JPS6014342A - Information processing system - Google Patents

Abstract

PURPOSE:To perform input/output operation and information processing in parallel with each other by providing an operating system to control both a channel table and a task table, and reading out an interruption signal in response to the sequence recorded to a shared memory. CONSTITUTION:The replacement and the preservation of both a channel table and a task table are carried out by an operating system OS which functions under the control of an interruption program. This system OS includes a communication control program and other service programs. The system OS refers to said both tables or perform the replacement of processing as one its functions. In other words, the channel table is retrieved by checking a channel which is used by an item of the task table. Then it is needed to know other tasks which uses channels. If no task is using presently a terminal, said terminal is accessed by the relevant task.

Description

[Detailed Description of the Invention] Technical Field The present invention relates to an information processing system, and more particularly to an information processing system equipped with an operating system that can perform input/output operations and processing of multiple channels in parallel. be.

Prior art operating systems are created with the goal of serving the users of computers, and therefore need to achieve the goal of extracting the computing power possible from the computer and making it easier for the users to use it. operating·
The system has four minimum required functions: a supervisor, a loader, a language processing program, and an input/output control system. Among these, the supervisor is a program that manages the entire system and is the one most closely connected to the hardware, and is responsible for exchanging input/output control signals with channels, allocating main storage, and scheduling and executing jobs. Controls the start and stop of the system, allocates peripheral devices u, etc. The supervisor is required to perform real-time processing, and unless processing of an interrupt caused by input/output or the like occurs within a certain period of time after it occurs, it becomes impossible to distinguish it from the next interrupt.

In this manner, when an interrupt occurs when a certain operation of an input/output channel is completed, the operating system operates by an interrupt processing program that accepts and processes the interrupt.

Interrupt requests to the operating system are accepted when instructions of programs other than the interrupt processing program are not being executed, but are not accepted and ignored when instructions of the interrupt processing program are being executed.

Therefore, even if processing capacity is available, only one input/output can be executed at a time. For this reason, a certain input/output channel becomes exclusive to a specific task, and for example, if batch processing is performed using tasks issued by a remote terminal, etc., there is a risk that the remote terminal, etc. becomes unusable. .

Thus, conventional operating systems do not utilize each device of the system efficiently.

Therefore, the present inventor has developed an information system in which a central processing unit connected to a time division multiplex line is provided with means for generating an interrupt signal at regular intervals so that none of the tasks to be multiplexed can be ignored. proposed a processing method (patent application 1982)
No.-21871 (filed on February 1, 1982) (see specification). In this method, the actual encoding of transmission frames using algebraic laws is also described.

Purpose The purpose of the present invention is to improve such conventional problems, to make it possible to perform input/output operations and processing of multiple channels in parallel, and to enable input/output devices to be used by multiple users. Also provides an information processing system with an operating system that does not mix up data between users.

Configuration The configuration of the present invention will be explained below using an example of actual fabric. 'As an example of the information processing method of the present invention, consider a case where a computer is installed in one of the edge devices in a straw in a loop transmission line connecting a plurality of terminal devices. and 1
．． An operating system is configured so that more of the program-dependent functions of the calculator are transferred to functions based on electronic circuits, and the computer and its m-side devices can be used more effectively.

In an information processing system, the central resource [C that uses a computer and its peripheral devices] is the operating system.
It is necessary to improve the functionality of the system software. This operating system is operated by a central computer.When many terminals are logically connected through lines, the information processing system becomes complicated and has a large amount of overhead.One reason for the complexity is that the This is due to inappropriate classification. Also, the division of functions between so-called hardware such as electronic circuits and operating systems is inappropriate. Inappropriate classification can be seen in both hardware and software.

The present invention pursues more appropriate application of information processing systems by correcting inappropriate classification and assignment. First, regarding hardware classification and division,
Optimization will be carried out using the following methods.

An example of a transmission system to which the present invention is applied is as follows.

The code format flowing over the transmission link consists of repeating transmission frames of fixed length, each frame consisting of code words that follow the algebraic laws described in the above-mentioned application. Each terminal becomes a node in the transmission system.

Alternatively, the computers may be connected in a loop as shown in FIG.

In FIG. 1, 'r1. ...l T l ...T
M is a terminal device, and S is an information processing system. α shown in FIG. 1 is a symbol representing the above-mentioned loop. Each of the two Norr stations share one transmission frame format. The 1" U information symbol field of a transmission frame is comprised of multiple parts, creating multiple channels for use as described below.

Terminal T1 in FIG. 17 is a general terminal station, and its transmission processing function is shown.

In FIG. 2, R is a receiving section of a line terminator that receives signals from an upper station. C is a master lock source, and the fundamental frequency can be automatically adjusted by voltage control &J. T.I.
This is a sample value data processing system that detects a signal approximately proportional to the timing deviation of the received bit clock from the scrambling baseband signal received by the system. The sampling clock is a bit clock, and a voltage is applied to the oscillation frequency control terminal of C, which is a mask clock source, to control the clock phase in a direction such that the output timing information becomes "O1". PR is operated by the bit clock output from C, and its input is the output of R descrampled by DS, and it checks whether it is a code word that follows the algebraic law of encoding, and the result is Save. The output of PR in FIG. 2 is 1. This is a code word whose pronunciation has been corrected according to the L test result, and is input into the SR located next to the PR. The SR consists of a register that records the corrected code word and a clock circuit that controls its contents, and this clock circuit is reset when the algebraic code word is found to be in accordance with the law. The transmission frame is synchronized by this.

The SR in FIG. 2 provides a buffer between the transmission system and the terminal device, and the output of the SR forms the information symbol portion of the transmission frame transmitted from the terminal to the "lower station". Among the information symbols input from PR to SR1, the transmission frame to be received by this terminal device is determined by the designation from the clock circuit, and this part is sent to the terminal device and at the same time the transmission frame is received by the terminal device. The transmission code from is input and updated. The above information symbol part outputted from SIL in this way has been partially updated by this terminal device, and is processed according to the above algebraic encoding law, and the result of this processing is It constitutes a code word and is transmitted from the terminal of the above-mentioned "lower station" as a transmission frame. Encoding of such a frame is performed by PS in FIG. 2, and the output of PS is S
It is scrambled by C and output to S.

In the present invention, a function code is provided at the beginning of a transmission frame, and by detecting this code in PR, the syndrome contents generated based on the test result of whether or not it is the above code word are transferred to another register. By making the above correction according to the contents of , and resetting the contents of this register, the code word in the PR of the next transmission frame can be received and inspection can be started immediately, so that the transmission efficiency does not decrease. be.

Each link of the loop network shown in FIG. 1 transmits the transmission frame created as described above in time series, and this transmission frame is divided into a plurality of fields as shown below.

In the second diagram, the symbol α refers to α in FIG. 1, and the transmission function circuit shown in FIG. 2 must be provided at each station on the α loop. In FIG. 1, terminal T1 is assumed to be a main station for synchronizing the operating clocks of each station, and a control station for allocating transmission channels. To
In this case, there is no wire (a) that controls the clock source C in FIG. synchronize. The above synchronization method is described in the above-mentioned application by the present inventor. The RD and SD in FIG. 2 transfer the contents of the SR to the RD and transmit information prepared in the SD to update the contents of the SR at the time when a transmission frame forming a codeword is generated in the SR. This is a register for transferring to SR.

The configuration of each I/register of IID and SD is the same in each station of "α loop", but 11□,...
.

The method of using RD and SD differs depending on the terminal device and information processing system in T..., TN, 31: and S.

First, in SR, the code 'NT! This section describes the structure of Field 1 in the transmission frame at the time when it is being formed. That is, it consists of a field constituting a control channel and a plurality of information fields for the purpose of transmitting information. The information processing system S is ]゛,...T...
It is jointly used by the TM, and S uses each of the above fields at the same time to multiplex tasks that use multiple information channels.

The terminals connected to these information channels are Go.

. . , T , . . . , TI, which corresponds to the number of information channels. The number of terminals is greater than the number of information channels, and each terminal selects one of the fields for the information channel in the It, D, and SD registers in FIG. 2 to communicate with S. become.

The fields for the RD and SD control channels in FIG. 2 are used in the same manner at each station on α except To. To transmits a start code at regular intervals from the control channel field of R1 and S, and
．． Examine the various signals from each station input from the D control channel. The start code interval τ is divided according to the number of stations in α-I-, and corresponds to each station in the name division α-1-. These stations count the divisions of their own station after detecting the start code, and if they need to communicate with S at this time, S
Record the calling code in the control channel field of the D register. At To, the presence or absence of a calling code sent to the control channel field of the RD register of the own station is detected by detecting the start code and finding the division corresponding to each station. Thereafter, T□ selects the information channel being set. Depending on the error, T1 sends a start code to the control channel field of the SD register, and when the number of divisions of that station is counted, a code indicating the information channel number or T1 indicates that the signal channel is not free. Record a connection code consisting of a code indicating . In the case of sending a calling code as described in (1,1)-F, each terminal station checks whether there is a connecting code before sending out a calling code, and connects it to S using the information channel specified for that code. connect. If the above code indicates busy, a meeting will be held. Each station on the α loop logically connects the data receiving section and transmitting section of each station by selecting the designated information channel from the fields of the RD and SD registers by detecting the connection code. By providing such hardware functions, the system that executes the connection procedure between the information processing system S in FIG. The operating system to be executed is responsible for updating the task table containing tasks developed by users of the information processing system S and the channel table containing input/output channels of S, and for logically distinguishing between tasks and channels. This creates one of the conditions that makes it possible. Connection between the loop network and the information processing system S is performed as follows.

0-2) The TB shown in FIG. 3 is T1. ..., T1. ....

Like T, etc., it is based on the same principle as R, D, SD and related parts shown in FIG. R.D.
The contents of 30 and 5D31 are divided into fields for a plurality of information channels and control channels, and are connected to input/output channels CHo and CH0-8, respectively. It is assumed that CHo is a control channel, and CH0 to CH8 are each ↑U information channels. Here, as an example, it is assumed that the number of information channels is three.

FIG. 3 shows the configuration of an information processing system that is shared, and CC 38 represents a central processing unit. Except for the contents of the program executed by the CC, there is no difference in the configuration depending on how the system S is applied. In the information system S, each of the above three information channels used by multiple terminals such as Tlf..., Tlf..., +TN, etc.
H,~CH8 are logically connected, and C in S
C performs multiple processing for questions or messages from these terminals and responses thereto. M35, FtE36 in Figure 3. R37 and the like are a type of external storage memory, and F stores programs or data. It is also possible that data consisting of image quality may be recorded. The RE records a fragment of the audio signal to send the audio response to the terminal. Saraji M35's memory (for speed conversion to enable input of images, images, or audio signals via P4J, converting them to low speed and transmitting them to the terminal. M's circuit and operation The principle is described in the above-mentioned application "Information Processing System" by the present inventor. However, in this application, the output of M is not directly connected to TB, but is connected to BUS like other f-devices. It is necessary to effectively utilize each block shown in FIG. 5, related software, and each area of the work memory of the CC. This is done by optimizing the

In other words, each block or line shown in Figure 81￥3 operates based on commands issued by the function of the program executed by the CC, but these operations are not related to the CC. - Executes in parallel without interruption, and communicates to CC by interrupt signal only when completion is complete. The CC has a work memory, and uses this memory area to process the operation results of each block or line described in -FU. This work memory area and C in Figure 3
Transfer with the CC reference block is performed by the input/output function of the CC, and execution of program instructions is performed by the command execution function of the CC.The work memory is shared by the above input/output function and the command execution function. The information transfer between each block shown in FIG. 3 is performed in a common memory T1.ES, and each block in FIG.
The right to access is in the form of time-sharing slots given to each block by the MPX. Slots for a CC are divided into its instruction execution function and input/output function, and each slot allows each function to access the work memory. The address area of the common memory RES is divided into a plurality of partial address areas, each partial address area having 6 partial addresses with a constant address value, and the partial addresses 0.5) and 6 in the address area.・S is the recorded Il[4i
They are logically connected so that they are read out in sequence, and by applying the above partial address, reading and recording are performed in accordance with the above connected order. The partial address for the CC is divided into two parts, one of which is assigned to the CC's instruction execution function and the other to the CC's input/output function.

In FIG. 3, BLJS is a plurality of M, RE.

It is occupied alternately by CC, TB, and F in a time-sharing manner. That is, this time division slot is performed by the line indicated by 1 on the BUS. The line 1 is an address line of several bits, and slot allocation is performed by changing the logical value combination of each bit. It is an input to the RES of B U S, and is composed of the logical sum of the output lines from each block above.
S3 is a parallel output line from RES to the above block.

BUS raw is common memory RI', S address line,
It is composed of the logical sum of output lines from each of the above blocks. B
The lines 2.3.4 of US are (16) for each block addressed by 1 of BUS. exchange. In a slot assigned to an arbitrary block, the partial address of the code transfer destination unit is designated and recorded in the first half of the slot, and the partial address corresponding to the block is designated and read in the second half of the slot. If a partial address is specified as described above, reading is performed from that address area in the recorded p+α order. CC instruction execution] Function 4: When performing code transfer between two blocks, including CC input/output functions, the partial address of each block in RES is a control word that specifies code transfer to these blocks. to be recorded. Then, the subsequent operation is determined by reading the control word from the own partial address in the data slot immediately applied to the above two blocks. When this operation is completed, the partial address corresponding to the instruction execution function of the CC is recorded as an interrupt signal. Even when an interrupt request instruction is executed by the instruction execution function of the CC, the interrupt signal is recorded in its own partial address depending on the instruction execution function of the CC.CC
Execution (In reality, it increments the instruction counter and executes the instruction in the work memory specified by the instruction counter.
When the execution of the instruction is completed, the instruction counter is incremented (1'). The processing program jumps to the address that has not been recorded and performs the processing J11! according to the contents of the interrupt signal.While the side-load processing program is running, reading from its own partial address is not possible. However, recording to this partial address continues.The TfJi partial address area corresponding to the instruction execution function of the CC ensures a sufficient number of addresses. This prevents interrupt errors from being ignored, makes the interrupt processing program sufficiently large, and allows flexible multi-processing.

Furthermore, each time a function character is received from the terminal device j1, the character is recorded in the partial address corresponding to the input/output function of the CC, and at the same time, an interrupt signal is recorded in the partial address corresponding to the command execution function of the CC. According to this method, by performing transmission control on a character-by-character basis, even if the frequency of interrupts increases, these interrupts will not be ignored. This is advantageous when programming, etc., in which conversations are frequently transmitted, is performed using a high-speed TTG terminal. However, the command issued to the line is not in the form of complete 1- with 7;11 included, but with CC
It is necessary to ensure that the character is not lost by completing the command G newly issued from the command execution function of the command execution function. RES and MP in Figure 3
The hardware created by Become. Therefore, C
If the processing capacity of C is large, there is a possibility that a plurality of transfer channels between peripheral devices 1f may occur at the same time. Therefore, occurrences of Warikomi Haku' may also occur one after another. This is because, in the case of the present invention, the interrupt signal 09) is not lost because it is waited for at the partial address for the interrupt signal of the RES.

However, controlling multiple channels in this way requires multiple processing, and the operating system creates a channel table with items corresponding to these channels, specifies the control status of each channel, and It displays how a task is used and performs appropriate processing each time an interrupt occurs.

Multi-processing in an information processing system is performed by the function of an interrupt processing program, which manages a task table consisting of many items. Although a task corresponds to a channel in a time division multiplex circuit, it is not multiplexed in the form of periodically assigning slots as in the case of a channel.

The interrupt processing function reads the interrupt signal, updates this task cable depending on which item the interrupt signal relates to, and searches for an item in the task table for which no input/output instruction is being executed. This task table records the contents and numbers of the commands of the interrupted program, and the priority of the items is G20).Therefore, the commands of the interrupted program are executed instead of the instructions of the interrupt processing program. If you change the program, control will be transferred to that program.

In this way, the role of the interrupt processing program 971 cat
s The time during input/output is effectively used for other tasks, but as mentioned above, ``If control is handed over to another program, there is no way to manage it unless an interrupt signal is detected. , since it has nothing to do with the priority of the task items mentioned above, there is a possibility that tasks waiting for control will be ignored.To prevent this situation, send the !'+'J signal at regular intervals. It is necessary to provide a block that issues a signal, that is, RT shown in FIG.

In the present invention, the classification of control is made clear by creating a channel table and a task table, and the operating system determines the action to be taken in interrupt processing by updating the indexes of both. This allows for easy and flexible responses. It is possible to connect CC, RES and MPX shown in Fig. 3, but this internal configuration is described in the separate application 41 [Joint Use 1/11 Report Processing System No. 1] by the present inventor. That's exactly what I did.

File F in Figure 3 contains 1 of the α loop shown in Figure 1.
(1) A message sent from a terminal named Tj to a terminal named Tj. -1 (Special application 1986-49)
60, 1151 in 'S specification Otsuteru)! It can be transferred using the method described above. This type of transfer is not just an image signal, but may also involve the transfer of a message with encoded meaning. For example, when creating a record from T1 to F, a part of the operating system executed by CC (depending on the actual situation) searches for an empty area in F and transfers the message sent from r1 to this area. .

Same J: uni, 'r, -naru. If it becomes necessary for batch processing due to drawings performed by L and M using the software system of information processing system S, T
□Record the job sequence specified by the user in a partial area of F. The content number of these files F is a message that is batch processed at an appropriate time and transferred to T j , but depending on whether it is a batch processing job, the former performs transfer control, and the latter is controlled by the user. The process specified by is performed, and the result &j is transferred to T as in the former case. In order for Cc in Fig. 3 to be able to create such a task, the service program executed by CC must have the function of creating a task table, which includes items corresponding to the tasks created by the user at T1. make. The detailed records for this item include records of work steps such as file manipulation and control signal exchange with T Make sure that multiple processing is possible when control is returned. When the work steps of a certain task item are recorded in a file and the contents are to be transferred to the user of the terminal Tj, the processing priority of the task item is lowered and the message is transferred to the terminal T□. Wait until the conditions are met.

03) As another form of message transfer from T□ to Tj or 'r4 itself, in the above explanation, when the recording from 1'□ to file F is completed, the processing step of that task item is completed, and the message is transferred by Tj. It is also possible for the user of T to access the record recorded by T1 in file F whenever it is convenient for the user of T. Of course, at this time, a new item corresponding to the work initiated by user Tj is created in the task table.

In the following, the configuration of the communication 1g1lIj and the communication processing have been explained, but it is necessary to clarify the communication control procedure in connection with this processing.

The detailed record of the task table described above is in the form of a kind of finite automaton, and the communication control procedure is also a finite automaton. This is a table consisting of a finite number of states and events that cause transition relationships between them, and the states and transition relationships from the established state of the transmission link to the reset state will be explained using diagrams.

FIG. 4 is a diagram showing the state that can be transmitted to the terminal device, ie, T1 in FIG. The states from 0 to 5 indicate the transition [L (1) when conversational communication is performed by the terminal device 1. The states in which the terminal device is communicating, that is,
] to 5, transmission frame control 411
It is necessary to send a code indicating that communication is in progress at the allotted time to this terminal device through a control channel created in the field.

As shown in FIG. 4, when communication is terminated or interrupted from the terminal side, return to the reset state O. At the same time, it stops the communication display sent from the control channel, stops the control at the control station To in FIG. 1, and stops the information processing device S from interrupting the communication display. In the state 1 in FIG. 4, the called station name is sent from the control channel together with the start code for the terminal to make a call.
That is, in this case, the information processing device S is called and T in FIG.
Station 0 transmits a call code, a calling station name, and a vacant information channel name from the control channel to the information processing device BZ S at the time assigned to dt1, and at the same time transmits one call code, the name of the calling station, and one vacant information channel name at the time assigned to the T1 terminal of the control channel. - Send the vacant information channel name. T□ receives a signal indicating that the connection has been completed from the information processing device through this channel and enters the state of 2 transmission (t il), and from then on 3.4
, Z5 and the transition line between them.

When the terminal is not used, and when the information processing device WS transfers a message to another terminal using the conversation communication described in ``2'', or when waiting for a response from the information processing bag ms, turn on the absent key. This will result in an absentee reset state. During this time, there is an incoming call at the time allotted to this terminal using the control channel created in the control field of the transmission frame, so the terminal prepares for a 7-absence response, and the terminal receives 1!, which will be described later. Automatically turns on the source, sets the paper, and performs self-checking to determine whether reception is ready. Even if the absent key is not operated, if an incoming call is detected from the control channel, the state 7 is reached. In this case, power has already been turned on. As a result of the self-diagnosis, if reception is possible, the corresponding control code is received.1. If the reception is unsuccessful, the corresponding control code is transmitted and the state returns to the absence reset state of 6. At this time, in the information processing bag US in the center of FIG. 3, the value priority of the corresponding task is lowered and processing of other tasks is started.

In the center, i:ji monitors the time when a control response should be received from the terminal as described above, and if a receivable code is not received at this time, it is assumed that an unreceivable control code is received. Perform similar treatment.

State 8 in FIG. 41 is a state in which the terminal is receiving or waiting to receive. At this time, if the received block is good, a control code requesting the next block is sent and the state is set to 0, and if it is bad, a control code is sent to request retransmission of the previous block, and the state is 9. become. The reason why the waiting state is divided into two is to take into account that the broadcast and retransmission request signals may also be erroneous. Monitor the time at which the control code for blowout or 14th feed should be present, and if the above response is received at this time, (27) It's possible to use a method such as passing the It'd adjustment code, vll or ENQ code on the control channel. conduct. On the terminal side, l:
By receiving the NQ code, retransmission, blowout, and universal code are sent (fi) at times 9 and 10, respectively.

If block reception is started in states 9 and 10,
The reception state becomes 8. Stopping the communication display code means the end of communication. In terminals with little cooperativeness, unnecessary power may be turned off. In terminals with the state diagram shown in Figure 4, 0 and 6 The power is OFF in the state of
is in a state of In a processing unit that operates only when the power is on, the above three states are identified and handled. Power off
When F, these states are distinguished depending on whether the absent key is operated and whether the hook is OFF.
The power is turned on by receiving a call and switching to the modem side. The timer is operated by the voltage applied at this time, and when the timer expires, the state identification registers 1 and 7 are set. At this time, depending on the state of the absent key and the hook of the telephone, it is necessary to determine whether the answer is 7 or the signal state is 1.

Figure 4 shows the l-
] This specifies the specifications for the communication control program. However, on the middle mounting 1d side, multiple processing is performed. The subject of multi-processing is the communication channel, not the tasks mentioned above.

This communication channel is defined to correspond to a terminal, and not to correspond to a task. Channel tables and task tables are both essential. That is, when one terminal is used at 2 Å or more, two or more tasks may occur for each item in the channel table. For example, even if one person among multiple people using one terminal sends batch processing, if the other person is doing conversational communication, communication to the terminal will not be possible even if the batch processing result is completed. Can not. This kind of judgment is made on the central side by looking at the channel table and task table described above. The operating system functions to update and save the channel table and task table under the control of an interrupt processing program, which includes the above communication control program and other service programs. The communication control program performs J-like processing in which each item in the channel table satisfies the state diagram shown in FIG. 4, but other functions of the operating system refer to both the channel table and the task table, or perform processing updates. In other words, to find the channel to be used by the -y hundredth item in the task table, search the channel table.
[You need to know what other tasks utilize the above channels.] If none of these tasks is making use of this terminal - the terminal can be accessed by said task.

Each of the task table and channel table can be considered to be a three-dimensional table. A normal table is a two-dimensional table consisting of new items and column items, and in the channel table example, the two-dimensional table is /r as shown in Figure 5(a),
This is for each channel. To represent the five dimensions,
Figure 5 (1) and (b) are required, and in (b), the channel name is written in the row "p+th ■, and the name of the task currently not being executed is written in the column item ■. Records tasks that are related to this channel but are not used due to batch processing or other reasons. In the two-dimensional table (a), the state name is recorded in the new item ``■'', and the event person and the task being executed are recorded in ``■''. (2) indicates to which state the state moves next when an event j occurs in state 1. ■ indicates the current state, and the name of the task currently using this channel is recorded in this matrix element.

Effects As explained above, according to the present invention, input/output operations and processing of multiple channels can be performed in parallel, and the channel table and task table are managed by the operating system. , Even if the input/output device is used by multiple users, data will not be mixed up among the users.

[Brief explanation of the drawing]

Fig. 1 is a block diagram of a loop transmission system to which the present invention is applied, Fig. 2 is a functional block diagram of the terminal device shown in Fig. 1, and Fig. 3 is an information processing system showing an embodiment of the present invention. FIG. 744 is a state transition diagram defining the specifications of the operating system of the present invention, and FIG. 5 is an explanatory diagram of a log table (task table) showing an embodiment of the present invention. 1: Information processing system (S), 2., 3, 4: Terminal device (T, ~T, ~T,), 36: Voice response file (RE
), 37: File (g), 38: Central processing unit (
CC), 39: Multiplexer (MPX), 4(l
Shared memory (RES), 32: Periodic interrupt generation circuit (
RT). c52), Figure g4 Kaisho GO-14342 (11) Figure 5 (a)

Claims (1)

[Scope of Claims] (1) In an information processing system having a central processing unit, a peripheral device connected to the central processing unit, and a memory shared by the peripheral device, “1. The system is equipped with an operating system that manages a channel table that records task names, etc., and a task table that records the status, etc. of each task, and handles interrupt signals associated with each input/output in the order recorded in the shared memory. Therefore, an information processing method is characterized in that a plurality of inputs and outputs are each processed in parallel by reading them. e) A patent characterized in that the plurality of inputs and outputs are processed under the control of subchannels created by dividing a transmission frame created by algebraic encoding into a plurality of fields. An information processing method according to claim 1. (3) The information processing system according to claim 14 or 2, wherein in the peripheral device, one input and output device is simultaneously used for a plurality of tasks with different processing modes.