 |
Description  |
|
|
BACKGROUND OF THE INVENTION
This invention relates to data storage devices and more particularly to optical disk data storage devices. Even more particularly, the invention relates to an apparatus and method for restricting the writing of information on optical disk media.
An optical disk is a data storage medium which is readable by a lasern-based reading device. Optical disks, known as "compact disks" or "CD's", have become increasingly popular during the last few years for recording music and audio-visual
works. Due to the huge storage capacity of optical disks as compared to conventional magnetic storage media, optical disks known as "ROM disks" have become popular for storing computer readable information. Recent technology has produced optical disks
which can be written as well as read by the computer, thus, in the future optical disks are expected to become increasingly more important in the computer industry and may eventually replace magnetically readable and writable storage media such as
"floppy disks" and "hard disks". Optical disks of the type used in computer applications are generally mounted in cartridges, and the reading devices generally read or write data through a slot provided on a surface of the cartridge.
One type of optical disk is often called "WORM" disks for Write-Once-Read-Many. WORM media is a type that can be written but cannot be erased, therefore, it can be written only once. If an attempt is made to write on this media a second or
subsequent times, the new data is written over the old data, resulting in garbled data which is unintelligible. A significant need exists for WORM media, however. This type of media is very useful for archive storage of data, wherein the data is
intended to be written only once and never erased. A very significant aspect of WORM media, is that it always leaves an audit trail. Since it can only be written once, the data in a sector will always be the original data written in the sector. If a
rewrite of the sector was attempted, it will be obvious from the garbled data and the CRC check data, which will be incorrect. In no case can the data in a sector be changed without leaving evidence of the change.
Recently a new type of optical disk media has been developed, called magneto-optic media or "MO" media. This type of media can be written, erased, and rewritten many times, in the same manner as magnetic media. Magneto-optic media is very
important as a direct replacement for magnetic media, since it performs the same functions and allows a much larger storage capacity on a given disk.
Both WORM and magneto-optic media have important applications, however, it is very difficult and costly to design a single disk drive that can process both types of media.
Another aspect is that WORM media is implemented in as many combinations of technology and formats as there are manufacturers. Hence, interchangability between systems is not possible. This lack of standardization in the WORM field is in
striking contrast to the firmly established standard in the MO field, where there is full interchangability between manufacturers.
There is a need then for a single media which can function either as WORM media, or as rewritable media. Also, there is need for a WORM implementation in the well standardized MO technology. Various features and components of such a media are
disclosed in U.S. patent application Ser. No. 07/426,834, filed Oct. 25, 1989, for MULTI-FUNCTION OPTICAL DISK DRIVE AND MEDIA of Hoyle L. Curtis and Terry Lynn Loseke, which is hereby specifically incorporated by reference for all that is disclosed
therein now abandoned.
A similar problem has been addressed with magnetic media. Floppy disks for example, have a write protect notch which, if covered, prevents writing on the media. Therefore, a device can write on the media while the notch is uncovered, and then
the notch can be covered to make the media "read-only". Magnetic tape has solved this problem in a similar way with a write ring, or in the case of tape cartridges with a record slide switch. These methods all suffer from the same drawback, that is,
the mechanism is very easily reversed to make the media writable once again. In many operating systems, for example DOS on personal computers, a file can be marked as read-only after it has been initially written. Again, however, the read-only status
is very temporary and can easily be reversed with another operating system command. Because the mechanism is easily reversed in all these cases, there is no audit trail.
This problem has also been partially addressed in a device, manufactured by Canon, Inc., called the "Canofile 250". This device allows an entire disk to be formatted as write-once. This device would appear to have two serious drawbacks,
however. The process of formatting a disk is usually done by an operating system within the computer, which means that the write-once format is known only to that operating system, and other operating systems, not knowing of this format, might write
over the media, leaving no audit trail. Another problem with this device is that the write-once status only applies to the entire media, therefore the media cannot be divided into write-once and rewritable portions.
Another problem that exists in converting rewritable media to write-once media is that drives that were manufactured prior to the design of the convertible media will not understand that the media is convertible and may write on the media even
though it has been marked as read-only. This problem is similar to the above described problem wherein the operating system simply sets a bit identifying a file as read-only. Because the previously manufactured drive is unaware of the mechanism by
which the media is converted to read-only, it will ignore such mechanism and write on the media. Therefore, one of the most difficult problems being faced in providing a convertible media, is the problem of designing a mechanism which will be recognized
by previously manufactured drives.
It is very desirable, then, to provide a system that will allow media to be converted from rewritable media to write-once media. Because the nature of its construction, WORM media is incapable of being written to more than once. Magneto-optic
media, however, can be rewritten many times. Therefore, there is need in the art to provide apparatus and method for allowing magneto-optic media to be converted to write-once media. There is a further need in the art for allowing portions of the media
to be converted to read-only, while retaining other portions of the media as rewritable. A still further need in the art is to allow each sector of the magneto-optic media to be independently converted to read-only.
SUMMARY OF THE INVENTION
It is an object of the present invention to provide an apparatus and method of restricting the writing of data on optical media.
It is another object of the present invention to provide apparatus and methods for defining writable media as read-only media.
Still another object of the invention is to provide apparatus and methods for defining portions of writable media as read-only media.
Yet another object of the invention is to provide apparatus and methods for dynamically changing a portion of the media from writable to read-only.
A further object of the invention is to provide apparatus and methods to define writable media wherein each data sector of such media can be redefined from writable to read-only.
A still further object of the invention is to provide a data structure, stored on the media, that defines which sectors of the media have been written and which sectors are available for writing.
Still another object of the invention is to provide apparatus and methods for erased sector management on rewritable media.
The above and other objects of the invention are accomplished in a system that stores information identifying which sectors have been written and which sectors are available to be written. This information is stored in a separate write
management directory on the media. When the media is first loaded into the drive, this write management directory is loaded from the media into RAM within the disk drive controller. Thereafter, whenever a write command is received by the drive, the
drive checks the directory to determine if the sector has already been written. If the directory indicates that the sector has already been written, the drive returns an error. Otherwise, the drive writes the sector and updates the write management
directory.
In the preferred embodiment, two separate methods are used to indicate that a sector has been written. The write management directory contains pairs of pointers, each pair defining a contiguously written area of the media. Also, each sector
within the media contains information, typically a flag bit, that indicates whether the sector has been previously written. When a write command is received by the drive, it first checks the write management directory and determines whether the address
of the sector to be written lies within any of the contiguous areas defined by the pointers. If the sector does lie within any of the contiguous areas, the drive returns an error since the sector has been previously written. If the sector is outside
the contiguous areas, the drive determines whether a new set of pointers needs to be created for a new contiguous area. If there is room in the write management directory, a new set of pointers is created. If the directory has become full, the use of
the directory is discontinued, and the flag bit within the sectors is then used. In this manner, the write management directory size can be limited while still providing coverage for all sectors on the disk.
In another embodiment, the write management directory pointers are used, however, sufficient space in the directory is allowed for the worst case size of the directory. Since the maximum directory size is allocated on the media, there is no need
for a flag bit in each sector.
In yet another embodiment, the write management directory contains a bit for each sector on the media. If the sector has been written, the bit is set to a logical one, otherwise, the bit is set to logical zero. When the drive receives a write
command for a sector, it finds the bit corresponding to the sector and examines the bit to determine whether the sector has been previously written.
In still another embodiment, the disk is logically divided into zones, and a pointer is kept for each zone. The pointer points to the next sector in the zone that may be written, thus, when the drive receives a write command for a sector, if the
sector number is greater than or equal to the pointer value, the sector may be written.
BRIEF DESCRIPTION OF THE DRAWINGS
The above and other objects, features, and advantages of the invention will be better understood by reading the following more particular description of the invention, presented in conjunction with the following drawings, wherein:
FIG. 1 is a block diagram of a computer system that incorporates the present invention;
FIG. 2 is a diagram of an optical disk for use with the present invention;
FIG. 2A shows an expanded view of the data storage area 202 of FIG. 2.
FIG. 3 shows a flowchart of a write operation in the present invention;
FIG. 4 shows a flowchart of determining sector status using a bit mapped directory;
FIG. 5 shows a flowchart of determining sector status using a pointers directory;
FIG. 6 shows a flowchart of determining sector status using the directory of the preferred embodiment; and
FIG. 7 shows a flowchart of determining sector status using the zone and pointer method.
DESCRIPTION OF THE PREFERRED EMBODIMENT
The following description is of the best presently contemplated mode of carrying out the present invention. This description is not to be taken in a limiting sense but is made merely for the purpose of describing the general principles of the
invention. The scope of the invention should be determined by referencing the appended claims.
In general, the invention comprises a system that allows rewritable media to be converted to write-once media or to manage erased sectors on rewritable media. The system performs this function by storing information identifying which sectors of
the media have been written and which sectors are available to be written. This information is stored in a separate write management directory on the media. When the media is first loaded into the drive, the write management directory is loaded from
the media into RAM within the disk drive controller. Thereafter, whenever a write command is received by the drive, it checks the directory to determine if the sector has already been written. If the directory indicates that the sector has already been
written, the drive returns an error. Otherwise, the drive writes the sector and updates the write management directory.
In a first embodiment, the write management directory contains a bit for each sector on the media. If the sector has been written, the bit is set to a logical one, otherwise, the bit is set to logical zero. When the drive receives a write
command for a sector, it finds the bit corresponding to the sector and examines the bit to determine whether the sector has been previously written.
In a second embodiment, the write management directory contains pairs of pointers, each pair defining a contiguously written area of the media. Sufficient space is provided in the write management directory for the worst case size of the
directory.
In a third, preferred, embodiment, two separate methods are used to indicate that a sector has been written. The write management directory contains pairs of pointers, each pair defining a contiguously written area of the media, as in the second
embodiment. Also, each sector within the media contains information, typically a flag bit, that indicates whether the sector has been previously written. When a write command is received by the drive, it first checks the write management directory and
determines whether the address of the sector to be written lies within any of the contiguous areas defined by the pointers. If the sector does lie within any of the contiguous areas, the drive returns an error since the sector has been previously
written. If the sector is outside the contiguous areas, the drive determines whether a new set of pointers needs to be created for a new contiguous area. If there is room in the write management directory, a new set of pointers is created. If the
write management directory has become full, its use is discontinued, and the flag bit within the sectors is then used. In the manner, the write management directory size can be limited while still providing coverage for all sectors on the disk.
A primary advantage of the preferred embodiment, is write management information redundancy. If either the write management directory or the flag bits are lost, they can be recovered by referencing the other information. That is, should the
write management directory become unusable, the flag bits of all sectors can be scanned and a new directory created. Furthermore, if the flag bits become unusable, the directory can be used to recreate the flag bits for all sectors on the disk unless
the directory has become full. Also, in the normal case, this type of media will be written sequentially, so that the write management directory will contain only a pair of pointers or at most a few pairs of pointers. Another advantage of the preferred
embodiment is that write operations are performed in a single pass until the write management directory becomes full.
In a fourth embodiment, the disk is logically divided into zones, and a pointer is kept for each zone. The pointer points to the next sector in the zone that may be written, thus, when the drive receives a write command for a sector, if the
sector number is greater than or equal to the pointer value, the sector may be written.
FIG. 1 shows a block diagram of a computer system that incorporates the present invention. Referring now to FIG. 1, a computer system 100 is shown having a processing element 102. Data is transferred between the processing element 102 and all
other parts of the system by a system bus 104. Attached to the system bus 104 is a keyboard 106 which allows a user to input data to the computer system 100. Also attached to the system bus 104 is a display 108 which allows the computer system 100 to
display data to the user. A main memory 110 is attached to the system 104 and is used to store data and programs. Stored in the main memory 110 is an operating system 112 and user software 114. Also attached to the system bus 104 is the optical data
storage device 120 of the present invention. Within the optical data storage device 120 is an interface 122 which connects the optical data storage device 120 to the system bus 104. Attached to the interface 122 is the drive controller 124 which
contains all the electronics and firmware for controlling the optical drive 126.
FIG. 2 is a diagram of an optical disk storage media suitable for use with the present invention. Referring now to FIG. 2, disk 200 contains a data storage area 202 used to store user data in sectors. The write management directory of the
present invention is a part of the data storage area 202. Outside the data storage area 202, and located toward the center of the disk 200, is a control track 204 which contains a media descriptor table which contains an indicator that the media is
magneto-optical with write-once-read-many (WORM) capability. A center hole 206 is used to center the media on the spindle of the optical drive 126 (FIG. 1). The disk 200 is typically surrounded by a cartridge (not shown) to protect the disk during
storage. The data storage area 202 and the control track 204 are accessible through a slot in the cartridge.
FIG. 2A shows an expanded view of the data storage area 202. Referring now to FIG. 2A, the data storage area 202 contains a write management directory 208, and a plurality of data sectors, for example data sectors 230 and 232. The write
management directory 208, which can be contained within data sectors located anywhere within the data storage area 202, may contain a plurality of bits, for example bits 209, for use with the first embodiment described above.
Alternatively, the write management directory 208 may contain a plurality of pairs of pointers, for example pointer pairs 210 and 212, for use with the second and third embodiments described above.
Each of the pointer pairs contains a beginning pointer, for example pointers 214 and 218, which points to the first sector of a contiguous area, and each of the pointer pairs contains an end pointer, for example pointers 216 and 220, which points
to the last sector of a contiguous area.
Each of the sectors, for example sectors 230 and 232, is organized in the same manner as a typical sector in a magnetic disk, wherein the sector contains a preamble area, for example areas 234 and 240, and a data area, for example data areas 238
and 244. Within the preamble area, each sector contains a flag bit, for example 236 and 242.
FIG. 3 shows a flowchart of a write operation in an optical drive of the present invention. Referring now to FIG. 3, after entry, block 302 checks the MO flag within the media to determine whether this is conventional magneto-optic media. If
the media is conventional magneto-optic, block 302 transfers to block 310 to simply write the sector. If the media is not conventional MO media, block 302 transfers to block 304 which determines whether the WORM MO flag has been set. The WORM MO flag
identifies this as magneto-optic memory capable of being re-formatted into write-once-read-many (WORM) media. If the WORM MO flag is not set, block 304 returns to the caller since this media cannot be used by the drive. If the WORM MO flag is set,
block 304 transfers to block 306 which calls either FIG. 4, FIG. 5, or FIG. 6 to get the state of the sector being written. In the preferred embodiment, block 306 will call FIG. 6. When using the bit mapped directory of the first embodiment, block 306
will call FIG. 4. When using the pointers only directory of the second embodiment, block 306 will call FIG. 5.
After determining the sector state, block 308 determines whether the sector is blank, and if the sector is not blank, block 308 returns to the caller since the sector cannot be written without overwriting pre-existing data. If the sector is
blank, block 308 transfers to block 310 which writes the sector and then returns to the caller. The write in block 310 could be either a conventional write or a write with verify.
FIG. 4 shows a flowchart of the process of determining the sector status using the bit mapped directory of the first embodiment. Referring now to FIG. 4, after entry, bit 402 calculates the bit in the write management directory which corresponds
to the sector being written. Ordinarily, when the media is inserted into the drive, the directory will be loaded into memory within the drive controller. Therefore, the bit will be contained in drive controller memory and this calculation simply
involves addressing the correct byte and bit of the directory within the memory. If the directory is not located in memory, the correct byte containing the bit corresponding to this sector would first be read from the drive. After calculating the bit,
the bit is retrieved and block 404 determines whether the bit is set. If the bit is set, block 404 transfers to block 406 which returns an error, because the sector is already written. If the bit is not set, block 404 transfers to block 408 which sets
the bit and then block 410 returns an indication that the sector is blank.
FIG. 5 shows a flowchart of the process of determining sector status in the second embodiment wherein the directory contains pointers to the sectors that have been written. In this embodiment, the pointers are stored in pairs, with the first
pointer of the pair pointing to the beginning of a contiguously written area, and the second pointer pointing to the end of a contiguously written area. Referring now to FIG. 5, after entry, block 502 gets the first or next pair of pointers from the
write management directory. Ordinarily, when the media is inserted into the drive, the directory will be loaded into RAM within the controller. Therefore, obtaining the next pointer pair simply involves obtaining them from RAM. After getting the next
pair of pointers, block 504 determines whether the address of the sector being written is contained within the range of the two pointers. If the sector address is between these two pointers, the sector has been previously written, and block 504
transfers to block 520 which returns an overwrite error. If the sector is not within this pair of pointers, block 504 transfers to block 506 which determines whether the directory contains additional pairs of pointers. If the directory does contain
additional pointers, block 506 transfers back to block 502 to check the next pair. If no more pointers exist, then the sector can be written so block 506 transfers to block 508.
Block 508 determines whether the address of the sector being written is adjacent to an existing pointer. If the address is adjacent an existing pointer, block 508 transfers to block 514 which simply updates this existing pointer and then
transfers to block 518 which returns a blank sector indication to the caller. If the address is not adjacent an existing pointer, block 508 transfers to block 510. Block 510 determines whether more pointer areas are available in the directory. If
additional pointer areas are available in the directory, block 510 transfers to block 512 which creates a new pair o | | |
