for music, the selector 16 is set to B
The ACIRC encoder 14 and the EFM modulator 15 are connected to the contacts, and thus function. In this case, the compressed data from the audio processing unit 10 is interleaved and the error correction code is added by the ACIRC encoder 14, and then the EFM modulation unit 15 performs the EFM modulation. Then, the EFM-modulated data is supplied to the magnetic head driver 17 via the selector 16, and the magnetic head 18 applies a magnetic field to the disk 90 based on the EFM-modulated data to perform data recording.

When recording on a high density disc, the selector 16
Is connected to the A contact, and thus the RS-LDC encoder 1
2 and the RLL (1-7) PP modulator 13 will function. In this case, the high-density data from the memory transfer controller 3 is interleaved by the RS-LDC encoder 12 and an error correction code is added by the RS-LDC method, and then the RLL (1-7) PP modulator 13 performs the RLL.
(1-7) Modulation is performed. Then, the recording data as the RLL (1-7) code string is supplied to the magnetic head driver 17 via the selector 16, and the magnetic head 18 applies a magnetic field to the disk 90 based on the modulation data, thereby recording data. Done.

The laser driver / APC 20 causes the laser diode to perform the laser emission operation during the reproduction and recording as described above.
c Lazer Power Control) also operates. That is, although not shown, a detector for laser power monitoring is provided in the optical head 19, and the monitor signal is fed back to the laser driver / APC 20. The laser driver / APC 20 compares the current laser power obtained as the monitor signal with the set laser power and reflects the error in the laser drive signal,
The laser power output from the laser diode is controlled to be stable at the set value. As the laser power, the values of the reproduction laser power and the recording laser power are set in the register inside the laser driver / APC 20 by the system controller 8.

The above operations (accesses, various servos, data writing, and data reading) are executed based on instructions from the system controller 8.

3. Basic Cache Control Method As described above with reference to FIG. 1, the recording / reproducing apparatus 1 serves as an AV device for the audio data (AV
In addition to recording / reproducing data), the device is connected to a personal computer (or network) 100 to record / reproduce PC data or AV data. Then, the data read from and written to the disk 90 passes through the cache memory 3.

In this example, the cache area of the cache memory 3 is divided into predetermined units N and controlled. Figure 6
Indicates a state in which the cache area is divided and set into N clusters × M blocks. One area (cell) indicated by a box in FIG. 6 is an area of one cluster (32 KBytes), and is an area of N × M clusters as a whole. For convenience of explanation, it is assumed that N clusters are shown in a vertical column and that M blocks are set as one division unit (block). In this way, the cache area in the cache memory 3 is logically divided and set to N clusters × M blocks.

Addresses on the cache memory 3 (cache addresses) are consecutive as cache IDs (Cid). FIG. 7 shows the address Cid of each cell when N = 8. That is, as the address Cid, physically, Cid = 0 to Cid = 8 for each cell.
(M-1) +7 are continuously given, but by setting this as N = 8, Cid = 0 to 7, 8
~ 15, 16-23 ... 8 (M-1) to 8 (M-1)
Up to +7, every eight cells are divided into one block, which is managed in a continuous state.

When data is written to such a cache area, cells are sequentially allocated starting from the remainder value obtained by dividing the data attribute (specifically, the address of the data on the disk 90) by N. N = 8 as shown in FIG.
In this case, cells corresponding to the remainders “0” to “7” exist in each block, but the cells corresponding to the remainder value in the block designated by the value of the pointer cnt are included in the data. It becomes the corresponding cell. The block including the cell serving as the starting point is designated by the value of the pointer cnt that is simply counted up. This pointer c
nt is a pointer that sequentially designates each block as it is counted up. Note that the pointer cn
When t reaches the last block “(M−1) N”, t is reset to a value designating the first block “0N”, so that the cache area can be used as a ring buffer capable of handling long data. It can also be used for.

That is, basically, if the addresses (adrs) on the disk 90 are continuous, the cache is filled vertically from the upper left cell in the figure, and when the bottom is reached, the pointer cnt is set. Go forward one. By repeating this, when the pointer cnt reaches the rightmost end (final block), the pointer cnt is returned to the left end (first block).

Specifically, the cache ID (Cid) is assigned to the address (adrs) on the disk of the data requested to be read from or written to the disk 90 by the following formula. Cid = (N x cnt) + remainer (adrs ÷ N)

Specific examples of such allocation of the cache address Cid are shown in FIGS. Although the address Cid of each cell in the cache area is not shown in FIGS. 8 to 11, it is assumed to be as shown in FIG. Now disk 90
Address as above (adrs) = 3 as 25, 26, 27
It is assumed that a request for writing or reading data for the cluster has occurred.

The "cluster" unit in the above-mentioned high-density disk and the "cluster" unit in the cache area are, for convenience of the disk format and the design of the device,
Although not necessarily coincident with each other, it is assumed here that they coincide with each other for simplification of description, and it is assumed that one cluster of data on the disk is stored in one cell of the cache area. Of course, even if they do not match, the cache control method of this example can be realized by performing the required address conversion calculation.

First, the address (adrs) of the data to be written in the cache area is 25, and the cache area is N = 8.
If the block "0N" is specified with the pointer cnt = 0 at the present, the above formula becomes Cid = (8 × 0) + remainer (25 ÷ 8), and it becomes the retainer (25 ÷ 8). The remainder value "1" is derived. That is, the address Cid = in the cache area
"1" is assigned and the address adrs is displayed as shown in FIG.
= 25 data is written in the cell with the cache address Cid “1”. Continued, address (adrs) = 26,
The respective data of 27 also have the residual values "2" and "3", respectively, and are continuously written in the cells of the cache addresses Cid "2" and "3".

Next, address (adrs) = 32, 33, 3
It is assumed that a request for writing or reading data for 3 clusters as No. 4 has occurred. Then, N of address (adrs) = 32
The remainder value divided by is 0, but in this case, the remaining cells of the block “0N” are Cid = 4,5,6,7, and past the cells of Cid = 0, so the pointer cn
t is incremented to cnt = 1. Therefore, the above formula becomes Cid = (8 × 1) + remainer (32 ÷ 8), and the address Cid = 8 is set. As shown in FIG. 9, the data at the address adrs = 32 becomes the cache address Cid “8”. Written to cell. The following data at addresses (adrs) = 33 and 34 also have Cid =
9, 10 are calculated, and the cache address Cid “9”
It is written in the cell of "10". In this case, Cid =
The cells of 4, 5, 6, and 7 are invalid cells.

Further next, addresses (adrs) = 14, 15,
If a request to write or read data for 3 clusters as 16 is issued, these data are similarly stored in C
The cells with id = 14, 15, 16 are allocated and written as shown in FIG. Similarly, the address (ad
rs) = 50, 51, 52, 53, data write or read request for 4 clusters, address (adrs) = 45,
FIG. 10 shows a state in which a write or read request for data of four clusters as 46, 47, and 48 is generated, and Cids are respectively assigned and written in the cache area.

In the above example, for example, about 3 clusters,
This is an example when a write or read request for data of a relatively short data size occurs, but the same applies to the case of long data such as AV data. For example, when the pointer cnt = 5, the address (adrs) = 100, 101, 102 ...
.. When a request for writing or reading a large amount of data is generated, as shown in FIG.
Cid = 44 is assigned to the data of = 100 by the above formula, and the cells of Cid = 44 are sequentially assigned and written as shown in the drawing.

Although a write / read request for continuous data of a large size as in the case of AV data is actually generated by being divided into a plurality of times, data having consecutive addresses (adrs) is sequentially written. / Because it is a read request, it will be written as a result as shown in FIG.

With such a cache control system, the following advantages can be obtained. First, continuity is ensured on the cache memory for consecutive addresses (adrs). However, as can be seen from FIGS. 8 to 10,
If the addresses (adrs) are not consecutive, maximum N-
A gap of 1 cluster (cell) is generated.

When searching the cache, the cache address Cid is changed from the requested address (adrs).
It is only necessary to search for one round to the left from that point as the starting point, and the number of searches (the number of search target cells) at the time of cache search is 1 / N of the case of searching all cells,
The search process can be speeded up. This cache search process will be described later.

Further, the cache address Cid allocation processing can be performed by the simple arithmetic processing shown by the above equation. In particular, it is a remarkably simple operation as compared with the allocation processing by the conventional hash algorithm. This makes it possible to simplify the calculation program, reduce the processing load, and speed up the calculation. In particular, by setting the value of N to a power of 2 as in the above “8”, it is possible to speed up the calculation of the cache address Cid by the CPU (system controller 8).

In such a cache control method, the possibility of a cache hit is increased as much as PC data, and the responsiveness is improved by improving the searchability. By making data to be written (or read) continuously on the cache memory 3 as well, the write / read operation to / from the disk 90 is made efficient, and the purpose is to continuously buffer a large data size such as AV data. It will be suitable for.

By the way, if the cache address Cid is allocated once in the N × M cache area, the cache address Cid may be allocated in the gap left as an invalid cell in the second cycle. However, in this case, in order to promote fragmentation of the cache data, it may be used to judge that the cache is full after one round and start deregistration.

Deletion of registration means returning a cell in which data has been written (a cell registered as a cell that has data and cannot be used for writing) to an empty cell in which data can be written. The data written in the cache memory 3 is necessary data at least until writing to the disk 90. Writing the data of the cache memory 3 to the disk 90 is called “flash”, but a cell having data that has not been flushed is in a state of being “registered” as a cell that cannot be written. Then, deregistration is performed on at least the flushed cell. However, flash is not a condition for deregistration. Considering a cache hit, for example, it is preferable to reserve data in a registered cell without deregistering it as long as possible. On the other hand, since the cache area is limited, it cannot be kept in the registered state forever. Under such circumstances, the registration deletion is performed at any time according to the cache status, capacity, request generation status, write-requested data size, and the like.

The data writing state in the cache memory 3 and the attribute of the written data are managed in the cache management memory 11 by the cache management data. As shown in FIG. 12A, the cache management data has management information corresponding to each cache address Cid (that is, each cell). Then, corresponding to each cache address Cid, as shown in FIG. 12B, the address Cid, the FAT cluster number (adrs),
It manages flash information and cache sector information.

The FAT cluster number (adrs) indicates the address on the disk 90 as an attribute of the data written in the cache address Cid. As described above, since the high density disc as the disc 90 records data managed by the FAT system, the (adrs) value is stored as the FAT cluster number correspondingly. That is, in the example of the present embodiment, the FAT cluster number is designated as an address when a write or read request is made from the disk. For example, as shown in FIG. 8, the address (adrs) as the FAT cluster number is added to the cell with Cid “1”.
When “= 25” is written, the FAT cluster number (adrs) = “25” is registered as the management information of Cid “1”.

The flush information is information indicating whether or not the data of the cache address Cid has been flushed. When the data is written to the cache address Cid, the value is set to "unwritten", and when the data is flushed, the value is updated to "disk written".

The cache sector information is information indicating whether or not data is written for each sector of the cache address Cid. One cache address Ci
d is added to cells in the region of 1 cluster, and if 1 cluster is composed of 32 sectors, for example, the cache sector information is 32 bit flag information in which 1 bit corresponds to 32 sectors, The presence or absence of data for each sector is indicated.

For example, by configuring the cache management data in this way, the system controller 8 refers to the cache management data in the cache management memory 11 to grasp the state of each cell in the cache area,
It is possible to perform necessary cache control such as flush processing, deregistration processing, and pointer cnt counting processing. Further, the cache search described later actually searches this cache management data.

4. Write / Read Operation Here, the flow of operation at the time of recording / reproduction performed via the cache memory 3 controlled as described above will be described. FIG. 13 shows a data write request to the disk 90 (when a disk is written), and FIGS. 14 and 15 show the disk 90.
3 shows the flow when a data read request is issued from the disk (when a disk is read). Note that FIG. 14 shows a case where no cache hit occurs, and FIG. 15 shows a case where a cache hit occurs.

In each figure, the recording / reproduction management task or P
Processing of C, file task or USB task, cache task, drive task is shown. The recording / reproducing management task, the file task, the USB task, the cache task, and the drive task are tasks activated by the system controller 8 (CPU). PC is a personal computer (or network) in Fig. 1.
The external processing is shown as 100.

The drive task includes a recording / reproducing operation in the storage unit 2 and the cache memory 3 and the storage unit 2.
It has a function of controlling data transfer between the two. The cache task controls the cache memory 3, that is,
It has a function of performing cache control including calculation of the cache address Cid. The USB task has a function of controlling data communication between the USB interface 4 and the personal computer 100 and writing / reading to / from the cache memory 3 for that purpose. The file task has a function of performing management / conversion by the FAT cluster for data input / output via the input / output processing unit 5 and address detection based on FAT or playlist.
The recording / reproducing management task has a function of instructing input / output of audio data via the input / output processing unit 5, that is, recording / reproducing to / from the disc 90 by the recording / reproducing apparatus 1 alone based on a user operation.

As described above, for recording / reproducing data to / from the disc 90, the recording / reproducing apparatus 1 is used alone as an AV device, or is connected to the personal computer (or network) 100 and is a storage device for them. There are two types when used as.
When connected to the personal computer 100, for example, the personal computer 100 issues a data write request and a data read request for the disk 90. In this case, the USB task, cache task, and drive task function to write or read data.

On the other hand, the recording / reproducing apparatus 1 can function as an AV device and can record / reproduce audio data or the like in accordance with a user's operation. In response, the recording / reproducing management task activated on the system controller 8 issues a data write request and a data read request for the disk 90. In this case, the file task, cache task, and drive task function to write or read data.

The operation regarding the data cache and the information to be transmitted, which are the ranges described here, are the same between the recording / reproducing management task and the personal computer 100.
Since the file task and USB task are the same,
These are commonly explained. It should be noted that the personal computer 100 or the like handles the address (adrs) on the disk 90 as a logical address, and the drive task handles the address (adrs) on the disk 90 as a physical address. Therefore, in practice, a logical / physical address conversion task, for example, U
It is interposed between the SB task and the cache task, and the cache task and the drive task use the physical address (adrs), and the USB task and the personal computer 100 side use the logical address (adrs).
Will be treated respectively, but for simplification of explanation,
The logical / physical address conversion is omitted.

First, the disk write operation will be described with reference to FIG. In step W1, the recording / reproduction management task (or PC) issues a write request (write request) and transfers write data. The write request includes information on the address (adrs) and the data length (length) that are the starting points of writing on the disk 90. Then, the write data (Wdata) is transferred.
When a write request is generated and the write data (Wdata) is supplied to the input / output processing unit 5 (or USB interface 4), the file task (or USB task) caches the cache task as the procedure W11. Make a write request. In this cache write request, information on the address (adrs) and the data length (length) is transmitted to the cache task. In step W21, the cache task uses the expression “Cid =” described above based on the address (adrs).
(N × cnt) + remainer (adrs ÷ N) ”, and the cache address Cid calculated in step W22 is used for the file task (or U).
SB task). That is, the cache memory 3 transmits the cache address to which the write data (Wdata) should be written.

In response to this, the file task (or USB
Task) is write data (Wdata) as procedure W12.
Is written in the cache memory 3 starting from the designated cache address Cid. Then, when the cache writing is completed, in step W13, the cache writing is notified to the cache task.
At this time, when the cache writing is completed, the recording / reproducing management task (or PC) is notified of the writing completion. For example, the personal computer 100 or the recording / reproducing management task, after transmitting a write request once, reports the completion of writing to the write request as a command OK notification, and after the command OK, the next write request becomes possible. . Therefore, when the personal computer 100 and the recording / reproducing management task continuously make write requests, the next write request can be generated at this point.

When the cache write completion is notified from the file task (or USB task), the cache task requests the drive task to write to the disk 90 in step W23. Drive task is procedure W3
1, in response to a write request from the cache task,
Write data (Wdata) held in the cache memory 3
Is read out, transferred to the storage unit 2, and recorded from the specified address (adrs).
Then, when the data writing to the disk 90 is completed in the storage unit 2, the completion of the disk writing is notified to the cache task in step W32. By the processing up to this point, the writing of the write data (Wdata) to the disc 90 is completed.

The flash operation for writing the data stored in the cache memory 3 to the disk 90 in this way does not have to be executed for each write request, and may be executed at a predetermined flash timing. For example, the data written in the cache memory 3 may not be flushed for a certain period of time, and may be collectively flushed at a required time point, so that the flush operation can be made efficient.
In particular, the write operation on the disk 90 can be made efficient. Further, at that time, the responsiveness of the cache memory 3 can be improved by performing a flash process with a specific priority rather than simply sequentially flushing each data. This processing will be described later in flash processing example <1><
2>.

By the way, the state of the cache memory 3 is managed by the cache management data as described above. In the cache management data, the FAT cluster number (adrs) shown in FIG. 12 is written for the corresponding cache address Cid, and the cache sector information is updated. Also, write data (Wdata) stored in the cache memory 3
Is written (that is, flushed) to the disk 90 in step W31, the write data (Wdat
The flush information of the cache management data is updated so that the cache address Cid in which a) was stored is flushed.

Next, the operation at the time of disc reading will be described with reference to FIG. This is when there is no cache hit. In step R1, the recording / reproduction management task (or PC) issues a read request (read request). As a lead request,
It includes information of an address (adrs) and a data length (length) which are the starting points of reading on the disk 90. When a read request is issued, the file task (or USB task) makes a cache read request to the cache task as procedure R11. In this cache read request, information of address (adrs) and data length (length) is transmitted to the cache task. In the procedure R21, the cache task uses the expression "Cid =" based on the address (adrs).
(N × cnt) + remainer (adrs ÷ N) ”, and the cache address Cid is calculated. Do a search. This search process will be described later. Here, it is assumed that the data of the requested address (adrs) is not found even if the cache search is performed. That is, since no cache hit occurs, the cache task continues with the procedure R22.
Then, a read request from the disk 90 is issued to the drive task. At this time, the address (adrs) to be read from the disk, the data length (length), and the cache address Cid of the cache memory 3 into which the read data is written are notified. In step R31, the drive task controls the storage unit 2 to reproduce the data of the specified data length (length) from the specified address (adrs) in response to the write request from the cache task. Then, the read data is written in the cache memory 3 based on the designated cache address Cid. Then, when the data reading from the disk 90 in the storage unit 2 and the writing to the cache memory 3 are completed, the drive task notifies the cache task of the disk reading completion as procedure R32.

Then, the cache task executes the procedure R23.
As a response corresponding to the cache read request in the above procedure R11, the cache address Ci to be read is to be read.
Notify d. That is, the cache address Cid in which the read data (Rdata) is written by the drive task is notified.

In response to this, the file task (or USB
Task) reads the read data (Rdata) from the specified cache address Cid in step R12,
The data is transferred to the input / output processing unit 5 or the personal computer 100 side. When the cache read and transfer are completed, the recording / reproduction management task (or PC) is executed as procedure R13.
Notify the completion of reading. Incidentally, in practice, the procedure R
The notification of the read completion in 13 may be issued at the same time as the data transfer in the procedure R12.

For example, the personal computer 100 or the recording / reproducing management task can make the next read request in response to the read completion notification. By the way, when the data is continuous and long data such as audio data, and it is necessary to maintain temporal continuity as an output from the input / output processing unit 5, a new completion of the reading is waited for after the notification of the reading completion. Requesting data may not be in time, and as the reproducing operation in the storage unit 2, even if the data is continuous on the disk 90, the reading is temporarily stopped and the read request is waited for. Starting reading again is a very inefficient operation and consumes extra power. Therefore, normally, on the drive task side, a pre-read operation is performed as a read operation of the storage unit 2 based on, for example, an instruction of the cache task, and data not requested yet is also buffered in the cache memory 3. .

For example, when the data requested in the procedure R31 is read from the disk 90 and written to the cache memory 3, the disk 90 is continuously read continuously, and the data is continuously written to the cache memory 3. . Of course, the address on the disk 90 (ad
Since rs) is continuous, the cache address Cid on the cache memory 3 is also written to continuous addresses. By such a prefetch operation, the buffering as shown in FIG. 11 is performed. In particular, in the case of audio data, it is highly probable that the next requested data will be a continuous data following the previously requested data. Therefore, if the read-ahead buffering is performed, the data can be transferred very quickly by the cache hit described below when the next read request is made. As a result, even when the input / output processing unit 5 outputs data continuously in terms of time, the input / output processing unit 5 should be in time.
Can be supplied with read data.

In the cache management data for managing the state of the cache memory 3, the read data (Rdata) is written in the cache memory 3 in the above procedure W32, and the cache address Cid shown in FIG. The FAT cluster number (adrs) shown in is written and the cache sector information is updated.

Next, as the operation at the time of disk read, the operation in the case of a cache hit will be described with reference to FIG. It is the same as in the case of FIG. 14 that the recording / reproduction management task (or PC) issues a read request in step R1 and the file task (or USB task) issues a cache read request to the cache task in step R11. . In the procedure R21, the cache task uses the expression “Cid = (N × cnt) + remainer (adrs) based on the address (adrs) in the procedure R21.
÷ N) ”to calculate the cache address Cid,
With the cache address Cid as a starting point, a search is performed on the rows having the same residual value in the above-described N × M matrix-set cache areas. Assuming that the data of the requested address (adrs) exists in the cache memory 3 as the cache search result, the cache task reads in step R23 as a response corresponding to the cache read request in step R11. Is notified of the cache address Cid to be executed. That is, the read data (Rd
The cache address Cid in which ata) is written is notified.

In response to this, the file task (or USB
Task) reads the read data (Rdata) from the specified cache address Cid in step R12,
The data is transferred to the input / output processing unit 5 or the personal computer 100 side. When the cache read and transfer are completed, the recording / reproduction management task (or PC) is executed as procedure R13.
Notify the completion of reading.

As described above, when the read data (Rdata) relating to the read request is already stored in the cache memory 3, the read operation in the storage unit 2 is not required and the input / output processing unit 5 or the personal computer is not required. The requested read data (Rdata) can be supplied to the computer 100 side, and the responsiveness becomes very good as compared with the case where no cache hit occurs as described above.

5. Cache Search Process The cache search process executed by the cache task as described above will be described. That is, this is a process of determining whether or not the requested data exists in the cache memory 3.

In the case of the present example in which the N × M matrix management of the cache area is performed as described above, a cell (cache address Ci
d) is “Cid = (N × cnt) + remainer (adrs ÷ when the address (adrs) on the disk 90 is designated.
N) ”. Here, the value of the pointer cnt is the count value at the time of writing the data in the cache memory 3, and does not always match the value at the time of search.
The value of (N × cnt) cannot be specified at the time of search, but the cache address Ci for storing the data of a certain address (adrs)
d will be defined by the term for obtaining the remainder value as remaining (adrs ÷ N).

For example, as shown in FIG. 7, if N = 8, there are rows with a remainder value of 0 to rows with a remainder value of 7 in the N × M matrix. Therefore, some address (ad
When the data of (rs) is requested, its address (adr
It is only necessary to retrieve the row of the residual value obtained by the operation of remainr (adrs ÷ N) using (s).

A specific example will be described. For example, consider a case where data is stored in the cache memory 3 as shown in FIG. It is assumed that the pointer cnt = 4 at this point. Here, it is assumed that the data of the address (adrs) = “25” is requested. At this time, the cache task is Ci
By performing the calculation of d = (N × cnt) + remainer (adrs ÷ N), the cache address Cid = 33 shown in FIG.
You will get Where remainr (adrs ÷ N) = 2
And Cid = 33 is the cache address of the row with the remainder value “2”. The cache task has Cid = 33
Starting from, each cache address Cid in the row of the remainder value 2 shown by the thick frame in FIG. 16 may be searched.

In this case, no data is stored in Cid = 33. Next, for Cid = 25, no data is stored. Next, no data is stored for Cid = 17. The data of adrs = 33 is stored in Cid = 9. Then, when Cid = 1 is confirmed, the data of adrs = 25 is discovered, and it can be determined that the cache hit.

If Cid = 1 does not apply, the search target is the cache address of the remainder value 2 of the final block (M-1) N (Cid = 8 (M-1) +1).
Proceed to. Finally, if the corresponding data does not exist even if all the cache addresses having the remainder value 2 are searched, it is determined that no cache hit has occurred.

As can be seen from this specific example, the cache search process is executed by the target cache address Cid.
(Cell) may be 1 / N of all cells. That is, normally, in the case of a cache search, it is necessary to check the presence / absence of requested data for all cache addresses Cid in the cache area. It can be speeded up. This makes it possible to improve the responsiveness when requesting data. The actual search is performed on the cache management data. That is, in the case of the above example, the process is to confirm the FAT cluster number of each cache management data for Cid = 33, 25, 17 ...

FIG. 17 shows a flow chart for realizing such a retrieval process in the cache task. In step F101, a cache address Cid is obtained by calculating Cid = (N × cnt) + remainer (adrs ÷ N) based on the requested address (adrs). In the case of the above example, the cache address Cid = 33 is obtained by this processing. Then, through step F102, F1
03, the cache address Cid (Cid =
With reference to the cache management data for 33), it is confirmed whether the requested data (adrs = 25 in the above example) is stored. If the requested data does not exist, the process proceeds to step F104, and the value of the cache address Cid to be searched is set to Cid-N. In the above example,
The new cache address Cid becomes Cid = 25 due to 33-8 = 25, and steps F102 → F10.
3, the cache address Cid (Cid =
With reference to the cache management data of 25), it is confirmed whether the requested data (adrs = 25 in the above example) is stored. With this loop, for example ad
When data of rs = 25 is requested, each cell in the row having the remainder value “1” is sequentially searched forward. In step F104, when the cache address Cid reaches the value of the first block (0N), it is set to the value of the cache address of the row having the same remainder value in the last block.

If the requested data is found in step F103, the flow advances to step F105 to notify the cache address Cid as a cache hit. That is, in step R23 of FIG. 15, the cache address Cid is notified to the file task (or USB task).

On the other hand, if the search is continued without finding the requested data in step F103 and the line having the surplus value is circled, the process proceeds from step F102 to F106, and it is assumed that no cache hit occurs. In this case, the processing shown in FIG. 14 is performed.

By the way, as can be seen from the cache search process, the cache search process can be made more efficient as the value of N is larger in the case of performing cache control by handling it in an N × M matrix. On the other hand, FIG.
As can be seen from the cache write operation described with reference to FIG. 10, the cache control method of this example uses the cache address C
When the id is determined, the larger the value of N, the higher the possibility that more invalid cells are left. That is, in the cache control method of this example, the search speed and the memory utilization efficiency (fragmentation suppression rate) are in a trade-off relationship.

For example, as schematically shown in FIG.
If (a) is a matrix with N = 8 and the N value is smaller than 8 as shown in FIG. 18 (b), the memory utilization efficiency can be improved. That is, on average, the number of cells left in an empty state during writing is relatively small. On the other hand, the number of cells in the horizontal direction, that is, the number of blocks M
Since b increases and the number of addresses having the same remainder value increases, a larger number of cache addresses Ci during search
The search of d becomes necessary, and the search speed decreases. On the other hand, when the N value is larger than 8 as shown in FIG. 18C, the number of cells in the horizontal direction, that is, the number of blocks Mb decreases, and the number of addresses having the same remainder value decreases. The number of cache addresses Cid that are sometimes searched is reduced, and the search speed is improved. On the other hand, on average,
Many cells are left in an empty state at the time of writing, and the memory utilization efficiency is reduced.

Under these circumstances, the N value in the cache control method of this example is preferably set to an appropriate value according to the usage mode of the device, the recording medium, the type of data to be handled, and the like. Further, for example, the setting of the N value may be changed according to the usage state or the user setting. Further, if the N value is automatically optimized and controlled according to the operating condition, the most preferable cache control can be executed.

6. Flush processing example <1> In this example, the cache address Cid is set and searched for the cache memory 3 as described above, but it is suitable as a cache control method for both AV data and PC data. Will be things. That is,
When AV data is downloaded from a network or a personal computer, or AV data is recorded / reproduced as an AV device, data continuity can be buffered and high-speed performance can be realized. Further, since the cache memory can be searched by the position corresponding to the remainder when the address on the disk is divided by N (predetermined value), efficient and quick search is realized. Therefore, it is possible to secure high-speed response when reading and writing PC data. Further, since complicated calculation such as a hash algorithm is not required, the address allocation on the cache can be simplified, the processing load can be reduced, and the cost of the device can be reduced. Further, since the continuity is ensured on the cache memory for the data of consecutive addresses, the recording / reproducing efficiency on the disk can be improved,
Power saving effect can be expected by reducing the number of access.

In the present example, in addition to such an effect, by performing the flash processing control according to the attribute of the data temporarily stored in the cache memory 3, the above effect can be further improved. . In the flash processing example <1> described below, when the data stored in the cache memory 3 is flashed, the time required for the flash processing for each data is determined from the attribute of each data, and the data with the short processing time is given priority. This is an example of a process of flashing.

Incidentally, for example, the personal computer 10
It is assumed that the data written to the cache memory 3 in response to the write request from 0 is a sector unit, while the storage unit 2 writes to the disk 90 in a cluster unit (for example, 1 cluster = 32 sectors). To do. As data having different flash processing times, it is assumed here that there are three patterns depending on the attributes of the data. There are three types of data: continuous data, discontinuous data, and blocking required data.

The continuous data means the data of each cluster in which the addresses (adrs) are physically continuous on the disk 90. When a plurality of clusters, which are continuous data, are continuously flash-processed, each cluster data can be written in the storage unit 2 by one continuous scan with respect to the track of the disk 90. That is, it is not necessary for the optical head 19 and the magnetic head 18 to access during writing of a plurality of clusters, and the most efficient
Moreover, it is possible to write a plurality of clusters in a short time. In other words, if you continuously flash the continuous data,
The flush time per cluster is very short.

The non-consecutive data is data of clusters whose addresses (adrs) are not physically continuous on the disk 90. That is, it is one piece of cluster data stored in the cache memory 3 to be flushed, but cluster data at addresses (adrs) before and after that is not stored in the cache memory 3 as data to be flushed. . When a plurality of non-contiguous data is flashed, the optical head 19 is written every time the cluster as the non-contiguous data is written to the disk 90.
Also, the magnetic head 18 needs to access the address (adrs). Therefore, the flush time per cluster becomes longer than that in the case of the continuous data.

The data necessary for blocking is the disk 9
This data requires a blocking process when writing to 0. As described above, the personal computer 100
Since the write data supplied from the disk unit is a sector unit and the write unit to the disk 90 is a cluster unit, a blocking process may be required in some cases. The blocking process will be described with reference to FIG.

FIG. 20A shows a cache address Cid = x, x + 1, x + in the cache memory 3.
The area of 3 clusters of 3 is shown. As described above, the disk 90 is assigned to the cell of each cache address Cid.
Data corresponding to a remainder value calculated by using the above address (adrs) is stored. Here, for example, it is assumed that the write data WD from the personal computer 100 is supplied in units of sectors and, as shown in the drawing, each of the data constitutes a part of a cluster. At this time, the cache address Cid = x, x + 1, x + 3
In each of the areas, in the portion where the write data WD is not written, currently invalid data such as a part of the data accumulated in the past exists. It is not possible to flush such write data WD as it is on the disk 90.
It cannot be done because the writing unit to is a cluster. This is because, for example, when the cluster data of the cache address Cid = x is flushed, even the invalid data portion is written to the disk 90, and the necessary data is overwritten and erased by the invalid data on the disk 90.

Therefore, in this case, the blocking process is performed. First, the write data WD from the disc 90
The cluster data of the address (adrs) corresponding to is read. For example, an area is secured in the RAM 10 and read data DSD from the disk 90 is stored as shown in FIG. Then, as shown in FIG. 20C, the sector data corresponding to the invalid data area in the read data DSD is written in the cache memory 3. Then, the cache address Cid = x, x + 1, x
The data at +3 is in a data state in which only the sector of the write data WD is rewritten with respect to the data DSD at the corresponding address (adrs) on the current disk 90. When the flash process is performed in cluster units by blocking in this way, the disk 9
On 0, only the sector of the write data WD is rewritten, that is, the disk writing according to the write request from the personal computer 100 or the like can be realized.

The data for which such a blocking process is required occurs not only in the case of a write request in units of small sectors but also in the case of a request for writing a relatively long data. For example, even when data having consecutive addresses (adrs) is supplied as the write data WD as shown in FIG. 20D, it is rare that the total data length is exactly an integral multiple of the cluster. Therefore, the data of the last several sectors is, for example, the cache address Cid = x.
It is stored in a part of +2. That is, the data stored at this cache address Cid = x + 2 becomes blocking required data.

The data requiring the above-mentioned blocking processing is called the blocking-required data. However, since the blocking processing is required at the time of the flush processing for this blocking-required data, the above-mentioned data is required for the flush processing. It will take longer than continuous or discontinuous data.

Regarding the data stored in the cache memory 3 as described above, it can be seen that there are three patterns of flush processing time per cluster when considering continuous data, discontinuous data, and blocking required data.
In the flash process of this example, when the flash is executed, the continuous data is preferentially flushed first, then the non-continuous data is flushed, and finally the blocking required data is flushed.

FIG. 19 shows a control procedure of flush processing by a cache task, for example. As described above with reference to FIG. 13, the flush process does not necessarily have to be executed each time a write request is made, and the cache memory 3
The write data can be stored in and then flashed. For example, in such a case, the cache task is
The flash operation is controlled by the process of 9.

First, in step F201, the attribute of the cache write data (data requested to be written and written in the cache memory 3), in this case the disk 9 is used.
The continuous data is searched from the address (adrs) on 0.
As described in FIG. 12, the cache management data includes
The FAT cluster number (adrs), flash information, and cache sector information are included for each cache address Cid. In the continuous data search process of step F201, these data may be confirmed for each cache address Cid. That is, the cache address Cid that has not yet been flushed from the flush information
To extract. Of the extracted cache addresses Cid, the cache address Cid in which the write data WD is written for all the sectors forming one cluster by referring to the cache sector information, that is, the cache address Cid that does not require blocking To extract. Finally, in each of the extracted cache addresses Cid, a plurality of cache addresses Cid having consecutive FAT cluster numbers (adrs) are extracted. The data written in each cache address Cid as the plurality of cache addresses Cid retrieved in this way is the “continuous data” that can execute the flush in a short time as described above.

Therefore, in step F202, flushing is performed for each "consecutive data" retrieved. That is, FIG.
In step W23, the disk write request for each "continuous data" is issued to the drive task. Upon completion of flushing all of the "continuous data", the process proceeds from step F203 to F204.

In step F204, non-continuous data that does not require blocking is searched. That is, this is to extract the cache address Cid that has not been extracted as continuous data from the cache addresses Cid that have not been flushed and that the blocking process is unnecessary as described above. When the "non-continuous data" are extracted in this way, in step F205, flushing is performed for each of the retrieved "non-continuous data". That is, the procedure W in FIG.
As indicated by 23, the disk write request for each “non-consecutive data” is issued to the drive task.

When there are a plurality of non-contiguous data, it is preferable that the non-contiguous addresses (adrs) are flushed in ascending order of the value. This makes the access of the optical head 19 and the magnetic head 18 which scans the spiral track on the disk 90 efficient. This is because, for example, when the addresses are relatively close to each other, the access by the track jump is unnecessary, or only the forward access is generated and the backward access is eliminated, so that the access time can be minimized. When the flush of all the "non-consecutive data" is completed, the process proceeds from step F206 to F207.

At step F207, the blocking necessary data is searched. That is, this is to extract the cache address Cid other than the cache address Cid for which the blocking process is unnecessary, from the cache addresses Cid that have not been flushed as described above. When "blocking required data" is extracted in this way,
In step F208, blocking and flushing are performed on each of the retrieved "blocking required data".
Although the blocking process is not shown in FIG. 13, in this case, the cache task requests the drive task to read the data necessary for blocking,
Blocking is performed on the cache memory 3. Then, as shown as procedure W23, a disk write request for each "blocking required data" is issued to the drive task. When the flush of all the "blocking required data" is completed, the flush process is completed.

The effects of such flash processing control will be described with reference to the models of FIGS. 21 and 22. FIG. 21 shows a case where the flash process is performed under the above control. FIG. 22 shows, for comparison, a case where the sequential flash processing is simply performed without depending on the above control.

21 (a) and 22 (a) show each cache address Cid (cell) of the cache memory 3, and each cache address Cid includes
As shown in the figure, three types of data with different flash processing times (continuous data,
It is assumed that non-consecutive data and blocking required data) are stored. Assuming that the flash processing time per cluster is 10 msec for continuous data and 2 for discontinuous data.
It is assumed that 0 msec and blocking required data are 40 msec.

A case where the data of each cache address Cid (cell) is simply flash-processed from the states of FIGS. 21A and 22A will be described with reference to FIG. When the flash processing is sequentially performed from the data of the head cell, the 14 cells from the head store the necessary blocking data that requires 40 msec for processing, as shown in FIG. After 500 msec from that, 12 cells are already flushed as shown in FIG. 22 (b) (the flushed cells are outlined). Further, when 1500 msec has elapsed from the start of flushing, 47 cells have been flushed, as shown in FIG.

In the model of FIG. 22A, there are 18 continuous data cells, 22 non-continuous data cells, and 28 blocking required data cells. Therefore, flush start of all cells is completed. Until (18 × 10) + (22 × 20) + (28 × 40) = 1
It will be 740 msec.

In the case of the flash processing under the control of the present example shown in FIG. 21, first, continuous data is flashed in each cell as shown in FIG. 21A. As shown in (b), 34 cells have been flushed. In addition, when 1500 msec has elapsed from the start of flashing, as shown in FIG.
As in (c), 62 cells have been flushed.
The time from the start of flashing to the completion of flashing of all cells is 1740 msec, which is the same as in the case of FIG.
Becomes

Comparing FIG. 21 and FIG. 22, the time until the completion of flushing all data is the same, but according to the control of this example, it is determined that a larger number of cells have been flushed at an early stage from the start of flushing. Become. For example 500 ms
At the time of ec, only 12 cells are flushed in FIG. 22, whereas 34 cells are flushed in FIG. It is possible to newly store the data in the flushed cell by the deregistration process. That is, according to the control of the present example, the usable area in the cache memory 3 can be secured as soon as possible, that is, the write request and the read request from the write data from the personal computer 100 can be quickly recovered. It is a thing. In particular, even when the unflashed data is accumulated in most of the cache area, it is possible to quickly secure the cells to be used for the subsequent processing.

As described above, it is efficient to collectively flush the data stored in the cache memory 3 to some extent. On the other hand, if the unflushed data is held for a long time, the usable area in the cache memory 3 decreases and the personal computer 1
A write request or a read request from 00 or the like cannot be immediately dealt with (in order to respond, a required number of cells must be flushed to bring them into a deregistration state). According to the control of this example, a large number of cells have been flushed at the earliest possible time after the start of flushing, and the cells can be deregistered and made usable so that write requests and read requests can be responded to. The responsiveness of 3 can be improved. Further, for this reason, it is also acceptable to save the unflashed data so that the flashes are collectively performed to some extent, and the efficiency of the flash processing can be improved.

7. Flush processing example <2> In the flash processing example <2>, too, flashing is performed by providing a predetermined priority order. In this case, the cache hit rate as an attribute of information is used as the reference of the priority order.
For example, with respect to the data stored in each cache address Cid, statistics of the read hit rate and the write hit rate are obtained, and if the write hit rate is high, the priority of the flash execution is lowered. If the read hit rate is high, the priority of deleting the registration from the cache memory 3 is lowered. A read hit is a case in which the data to be read request remains in the cache memory 3 without being deregistered. The write hit is a case where the cluster data including the sector data supplied by the write request remains without being deregistered. The read hit rate and the write hit rate for the data stored in each cache address Cid may be, for example, the ratio of the number of hits to the number of read requests and write requests, or simply count the number of hits. Good.

For example, in this case, as the cache management data, as shown in FIG.
In the management data corresponding to id, cache hit information, flush rank information, and registration deletion rank information are stored in addition to the contents described in FIG.

The cache hit information is the read hit rate and the write hit rate of the data stored in the cache address Cid. These are updated each time a cache hit occurs. The flash order information stores the priority order of the flash for the data stored at the cache address Cid. Priority is 2 levels, 3
It may be set in a plurality of stages or more.
The registration deletion order information stores the priority order when the registration deletion is performed on the data stored in the cache address Cid. The priority may be set in two stages, three stages, or a plurality of stages higher than that. For example, for explanation,
The flash ranking information and deregistration ranking information are the first
Three levels of level (priority), second level (normal), and third level (non-priority) are set. The management information is cleared by deleting the registration.

The cache task is the cache memory 3
In response to the data being written to a certain cache address Cid, the cache management data corresponding to that cache address Cid is updated. That is, regarding the cache management data having the structure shown in FIG. 23 stored in the cache management memory 11, the cache address C
As the id management information, the FAT cluster number, flash information, cache sector information, cache hit information, flash order information, and registration deletion order information are set. At this time, the cache hit information is "0".
The flash order information registration deletion order information is set to "first level (priority)".

After that, when a cache hit occurs,
The read hit rate and the write hit rate are updated as cache hit information. When the read hit rate and the write hit rate are updated, it is determined whether the read hit rate and the write hit rate have exceeded the first and second predetermined values respectively set, and the flash result is determined based on the result. Update ranking information and deregistration ranking information. That is, when the read hit rate exceeds the first predetermined value, the registration deletion order information is set to "second level (normal)". When the read hit rate exceeds the second predetermined value, the registration deletion order information is set to "third level (non-priority)". If the write hit rate exceeds the first predetermined value, the flash rank information is set to "second level (normal)". When the write hit rate exceeds the second predetermined value, the flash rank information is set to "third level (non-priority)".

As described above, the cache task updates the cache hit information, the flush order information, and the registration deletion order information in the cache management data as needed. That is, the higher the hit rate, the lower the priority for flushing or deregistration.

When the data stored in the cache memory 3 (data that has not been flushed) is flushed, the cache task performs the processing shown in FIG. First, in step F301, the variable n is set to "1", and in step F302, the nth priority data regarding the flash is searched. That is, first, the data set as the "first level (priority)" is searched. For this reason, the flash information and the flash order information in the cache management data are confirmed, and the cache address Cid whose flash order information is not flushed and whose flash order information is "first level (priority)" is extracted. Go. Then, in step F303, the data written in the one or more cache addresses Cid thus searched is flushed. That is, as shown as procedure W23 in FIG. 13, the disk write request is issued to the drive task.

When flushing of all the retrieved data of each cache address Cid is completed, the process proceeds from step F304 to F305. If there is data that has not been flushed, the variable n is incremented in step F306 and the process returns to step F302.
In this case, the variable n = 2, and in step F302,
Second priority data, ie "second level (normal)"
The data (cache address Cid) which is said to be searched. Then, the cache address Cid retrieved in step F303 is flushed. Also,
If flushing is completed for all the "second level (normal)" data and there is data that has not been flushed yet, the variable n is incremented in step F306 and the process returns to step F302. In this case, the variable n = 3, and in step F302, the third priority data, that is, the data (cache address Cid) that is the “third level (non-priority)” is searched. Then, the cache address Cid retrieved in step F303 is flushed.

When the flush is completed for all the data to be flushed, the flush process is terminated from step F305.

According to such a flash process, the flash process is delayed for data having a high write hit rate. Data with a high write hit rate is likely to be written immediately thereafter, so delaying the flash as much as possible can eliminate wasteful flash processing, and it is also desirable to delay the flash. If so, the deregistration will be delayed accordingly, so the hit rate is expected to improve. For example, if a write request is made for one sector in a cluster and then another sector in the same cluster is write requested, a write hit occurs. However, if there are many such write hits, it is possible to delay the flush as much as possible. Flashes corresponding to a plurality of write requests can be collectively executed by one flash. Further, when blocking is necessary, blocking can be executed collectively. As a result, the flash processing becomes efficient.

Regarding the flashed data,
By deleting the registration, the area of the cache address Cid can be used for new cache writing thereafter, but no cache hit occurs for that data. Therefore, for data having a high read hit rate, the cache hit rate can be improved by delaying registration deletion as much as possible. Therefore, the deregistration process is performed in a procedure substantially similar to the flush process shown in FIG. 24, as shown in FIG.

FIG. 25 shows the registration deletion processing by the cache task. First, in step F401,
The variable n is set to "1", and in step F402, the nth priority data for deregistration is searched. That is, first, the data set as the "first level (priority)" is searched. In this case, the flash information and the registration deletion order information in the cache management data are confirmed, and the cache address Cid in which the flash information is flushed and the registration deletion order information is "first level (priority)" is extracted. I will do it. And step F403
Then, the one or more cache addresses Cid thus searched are deregistered. That is, the content of the cache management information for the cache address Cid is cleared.

In step F404, it is determined whether or not the necessary amount of registration deletion has been completed. The deregistration is a process of making the cache address Cid a newly usable area, and the size of the area to be deregistered may be determined according to the cache situation at that time. . Therefore, all of the flushed cache address Cid may be deregistered, or a part of the flushed cache address Cid may be deregistered. In step F404, the registration deletion processing determines whether or not the area size to be deleted this time has been reached.

If deregistration of the required amount is not completed,
In step F405, it is judged whether or not there is any unregistered cache address Cid of the above-mentioned searched "first level (priority)". If there is any, the process returns to step F403 and the registration is similarly deleted. To do. If it is determined in step F405 that the registration deletion has been completed for all the data of the retrieved cache addresses Cid, the process proceeds from step F405 to F406.
The variable n is incremented and the process returns to step F402.

When the variable n is incremented in step F406 and the variable n = 2 is set and the process returns to step F302, the second priority data, that is, flushed and the registration deletion priority information is "second level (normal)". The data (cache address Cid) which is said to be "is searched.
Then, the registration of the cache address Cid retrieved in step F403 is deleted.

If the deregistration of all the data of the "second level (normal)" is completed at the stage where the deregistration of the required amount is not completed and the deregistration is still continued, the variable n is set in step F406. Increment and step F
Return to 402. In this case, the variable n = 3, and in step F402, the data having the third priority, that is, the data that has been flushed and the registration deletion order information is “third level (non-priority)” (cache address) Ci
Search d). Then, the registration of the cache address Cid retrieved in step F403 is deleted.

At the time when the necessary amount of registration deletion is completed, the registration deletion process is ended from step F405.

According to such registration deletion processing, the registration deletion is delayed for data having a high read hit rate,
That is, it is stored in the cache memory 3 for a relatively long time. Therefore, it is possible to increase the probability of a cache hit, improve cache responsiveness, and
It is possible to reduce the need for the disk read operation.

The flash process shown in FIG. 24 or the registration deletion process shown in FIG. 25 may be combined with the process shown in FIG. 19 described in the flash process example <1>. For example, steps F202, F205, and F2 in FIG.
When the flush order is set by the processing of FIG. 24 when the flush is executed at 08, the flush is delayed as the data having a higher hit rate.
For example, in step F205, the non-consecutive data having a smaller hit rate is flushed, and the one having a high hit rate is left for a relatively long time. As a result, the responsiveness effect of the processing of FIG. 19 and the hit rate improvement effect of the processing of FIG. 24 can be obtained.

The embodiments of the storage device, the recording and / or reproduction device, the information storage system, and the storage method according to the present invention having the configuration of the recording / reproduction device 1 and the personal computer 100 have been described above. Storage device, recording and / or reproducing device, information storage system,
The specific processing procedure of the storage method is not limited to the above example,
Various modifications are possible.

For example, as the flush processing, when the cache area is insufficient due to a read request immediately after writing to the cache and the cache area has to be flushed, the data size requested to be read is used as a reference, or slightly the same size. It is possible to flush only a large amount of data. That is, immediately after writing to the cache memory 3 and when most of the cache area is in the registered state, flush is performed for data cache in response to a read request and used for reading data from the disk 90. It is necessary to secure a possible cache area, but at that time, only the cache area having the same size as the read data size or a slightly larger size is flushed. By doing so, it is possible to secure the cache area for the read data from the disk 90, minimize the flush process, and improve the response speed of the cache process to the read request.

The program of the present invention is a program for executing the operation of the above-described storage method in a storage device, a recording and / or reproducing device, and an information storage system. And / or a reproducing device and an information storage system can be realized. Furthermore, according to the recording medium of the present invention in which the program of the present invention is recorded, it is easy to provide the program for realizing the present invention, and it is suitable for device design and system construction. The recording medium for recording the program is C
It can be realized by a D type, DVD type, MD type optical disc, a magneto-optical disc, a magnetic disc such as a flexible disc, an HDD (hard disk drive), a memory card using a solid-state memory, and the like.

[0176]

As can be understood from the above description, according to the present invention, the information temporarily stored in the temporary storage means is read according to the attribute of the information stored in the cache memory (temporary storage means). Operation of writing to the recording medium,
That is, the flash operation is controlled. In particular, continuous data,
The flash processing is preferentially performed from the data having a short flash processing time such as non-continuous data and data requiring blocking. By such control, it becomes possible to increase the free area in which data can be written in the cache memory at an early stage. Therefore, the cache response speed can be improved. In addition, since continuous data will be collectively flashed, efficient write operations can be realized on recording media such as discs, and the number of accesses to discs and the like can be reduced at the same time. It is possible to improve the operation efficiency and save electricity. In addition, the cache hit ratio is improved by preferentially performing the flush process on the data having the low hit state value and delaying the flush process for the data having the high hit state value. Realize the improvement of. Controlling the flash process according to the continuity of data and the hit status in this way can be suitable as control for a cache memory in which PC data and AV data are mixed. That is, AV with high continuity
Efficient flash processing that takes advantage of the characteristics of the data and P
Flush processing that maintains the probability of a cache hit for C data is compatible.

Further, according to the program of the present invention or the recording medium recording the program, a storage device, an information storage system, which performs cache control for realizing the above-mentioned effects,
The recording and / or reproducing apparatus can be easily realized.

[Brief description of drawings]

FIG. 1 is a block diagram of a recording / reproducing apparatus according to an embodiment of the present invention.

FIG. 2 is an explanatory diagram of a disk format according to the embodiment.

FIG. 3 is an explanatory diagram of an area structure of the disc according to the embodiment.

FIG. 4 is an explanatory diagram of a disc management structure according to the embodiment.

FIG. 5 is a block diagram of a storage unit of the recording / reproducing apparatus of the embodiment.

FIG. 6 is an explanatory diagram of a cache control method according to the embodiment.

FIG. 7 is an explanatory diagram of a cache ID according to the embodiment.

FIG. 8 is an explanatory diagram of a data write operation to a cache according to the embodiment.

FIG. 9 is an explanatory diagram of a data write operation to the cache according to the embodiment.

FIG. 10 is an explanatory diagram of a data write operation to the cache according to the embodiment.

FIG. 11 is an explanatory diagram of a data write operation to the cache according to the embodiment.

FIG. 12 is an explanatory diagram of cache management data according to the embodiment.

FIG. 13 is an explanatory diagram of an operation at the time of disc writing according to the embodiment.

FIG. 14 is an explanatory diagram of an operation when reading a disc according to the embodiment.

FIG. 15 is an explanatory diagram of an operation in the case of a cache hit at the time of disk read according to the embodiment.

FIG. 16 is an explanatory diagram of cache search processing according to the embodiment.

FIG. 17 is a flowchart of cache search processing according to the embodiment.

FIG. 18 is an explanatory diagram of a division unit of cache control according to the embodiment.

FIG. 19 is a flowchart of flash processing according to the embodiment.

FIG. 20 is an explanatory diagram of blocking processing according to the embodiment.

FIG. 21 is an explanatory diagram of state transition by the flash operation according to the embodiment.

FIG. 22 is an explanatory diagram of state transition due to a normal flash operation.

FIG. 23 is an explanatory diagram of another cache management data content according to the embodiment.

FIG. 24 is a flowchart of a flash processing procedure according to the embodiment.

FIG. 25 is a flowchart of a registration deletion processing procedure according to the embodiment.

[Explanation of symbols]

1 recording / reproducing apparatus, 2 storage section, 3 cache memory, 4 USB interface, 5 input / output processing section, 6 display section, 7 operation section, 8 system controller, 9 ROM, 10 RAM, 11 cache management memory, 12 NV-RAM, 100 Personal Computer / Network

Continuation of front page (51) Int.Cl. ⁷ Identification code FI theme code (reference) G11B 20/10 301 G11B 20/10 301Z H04N 5/85 H04N 5/85 Z 5/91 5/91 Z (72) Invention Person Takai Motoyuki 6-35 Kita-Shinagawa, Shinagawa-ku, Tokyo F-term in Sony Corporation (reference) 5B005 JJ11 MM11 NN02 NN15 PP03 QQ04 VV04 5B065 BA01 CE12 CH02 5C052 AA03 AB04 CC11 CC12 DD06 5C053 FA23 GA01 GB06 GB07 GB11 GB15 JA01 JA07 KA01 LA11 LA14 5D044 AB01 AB05 AB07 BC06 CC06 DE12 DE49 EF03 FG10 HH07

Claims (30)

[Claims] 1. A temporary storage means for temporarily storing information read from a recording medium and / or information to be written to the recording medium, and attribute determination means for determining an attribute of the information temporarily stored in the temporary storage means, A storage device, comprising: a control unit that controls to read out information temporarily stored in the temporary storage unit and write the information in the recording medium according to a determination result by the attribute determination unit. 2. The attribute discriminating means, regarding the information temporarily stored in the temporary storing means, continuous information continuously written in the recording medium and discontinuous in the recording medium in comparison with the continuous information. 2. The storage device according to claim 1, wherein the control unit determines that the continuous information is written in step S <b> 3, and controls the sequential information to be preferentially read and written in the recording medium. 3. The attribute discriminating means discriminates, from the information temporarily stored in the temporary storage means, blocking necessary information that requires blocking when writing to the recording medium, and the control means makes the blocking necessary information. The storage device according to claim 1, wherein information other than the above is controlled to be preferentially read and written in the recording medium. 4. The recording means performs a blocking process of forming a predetermined write data unit using the information read from the recording medium for the blocking necessary information, and then the recording medium. The storage device according to claim 3, wherein the storage device is controlled to write to the storage device. 5. The attribute discriminating means discriminates a hit situation with respect to the information temporarily stored in the temporary storage means, and the control means preferentially reads out information having a low value indicating the hit situation. The storage device according to claim 1, wherein the storage device is controlled to be written in the recording medium. 6. A writing / reading means for writing and / or reading information to / from a recording medium, and information read from the recording medium by the writing / reading means and / or the writing / reading means. The temporary storage means for temporarily storing the information to be written to the recording medium, the attribute determination means for determining the attribute of the information temporarily stored in the temporary storage means, and the determination result by the attribute determination means. A recording and / or reproducing apparatus, comprising: a control unit that controls to read information temporarily stored in the temporary storage unit and write the information in the recording medium. 7. The attribute discriminating means is, for the information temporarily stored in the temporary storage means, continuous information continuously written in the recording medium and discontinuous in the recording medium in comparison with the continuous information. Discriminating the non-contiguous information written by the control means, the control means controls the consecutive information to be preferentially read from the temporary storage means and written to the recording medium by the writing / reading means. The recording and / or reproducing apparatus according to claim 6. 8. The attribute discriminating means discriminates, from the information temporarily stored in the temporary storage means, blocking necessary information that requires blocking when writing to the recording medium, and the control means makes the blocking necessary information. 7. The recording and / or recording according to claim 6, wherein the information other than is controlled to be preferentially read from the temporary storage means and written to the recording medium by the writing / reading means.
Or a playback device. 9. The control means causes the writing / reading means to execute a blocking process for forming a predetermined write data unit using the information read from the recording medium by the writing / reading means. 9. The recording and / or reproducing apparatus according to claim 8, wherein the writing / reading means controls the writing / reading means to write on the recording medium. 10. The attribute discriminating means discriminates a hit situation with respect to the information temporarily stored in the temporary storage means, and the control means preferentially gives the temporary information from the information having a low value indicating the hit situation. 7. The recording and / or reproducing apparatus according to claim 6, wherein the recording and / or reproducing apparatus is controlled to read from the storage means and to write to the recording medium by the writing / reading means. 11. Information and / or information read from a recording medium
Alternatively, temporary storage means for temporarily storing information to be written in the recording medium, attribute determination means for determining the attribute of the information temporarily stored in the temporary storage means, and the temporary storage means according to the determination result by the attribute determination means. Control means for controlling the information stored in the storage means to be read out and written in the recording medium, and an address indicating an access position to the recording medium for writing and / or reading the information in the recording medium. An information storage system comprising: a writing / reading request means for notifying the control means. 12. The attribute discriminating means, regarding the information temporarily stored in the temporary storing means, continuous information continuously written in the recording medium and discontinuous in the recording medium in comparison with the continuous information. 12. The information storage system according to claim 11, wherein the control means determines that the continuous information is written in step S4, and controls the continuous information to be preferentially read and written in the recording medium. 13. The attribute discriminating means discriminates, from the information temporarily stored in the temporary storage means, blocking necessary information that requires blocking when writing to the recording medium, and the control means makes the blocking necessary information. The information storage system according to claim 11, wherein information other than the above is controlled to be preferentially read and written in the recording medium. 14. The control means performs a blocking process for forming a predetermined write data unit using the information read from the recording medium for the blocking necessary information, and then the recording medium. The information storage system according to claim 13, wherein the information storage system is controlled so that the information is written in 15. The attribute discriminating means discriminates a hit situation with respect to the information temporarily stored in the temporary storage means, and the control means preferentially reads out information having a low value indicating the hit situation. The information storage system according to claim 11, wherein the information storage system is controlled so as to be written in the recording medium. 16. When the information read from the recording medium and / or the information to be written to the recording medium is temporarily stored in the temporary storage means, the attribute of the information temporarily stored in the temporary storage means is determined, and the A storage method, characterized in that, in accordance with a result of the determination, the information temporarily stored in the temporary storage means is read out and written in the recording medium. 17. Regarding information temporarily stored in said temporary storage means, continuous information continuously written in said recording medium, and discontinuous information written discontinuously in said recording medium as compared with said continuous information. The control is performed such that the continuous information is read out preferentially and written to the recording medium.
6. The storage method according to item 6. 18. With respect to the information temporarily stored in the temporary storage means, the blocking necessary information that requires blocking when writing to the recording medium is discriminated, and the information other than the blocking necessary information is read out preferentially. The storage method according to claim 16, wherein the storage method is controlled so as to write to a recording medium. 19. The blocking necessary information is controlled to be written in the recording medium after executing a blocking process for forming a predetermined write data unit using the information read from the recording medium. The storage method according to claim 18, further comprising: 20. A hit state is determined for information temporarily stored in the temporary storage means, and control is performed so that information having a lower value indicating the hit state is preferentially read and written in the recording medium. The storage method according to claim 16, wherein the storage method is a storage medium. 21. When the information read from the recording medium and / or the information to be written to the recording medium is temporarily stored in the temporary storage means, the attribute of the information temporarily stored in the temporary storage means is determined, and the A program for executing a process of controlling to read the information temporarily stored in the temporary storage unit and write the information in the recording medium according to the determination result. 22. Regarding information temporarily stored in the temporary storage means, continuous information continuously written to the recording medium, and discontinuous information written discontinuously to the recording medium as compared with the continuous information. 3. The control is performed so that the continuous information is read out preferentially and written to the recording medium.
The program according to 1. 23. With respect to the information temporarily stored in the temporary storage means, the blocking necessary information that requires blocking when writing to the recording medium is discriminated, and the information other than the blocking necessary information is read out preferentially. 22. The program according to claim 21, which is controlled so as to be written in a recording medium. 24. The blocking necessary information is controlled to be written in the recording medium after executing a blocking process for forming a predetermined write data unit using the information read from the recording medium. 24. The program according to claim 23, wherein 25. It is determined that the hit status is determined for the information temporarily stored in the temporary storage means, and the information having a lower value indicating the hit status is read out preferentially and written in the recording medium. 22. The program according to claim 21, characterized in that 26. When the information read from the recording medium and / or the information to be written to the recording medium is temporarily stored in the temporary storage means, the attribute of the information temporarily stored in the temporary storage means is determined, and the A recording medium having a program recorded thereon for executing a process of reading information temporarily stored in the temporary storage unit and controlling the writing of the information in the recording medium according to a determination result. 27. The program, for information temporarily stored in the temporary storage means, continuous information continuously written in the recording medium, and discontinuously written in the recording medium compared to the continuous information. The non-continuous information to be recorded is discriminated, and the continuous information is controlled to be preferentially read and written in the recording medium.
The recording medium according to 6. 28. The program determines blocking necessary information that requires blocking when writing to the recording medium for information temporarily stored in the temporary storage means, and gives priority to information other than the blocking necessary information. 27. The recording medium according to claim 26, wherein the recording medium is controlled to be read out and written to the recording medium. 29. The program causes the recording medium to execute a blocking process of forming a predetermined write data unit using the information read from the recording medium for the blocking necessary information. 29. The recording medium according to claim 28, wherein the recording medium is controlled to be written. 30. The program discriminates a hit situation with respect to the information temporarily stored in the temporary storage means, and preferentially reads and writes in the recording medium from the information having a lower value indicating the hit situation. The recording medium according to claim 26, which is controlled.