JPH0318928A - System load processing method and disk controller - Google Patents

Abstract

PURPOSE:To load an operating system (OS) included in plural system resident volume (SYSRES) areas of a secondary storage without using a load processor by obtaining the start and end addresses in a corresponding disk based on the OS name. CONSTITUTION:A system load program 112 registers the divided SYSRESs stored in a SYSRES store position storage part 113 of a disk device 101, the register names (OS name and divided disk identifier) of divided disks, and the start and end addresses in a disk into a disk controller 103. Then a load program 112 retrieves the store position of the SYSRES area containing a designated OS by the controller 103 and inputs an initial program load (IPL) record 111 of the designated OS into a main storage 102 from the SYSRES area to perform the IPL simulation. Thus it is possible to load an optional OS through plural SYSRES areas in the disk without using any system load processor.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a computer that provides multiple types of operating systems. The present invention relates to a system loading method for selectively loading and a disk control device.

[Conventional technology]

Publicly known example closest to the invention: JP-A-63-116254 "System load processing method": Document 1: JP-A-1983-116254
Publication No. 62944 [Disk Addressing Method]: Document 2 According to the method of Document 1, it is possible to store multiple OSs with different storage formats in the same secondary storage device, and it is possible to store multiple OSs with different storage formats in the same secondary storage device. Different OSs can be provided on the next storage device. Furthermore, this method selectively loads the OS into the main storage device using a dedicated system load processing device.

On the other hand, Document 2 discloses a split disk method in which a disk volume is divided into a plurality of parts and a plurality of programs use them completely independently. In this conventional technology, in a virtual computer system in which multiple OSes are operated simultaneously on the same computer system, the overhead caused by collectively managing divided disks using a virtual machine control program (VMCP) is reduced by address translation using a disk controller. This paper discloses an efficient disk addressing method for virtual machine systems that reduces the number of disk addresses by performing the following steps.

[Problem to be solved by the invention]

However, the above conventional technology has the following problems.

The formats and loading methods of multiple OSs provided on one computer system are all stored in advance in the system load processing device, or the virtual machine control program (VMCP) does not intervene on the virtual computer system. , - It is not possible to load multiple systems from one disk device. Therefore, a new OS storage area (S
YSRES area) to the secondary storage device or change the loading means, the system load processing device must be rebuilt.

Further, the storage format information of the secondary storage device is stored only in the system load processing device. Therefore, unless there is a synergistic effect between the system load processing unit and the instruction processing unit,
You can use the secondary storage device on the actual computer other than when waiting for system load processing (for example, when executing a job), or when using one OS.
cannot use different divided storage areas (divided disks) on the device.

An object of the present invention is to load one OS in a plurality of 5YSRES areas in a secondary storage device without using a load processing device on an actual computer in order to solve the above-mentioned problems of the conventional method. It is in. Another purpose is to enable one OS on a real computer to use a plurality of divided disks defined in the secondary storage device and other secondary storage devices. As a result, multiple systems can be operated with a small number of disk devices within the limited computing center space.

[Means to solve the problem]

In order to achieve the above object, the present invention limits the secondary storage device to a disk device. a command check circuit for determining whether data has an address in the disk; a means for determining a real address in the disk device from an address in the disk specified by the OS and the starting address in the disk; 5YS) Determine whether the address is inside the IEs using the final address in the disk, and if the address is not inside the 5YSRES area, end the input/output process with the disk device with an error, and if the address is inside the 5YSRES area, the disk device and means for performing input/output processing from the corresponding address.

On the other hand, within the disk device, there is a storage unit containing a set of a 5YSHES area storage location and an OS name corresponding to each of the plurality of OSs, a storage location of a divided disk area having general data, and a set of its divided disk identifier, and One of the OS
and a first initial program load (IPL) record for loading the system load program. This ^ stem load program includes a first step of registering, in the disk control device, the storage positions of the plurality of 5YSRES areas and the divided disk areas in the disk device, the corresponding OS name and the divided disk identifier;
The OS name specified from among the multiple OS names above and the OS name
a second step of inputting the identifier of the divided disk area used by S and the virtual device number into the disk controller, and registering the virtual device number in the entry in which the identifier is registered in the disk controller; The third step is to store a correspondence table between the number and the actual device number on the main storage device, and to have the disk controller select the storage location in the disk of the 5YSRES area of the specified OS to create an IPL record of the OS. is input into the main memory and the IP
and a fourth step of performing L simulation and switching to the control state of the OS.

When the OS issues an input/output command by specifying a device number (which is a real device number for the OS, but is actually a virtual device number of a partitioned disk), the instruction processing device stores the above virtual device number on the main storage device. It has means for determining whether a device number is stored, and if it is, converting it into a corresponding real device number, and issuing an input/output command having the virtual device number as an operand to the input/output processing device. The disk control device selected by the input/output processing device has means for performing address translation within the disk using disk partition information registered by the system load program when loading the OS.

[Effect]

The first IPL record, system load program, SYSRES contents of multiple OS, and these 5YSREs
Specify the large capacity disk device number that stored the S area storage location and perform IPL. The first IPL record deploys the system load program onto main storage. This system load program loads multiple 5Y files from the disk device.
Read the OS name, split disk storage location, and split disk identifier that correspond to the SRIES area storage location, and register them in the disk controller. Next, write the specified identifier and virtual device number to the disk controller to be loaded from now on. A correspondence table between the virtual device number and the real device number is also stored on the main storage device. Next, the system load program causes the disk controller to search for the storage location of the S'1SHES area that has the specified OS, and reads the specified OS from this 5YSRES area.
The IPL record of S is input to the main storage and an IPL simulation is performed.

This ends the control of the system load program, and the designated OS starts controlling the computer system.

When the OS that started control issues an input/output command.

The instruction processing device compares the device number specified by the OS with the virtual device number stored on the main storage device. If they match,
Since this device is a split disk, the real disk device number is obtained from the correspondence table on the main storage device, and this is used to issue the input/output command to the input/output processing device. At this time, the virtual device number is also passed as a command operand to the input/output processing device, the input/output processing device selects the relevant disk control device, and sends the virtual device number together with the real device number to the disk control device. The disk control device selected by the input/output processing device determines whether the input/output command has an in-disk address in its data, and if it does, it determines the start address of the divided disk from the virtual device number registration entry. Select the disk address in the data,
This divided disk start address is used as a reference to convert it into a real disk address. In addition, in order to prohibit access to another divided disk area, the disk control device performs an intra-disk address check using the final address of the divided disk, and if it is outside this divided disk area, replaces the input/output data with invalid data. Returns an input/output processing error.

As a result of the above, multiple 550p files on a large capacity foot disk can be
It becomes possible to load any OS from the YSRES area and use multiple divided disk areas on that OS.

〔Example〕

The present invention will be explained in detail below. FIG. 1 is a block diagram showing a load processing circuit of a computer system in an embodiment. 101 is a disk device, 102 is a main storage device, 103 is a disk control device, 104 is a system console, 105 is an instruction processing device, and 106 is an input/output processing device. The disk device 101 includes a first IPL record 111 and a system load program 112 SYSR.
There is an ES storage location storage section 113. The main storage device 102 has a correspondence table 122 between virtual device numbers and real device numbers.

The disk control device 103 includes a disk partition information registration section 131. Input/output command check circuit 132. Adder 1
33. Disk address check circuit 134. There is an error handling circuit 135.

A method for loading an OS performed in a computer system having the above-mentioned components will be described in detail below using the flowchart shown in FIG. When the disk device 101 storing 5YSRES is specified and IPL is performed, the first IPL record 111 is input from the specific address of the disk device to the system program area 121 of the main storage device 102 (step 201). According to the output command, the system load program 112 is loaded from the disk device 101 into the system program section 1.
21 (step 2o2). The system load program 112 uses the 5YSRE of the disk device 101.
The divided 5YS stored in the S storage location storage unit 113
RH3 and the registered name (OS name and divided disk identifier) of the divided disk, and the start address and end address within the disk are stored in the disk division information 9. in the disk control device 103. The system load program 113 registers the information in the green section 131 (step 203) and repeats this until all disk partition information is registered (step 204). Next, the system load program 113 registers the OS name of the OS that will be IPLed from now on and the information that this OS will use. The partitioned disk identifier and virtual device number are input from the system console 104 (step 205). It is determined whether the input data includes a virtual device number (step 206). If the virtual device number is input, the disk controller 103 The corresponding partitioned disk identifier in the disk partition information axis registration unit 131 is replaced with this virtual device number (step 207), and the virtual device number and real device number are added to the virtual device number and real device number correspondence table 122 of the main storage device 102. Storing the Device Number (Step 2OS) Step 207 described above will be explained using FIG. When the disc division information registration unit 131' DIS
This means that the entry 301 called K(C) is rewritten to the device number 441 (entry 301').

If the virtual device number is not input in step 206, the system load program 112 determines that this input information is the OS name of the system to be booted from now on, and stores the corresponding entry 302 in the disk partition information registration section 131' as the disk device. 1101 is rewritten to the actual device number (440) (step 209), and then the 5YSRES area start address (α) 303 of this entry 302' is
The arm of the disk device 101 is moved to the address α in the disk (step 210). There is an LPL record of the specified OS at the address α in the disk, and this is stored in the system program section 12 of the main storage 102.
1 (step 211), the system load program performs an IPL simulation using the IPL record, the process ends in the system load program 112, and the system is controlled by the OS with the specified OS name (step 211). 212), with the above, multiple 5YS
System load processing can be performed on a real computer from a disk device having a RES area, etc. In addition, when multiple users divide and share the same disk device, it is possible to add a password to each divided disk area or to limit the OS that can be used for each divided disk for security protection. .

Next, the operations of the instruction processing unit 105 and the disk control unit 103 when the OS loaded in the above embodiment uses the divided disk area on the real computer will be described using the flowchart shown in FIG. When this OS issues an input/output command, the instruction processing unit [(105) writes the O
Checks whether there is a virtual device number that matches the device number 151 specified by S (step 401). If it is determined in step 402 that it is stored in the main storage device 102, the corresponding real device number is selected. (Step 403), Next, issue an input/output command that has the virtual device number as an operand and specifies the real device number to the input/output processing device 106, and selects the disk control device 103 (Step 404). Then, the disk control device 1103 selected as the input/output processing device 106 inputs the input/output command to the command check circuit 132, and checks whether this command has an address in the disk as data (step 405). If the data does not contain an address, the disk control device 1103 performs input/output processing from the disk device 101 as before, but if the data contains a disk address, the virtual device number is input to the partition information registration unit 131 (step 406). , select the entry that matches this virtual device number (step 4
07), for example, division information registration section 131#4 in FIG.
41 is selected, the adder 133
In step 4O, add the disk address and the corresponding divided disk area start address (γ) 304 in the corresponding entry.
S) The address check circuit 134 compares the sum of in-disk addresses with the final address (δ) 305 of the divided disk area in the entry (step 409). If it is determined that there is one (step 410), the error processing circuit 13
Control is transferred to step 5, and the sum of the above-mentioned addresses in the disk is converted into an invalid address (step 411).

If you access the disk using this invalid address,
An input/output error occurs. On the other hand, if it is determined in step 410 that it is within the range of the corresponding divided disk area, input/output processing of the disk device 101 is performed using the sum of the above-mentioned intra-disk addresses.

Note that the first IPL record 111, system load program 112, and divided disk storage location storage unit 113 must be stored in a disk device other than the system IPL volume 11 of the virtual machine system to which the divided disk area is allocated on the virtual machine. For example, the above method makes it possible to use the divided disk area even on a real computer.

In this way, according to this embodiment, the disk device that stores the 5YSRES areas and divided disk areas of multiple OSes is used not only on the virtual computer but also on the real computer, without using a system load processing device. becomes possible.

〔Effect of the invention〕

According to the present invention, the contents of the system resident volume area of multiple OSes in any format are stored in one disk device, and one of the OSs is loaded on an actual computer without using a system load processing device. It becomes possible to do so. Furthermore, it becomes possible to use a plurality of split disks on a real computer, which used to make one disk device look like multiple disk devices through simulation of the virtual computer control program. Therefore, multiple systems can be operated with a small number of disk devices within a limited calculation center space.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing a load processing circuit of a computer system according to an embodiment of the present invention, FIG. 2 is a flowchart showing a system load processing procedure according to an embodiment of the present invention, and FIG.
FIG. 4 is an explanatory diagram of a disk division information registration unit in a disk control device according to an embodiment of the present invention, and FIG. 4 is a flowchart showing an input/output processing procedure within the disk control device according to an embodiment of the present invention. 101... Disk device, 102... Main storage device,
103...Disk control device, 104...System console, 105...Instruction processing device, 106...
Input/output processing unit, 111...first IPL record. 112...System load program, 113...
Disk division position storage unit, 131...Disk division information registration unit, 132...I/O command check circuit,
133...Adder, 134...Disk address check circuit, 135...Error processing circuit. η 2nd day Hokuni

Claims (1)

[Claims] 1. System resident volumes (SYSRES) of multiple operating systems (OS) in the same disk device
) and can run multiple types of OS, register disk partition information, which is a correspondence table between the OS name, the start address in the disk, and the end address, and calculate the start address in the disk corresponding to the OS name. a command check circuit that determines whether the input/output command has an address in the disk as data; and a means for determining the actual address in the disk device from the address in the disk designated by the OS and the corresponding start address in the disk. determining means for determining whether the in-disk address obtained by the above-mentioned means is an address within the SYSRES area from the above-mentioned final address in the disk;
A disk control device characterized by having means for causing an error in input/output processing with the disk device if the address is not in YSRES, and performing input/output processing from the disk device if the address is in the SYSRES. 2. In the computer system described in paragraph 1, a storage section for the OS name corresponding to the SYSRES content storage location of the plurality of OSes in the disk device, and a system load for loading one of the plurality of OSs. program, and a first initial program load (IPL) record for loading the system load program, and the system load program stores the SYSRES of the plurality of OSs in the disk controller described in item 1 above. A first step of registering the content storage location and OS name, a second step of inputting the OS name of one of the plurality of OSs to the disk control device, and inputting the IPL record of the OS from the disk device. and the said I
A system load processing method, comprising: performing an IPL simulation of the OS using the PL record, and taking over system control to the OS. 3. In the computer system according to item 2 above, which has a plurality of divided disk areas different from the SYSRES area, and a storage unit for storing the divided disk area identifier and the divided disk storage location in the disk device, the first step includes:
The second step also includes registering the split disk area identifier and the split disk storage location in the disk controller, and the second step includes registering the split disk identifier and virtual device number of the split disk area used by the OS along with the OS name. input, register the virtual device number in the entry of the corresponding partitioned disk identifier in the disk control device, and also store a correspondence table of the virtual device number and real device number on the main storage device. How to handle system load. 4. In the computer system described in item 3 above, when the OS loaded by the system load processing method described in item 3 issues an input/output instruction, the instruction processing device uses the virtual device number in the above correspondence table on the main storage device. Determines whether the OS stores the specified device number, and if it does, the device is a split disk, and the actual device number corresponding to the input/output command that has the virtual device number as an operand is determined. It has a means for issuing a message to an input/output processing device to select a disk controller of a disk device having the divided disk area, and the selected disk controller registers the disk partition information registration unit using the virtual device number. 1. A disk control device accessing a disk device by means of searching for a divided disk area from a disk, and the means described in item 1 above.