JPH0926853A - File partition extension method and device therefor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To guarantee interchangeability with an existing program at a binary level by connecting an external memory device to a bus interface device via an input/output bus and employing two offset methods in byte unit and block unit when addressing on the external memory device is performed. SOLUTION: A magnetic disk device 170 is connected to an I/O bus interface 150 via the I/O bus 160. Read/write from memory 110 as a main memory device is performed on the magnetic disk device, and offset to an access point is managed in byte unit and block unit independently. At this time, two offset methods in byte unit and block unit are employed when the addressing on the magnetic disk device 170 is performed. Thereby, the interchangeability with the existing program at the binary level is guaranteed. Concretely, access on the area of the disk device exceeding the maximum address value of a CPU is performed.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a file partition expanding method and device used in a computer system such as a workstation and a server, and more particularly to a file partition expanding method and a device having a file partition expanding function in a large capacity disk device.

[0002]

2. Description of the Related Art In recent years, the capacities of disk devices and array disk devices have increased dramatically. When a file partition is constructed on these secondary storage devices, the maximum capacity thereof is limited by the addressing width of the CPU. For example, the maximum address value in a 32-bit computer system is 4G-1 bytes (2 ³² -1 bytes)
Even if a single disk device having a capacity exceeding this maximum value was connected, the file partition could not be constructed exceeding this maximum value. As a software approach to solve this problem, it is conceivable to use a 64-bit compiler. This method handles 64-bit integer values by combining two 32-bit general registers. As a result, a new 64-bit variable value is declared from the program, and addressing up to 2 ⁶⁴ -1 bytes becomes possible.
Regarding this, for example, the GNU C compiler has already been put to practical use, and the source code has been released. In addition to this, there is also a method in which a new interface is prepared for the operating system and the library to make it independent from the conventional interface. Please note that
For example, Dave Shaver, Eric Schnoebelen, “An Im
plementation of Large Files for BSD UNI
X ”, USENIX Winter ConferenceProceedings,
pp. 61-68, Winter 1992.

[0003]

As described above, if a compiler is used, a file partition with a size of up to 2 ⁶⁴ -1 bytes can be easily constructed. However, in this method, to change the compiler,
There is a problem of losing compatibility with existing programs at the binary level, and it is necessary to prepare a library to maintain compatibility. The present invention has been made in view of the above circumstances, and an object of the present invention is to solve the above-mentioned problems in the conventional technique, guarantee compatibility with an existing program at a binary level, and further It is an object of the present invention to provide a file partition expansion method and device which can be realized by adding an interface capable of expanding the addressing range with a minimum change without changing the operating system.

[0004]

SUMMARY OF THE INVENTION The object of the present invention is to provide C
A PU, a main storage device, and at least one bus interface device, and an external storage device is connected to the bus interface device via an input / output bus, and the main storage device writes to or reads from the external storage device. In the computer system for performing the above, a file partition expansion method characterized by using both byte unit and block unit offset methods when addressing to the external storage device, or a CPU, a main storage device and at least one bus interface In the computer system that includes a device, an external storage device is connected to the bus interface device via an input / output bus, and the external storage device writes to or reads from the main storage device, the device name of the operation target And offset value to access point The external storage device has access control means for writing or reading data to or from the external storage device, and file management means for independently managing the offset to the access point in byte units or block units. This is achieved by a file partition expansion device characterized by referring to the file management means when accessing any of the access points. More specifically, in the seek step for addressing the entire file partition, in addition to the conventional offset designation method in units of bytes, a new offset designation method in units of blocks is added. In addition to byte designation, a mode that allows block designation is added to the argument of the seek step, and the file partition is extended to block size × (4G−1) bytes by accessing in block units. Further, since the file operation interface is not changed, the existing program interprets it as a byte-by-byte offset method, and performs addressing within 4 GB as in the conventional case. In order to realize these, the access control means that interprets the offset specification method from the program and manages the information necessary to access the file partition on the disk device, is specified by each offset specification method. A file management means for managing the offset value is provided.

[0005]

In the file partition expansion method according to the present invention, the file management means is obtained from the designated device name of the operation target. Furthermore, the offset value to the desired access point is obtained in block units and byte units, and is managed by the file management means. The access control means can obtain the attribute information of the file and the information necessary for the I / O operation from the file management means, and can expand the file partition on the disk device by this.

[0006]

Embodiments of the present invention will be described below in detail with reference to the drawings. FIG. 1 is an example of a system configuration for performing a file partition expansion method according to an embodiment of the present invention. The system according to the present embodiment has a CPU 100, a memory 110, a CRT display 120, a keyboard 130, a network interface 140 and an I / O bus interface 150 which are connected to a common data bus.
It is composed of I / O bus interface 150
To the magnetic disk device 17 via the I / O bus 160.
0 is connected.

The network interface 140 is connected to a network (LAN) 180, and a remote workstation connected to the network 180.
(WS) 190 and other remote access is possible. The access control program 200 is a program resident in the memory 110 of the computer system. The file management table 210 is held in the memory as a file management means. The access control program 200 refers to the above-mentioned file management table 210 to perform control when accessing a file partition. The specific configuration of the file management table 210 and the procedure for accessing the file partition will be described below.

FIG. 2 shows the above-mentioned file management table 21.
FIG. 9 is a diagram illustrating a configuration example of a zero. File management table 21
0 consists of a file attribute area and an I / O management area.
The file attribute area is an access type entry 300 for storing information indicating a unit of offset start position and offset size for a file, read / write.
The te flag 310 and the device name 320 are included. The I / O management area consists of memory address 330, transfer byte count 340, block offset area 350, and byte offset area 360. The block offset area 350 stores an offset value for each block, and the byte offset area 3
An offset value in bytes is stored in 60.

FIG. 3 shows the system according to this embodiment.
This is the overall flow when accessing a file partition. Each step corresponds to a command from the user or a function call from the program. When each step is called, the access control program 200 resident in the memory is activated to execute the control flow. (1)
Open step (see Figure 4 for details)
Allocate file management tables and file descriptors to the devices that make up the file partition.
In the seek step (2) (see FIG. 5 for details), the offset to the file partition is designated. In the read step (3) (see FIG. 6 for details), data is read from the file partition.

Similarly, in the write step (3) (see FIG. 7 for details), data is written in the file partition. (4) close step
In the details (see FIG. 8), the device closing process, that is, the file descriptor is released. The detailed flow of each step will be described below. Figure 4 is open
It is a control flow of a step. In step 41, a file management table is allocated. In step 42, the file is opened and a file descriptor (referred to as "fd") is allocated. In step 43, the file attribute information is registered in the file management table. In step 44, fd is returned as a return value.

FIG. 5 is a control flow of the seek step. In step 51, the file descriptor that is an argument
The file management table is obtained from (fd). Step 52
Then, it is determined from the mode flag which is an argument whether the block mode is designated. If the offset is designated in block units, the process proceeds to step 53. If the offset is designated in byte units, the process proceeds to step 54. In step 53, the offset value is registered in the offset entry for each block of the file management table, and the process proceeds to step 56. In step 54, the offset value in bytes is converted into blocks and registered in the offset entry in blocks of the file management table. In step 55,
After the conversion in step 54, if there is a fraction,
Register the fractional byte value in the byte offset entry of the file management table.

In step 56, if the offset is designated in block units, the offset value in block units is
If the offset is specified in bytes, the offset value in bytes is returned as a return value. Next, FIG.
Is a control flow of the read step. Step 6
In 1, the file management table is obtained from the file descriptor (fd) that is an argument. In step 62, the read destination address and the number of bytes to be read, which are arguments, are registered in the file management table. In step 63, the logical block number of the disk to be read is obtained from the offset value in block units and the offset value in byte units written in the file management table. In step 64, the data block is read from the above logical block number.

In step 65, the number of bytes read is added to the offset entry in byte units of the file management table. In step 66, after converting the byte unit offset value of the file management table into block units, it is added to the block unit offset value, and the remainder is registered in the byte unit offset entry. In step 67, it is determined whether all the data blocks to be read have been read. If there is a data block to be read, the process returns to step 63 and the process is repeated. Next, FIG. 7 is a control flow of the write step. In step 71, a file management table is obtained from the file descriptor (fd) which is an argument.

In step 72, the address of the data to be written and the number of bytes to be written, which are arguments, are registered in the file management table. In step 73, the logical block number of the write destination disk is obtained from the offset value in block units and the offset value in byte units written in the file management table. In step 74, the data block is written to the logical block number. In step 75, the number of written bytes is added to the offset entry in byte unit of the file management table. Step 76
Then, after converting the byte unit offset value of the file management table into block units, it is added to the block unit offset value, and the remainder is registered in the byte unit offset entry. In step 77, it is determined whether all the data blocks to be written have been written. If there is a data block to be written, the process returns to step 73 and the process is repeated.

FIG. 8 is a control flow of the close step. In step 81, the file descriptor that is an argument
Close the file (fd). In step 82,
Release the file management table. Next, based on an example of a specific program, the file management table 210
And changes in the position pointed to by the offset pointer on the file partition will be described with reference to FIGS. In the example shown in FIG. 9, the unit of block offset is 1 Kbyte. The block offset unit may be any other size as long as it is a fixed value.
First, in the (1) open step, one file management table is allocated. At this point, each block / byte offset entry in the file management table contains'
0'is written.

(2) Next, in the seek step, the offset pointer is moved to the desired position. Here, the offset pointer is moved to the position of 5000000 blocks (1K × 5000000 = 5 Gbytes) from the beginning by the offset designation method in block units. At this point, "5000000" is registered in the block offset entry in the file management table. (3) Further, in the seek step, the offset pointer is moved from the current position to the 500th byte by specifying the byte offset. At this point, '5000000' is registered in the block offset entry and '500' is registered in the byte offset entry of the file management table. (4) Here, by executing the read step,
1200 bytes of data will be read from the position of (5G + 500) bytes. When the reading is completed, each offset entry of the file management table is updated, and "5000000" is registered in the block offset entry and "1700" is registered in the byte offset entry.

(5) c with the file descriptor (fd) as an argument
By executing the lose step, each offset entry of the file management table is cleared and the file management table is released. According to the above-described embodiment, it is possible to obtain the effect that the user can access the offset of 4 Gbytes or more by the offset value of the block unit and the byte unit registered in the file management table. Further, since the file operation interface is the same as the conventional one, even if the offset method of two steps of block unit and byte unit is replaced, the effect that the modification of the application program can be used with a minimum is obtained. The existing application program is binary and can be operated in the same addressing range as the conventional one. It should be noted that the above embodiment is an example of the present invention, and it is needless to say that the present invention is not limited to this.

[0018]

As described above in detail, according to the present invention, compatibility with an existing program at the binary level is guaranteed, and the addressing range can be minimized without changing the library. It is possible to realize a file partition expanding method and device which can be realized only by adding an interface capable of expanding the file system to the operating system. More specifically, in the single disk device connected to the computer system, by adopting both the block unit and the byte unit offset designation method registered in the file management table, the area exceeding the maximum address value in the CPU can be stored. Access can be realized. Moreover, the user can use the application program without modification because the file operation interface is the same as the conventional one.

[Brief description of drawings]

FIG. 1 is a system configuration example for performing a file partition expansion method according to an embodiment of the present invention.

FIG. 2 is a diagram showing a configuration example of a file management table 210 according to the embodiment.

FIG. 3 is an overall flowchart of file partition access according to an embodiment.

FIG. 4 is a control flow diagram of an open step according to the embodiment.

FIG. 5 is a control flow chart of a seek step according to the embodiment.

FIG. 6 is a control flow diagram of a read step according to the embodiment.

FIG. 7 is a control flow diagram of a write step according to the embodiment.

FIG. 8 is a control flow chart of a close step according to the embodiment.

FIG. 9 is a diagram showing a change (part 1) in the file management table.

FIG. 10 is a diagram showing a change (2) in the file management table.

[Explanation of symbols]

 100 CPU 110 Memory 120 CRT Display 130 Keyboard 140 Network Interface 150 I / O Bus Interface 160 I / O Bus 170 Magnetic Disk Unit 180 Network 190 Workstation 200 Access Control Program 210 File Management Table 300 Access Type 310 read / write Flag 320 Device Name 330 Memory address 340 Number of transfer bytes 350 Block offset area 360 byte offset area

Claims (5)

[Claims] 1. A CPU, a main storage device, and at least one bus interface device, an external storage device is connected to the bus interface device via an input / output bus, and the main storage device is connected to the external storage device. A method for expanding a file partition, characterized in that in a computer system for writing or reading from a device, both a byte unit and a block unit offset method are used when addressing to the external storage device. 2. A CPU, a main storage device and at least one bus interface device, an external storage device is connected to the bus interface device via an input / output bus, and the main storage device is connected to the external storage device. In a computer system that writes or reads data from a device, an access control unit that obtains the device name of the operation target and an offset value to the access point, writes or reads the external storage device, and an offset to the access point A file partition extension characterized by having a file management means for independently managing in byte units and block units, and referring to the file management means when accessing an arbitrary access point on the external storage device. apparatus. 3. The file partition expansion device according to claim 2, wherein the access control means is an access control program stored in the main storage device. 4. The file partition expansion device according to claim 2, wherein a plurality of said file management means are provided so that they can be operated in parallel independently of a plurality of application programs. 5. The file partition expansion device according to claim 2, wherein the file management means is a file management table stored in the main storage device.