JPH05502528A - Virtual network architecture and loader - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed description of the invention] Virtual Enoto Work Architecture and Loader Main Team ■Background Acquired country Since this invention also relates to computer systems, it is particularly applicable to virtual networks. connected by a single virtual computer and through a virtual network It functions as either of two interconnected computers that send and receive information. pertains to a computer system that includes at least two computers that Ru.

l rice limb stop prolapse■ Many different computer systems are available today. but However, many of these systems lack basic components, as shown in Figure 1, for example. It is built around location. In general, the basic components of a computer are input devices. It is connected to the input device 12 via the input lotus 11, and to the output device 14 via the output lotus 13. It includes a central processing unit 10 (CPtJ) connected thereto. CPtJlo is It is connected to a memory device 17 via a bus 15,16.

CPU10 performs control and calculation functions. The input and output devices 12.14 are computer It is used for communication between the computer user and CPU10. The input device 12 is a CPU Provide information to l0. Common input devices are a keyboard and mouse.

The output device 14 displays information from the central processing unit 10. Common output devices are Virtual display on video display monitors, printers, e.g. plotters, etc. Contains steps. Input device 12 and output device 14 are often input/output (Ilo) called a unit.

Memory device 17 typically contains two general types of information: instructions and data. Ko A computer program is a sequence of instructions that enable a computer to perform a given function. It is a command. The data stored in the memory device 17 is executed by the CPU 10. Processed by CPU10 in accordance with instructions from the computer program.

The memory device 17 generally includes a large capacity memory 17A also called a secondary storage device. It includes an in-memory 17B. Main memory 17B is relatively high-speed memory. , typically having an access time of about 20 to 400 nanoseconds.

Access time is the time from when the cpuio requests data from memory to when the data is requested. This is the time until the data is given to CPU10.

The main memory 17B normally stores a small number of programs currently being executed by the CPU10. Stores at least some of the data needed by this program. For example, The large-capacity memory 17A, such as a disk or tape, stores programs, data or CPLl. Programs and data that are not needed by lo, or main memory 17B Stores programs and data that cannot be stored in main memory due to capacity constraints. do. Programs and/or data are stored in large-capacity memory under the control of CPU10. Since the data is read and written to the memory 17A, the large capacity memory is It has 10 units. The access time for large memory is typically in the tens of It is on the order of milliseconds.

The computer's operating system consists of the CPU10 (Figure 1), the main Used to control the operation of memory 17B, I10 unit 12.14.17A. It will be done. Additionally, the operating system uses the computer's user apps. interface applications and hardware. In this specification, hard Software is used to refer to the physical components of a computer system. Ru. A computer's operating system has a kernel. , this kernel (i) spawns a user process, as described below, and (i i) scheduling the execution of a user process; Provides system services to processes; (iv) hardware interrupt handling, exceptions; Perform processing. A computer's operating system is usually the computer's It is loaded into the main memory 17B at boot time.

Here, "boot" refers to when the computer is turned on or when the computer refers to a series of actions performed when

The definition of "user process" is the user's source code, i.e. the computer program. It requires an understanding of the successive steps of converting the Essential. Apps that run on a computer using a given operating system User source code for application programs, typically executable images. A binary file known as a page is compiled and linked. " 'user process' is this executable by the operating system's kernel This is the execution of a great image.

For example, Microsoft Corporation's MS-DOS operating system For computers using the system (MS-DO3), the operating To interface the system and hardware, the basic human output system ( BIOS), this BIOS is usually a computer glue-on system. It is stored in memory (ROM).

(MS-DO3 is a registered trademark of Microsoft Corporation.) B IOS is based on instructions from the operating system called commands. Drive the hardware.

In particular, the MS-DOS operating system uses a series of commands called interrupts. interfaces to the BIOS through a command. The processing is usually carried out on CPU10. It is started by the software that is running. In response to the interrupt, CPU 1 0 is the interrupt vector table interrupt service stored in the main memory 17B Get the address of the routine. Generally, interrupt service routines are stored in ROM Address of main memory 17B storing BIOS and interrupt service routine tells the operating system where it is stored. Therefore, main memory 17B typically contains user processes, operating systems, and CPUs. Interface the operating system with the hardware connected to the have the information to do so.

Using two or more computers similar to computer 20 in FIG. If so, this computer information must be distributed. independent combination Several methods have been developed to transfer information between monitor devices. the first one As shown in FIG. 2A, the method stores information on the computer 20-1, for example, on Or copy it to the tape (disk/tape) 25 and copy it to the disk/tape 25. Transfer information from disk/tape 25 to computer 20-2. This is a method of copying to 0-2. However, computers 20-1 and 20- 2 are not installed in close proximity, the disk/tape can be -1 to computer 20-2, the preferred method It cannot be called a law. Furthermore, a more important problem is the disk drive for computer 20-1. The disk/tape 25 may not be compatible with the computer 20-2. It is. For example, computer 20-1 has a 5 1/4 inch floppy disk. The computer 20-2 uses a 3 1/2 inch floppy disk. This is the case when using .

For this reason, computers 20-1.20-2 often have modems 26.27 and 26.27. (FIG. 2B). The computer 26 It is connected to the modem via the serial port of the computer 20-1. computer 20-2 is similarly coupled to a modem. Modems 26.27 are generally Connected via telephone line.

A modem converts the digital signals provided by a computer's serial port into Convert it to an analog signal and send this analog signal to the other modem via the telephone line. be transferred. The second modem converts the analog signal received over the telephone line into a digital convert it into a digital signal and feed it to the computer via the serial port. The modem is , transfer data at a baud rate of 9600 baud or less. The transmission speed is Limited by the quality of the wire that connects the modem. In general, data between modems is The transfer rate is determined according to the transmission state at the time of connection, and is not changed thereafter. subordinate So if the data is first transferred at a high transfer rate and then connected to the line , an error occurs in the data transfer, but the transfer rate is high regardless of the quality of the transferred data. The case will be filed quickly. Similarly, if the initial line quality is poor (determined to be a low transfer rate) And this transfer rate will not be changed later.

Error checking of transmitted information ensures that the transmitted information is the same as the received information. It is done depending on whether or not. − For each block of data transferred to the ship, sender The computer calculates the error code and sends it as part of the data block. be done. The receiving computer calculates an error code on the block of data. and compare the calculated error code with the received error code. error code If the codes do not match, the receiving computer sends an error message to the sending computer. the block of data is retransmitted.

Still another form of transfer is to connect the computer 20-1 with the cable 29. The program is connected to the computer 20-2 (FIG. 2C) and executed on the computer 20-1. a software program and the same software executed by computer 20-2; Data is linked via cable 29 by the software program. general In these operations, the connected systems, i.e. computers 2o-1, 20-2, the cable 29 and the two executed software programs are It was set up for the users of. In other words, input/output devices such as keyboards , any of which can operate similarly to a computer video display. but However, in most cases directly connected computers 20-1.20-2 The transfer of data between is performed by - users.

Therefore, the operation of one keyboard and video display becomes redundant and unnecessary. The key point. In any case, the current technology for connecting two computers is as follows. There is a possibility that Furthermore, depending on the format of computers 2o-1 and 20-2, Therefore, the specific software program linking between computers 20-1 and 20-2 is Program execution operations are restricted. Links between computers are linked to other users' Depending on the application, access may not be possible.

In view of the shortcomings in the above methods for transferring information between such computers, Computer networks were developed. computer network , formed by several computers 20-1 to 20-n as shown in FIG. It will be done. This network typically includes computers similar to computer 20 in FIG. computers are used and are combined into a common computer system called a server.

Server 21 is accessed via the network. The network 22 is It enables sharing of data files and I10 devices connected to the server 21.

The operating system for the network is Not only the operating system of the computer 20-1 to 20- n and additional software for coordinating operations with respect to the server's shared resources. It also includes a.

Traditionally, programs executable by MS-DO3 are user executable by the operating system depending on the file name of the program you type. Read from disk and loaded into memory. MS-DOS operating In a programming system, an executable program file is an extension of "EXE". have a child. This type of file is called an ``r.EXE file''. Execution file Other executable files in the file are executable files with the extension “,COM”. It is a program file. , EXE file and 1. The difference between C0M files is It is obvious to the trader.

, the process of loading an executable program stored as an EXE Allocate the AM area to the program, and set the program prefix and suffix (PSP). This includes setting up the program and reading the program from disk. Often 1, E Executable programs stored as Contains "relocation J processing" that is performed at time and execution time.In the relocation processing, M The EXEC function of S-DO3 creates a relocation table at the start of an executable file. Use to adjust various values, pre-divide addresses, and perform value corrections.

For example, in assembly language for the Intel 8088 microprocessor, In the code written below, there is one relocation.

, 1Iodel small mov ax, Data; data load segment mov ds, ax ;To ds mov ah, 4ch; 005 end; 4ch function int 21h n+ain endp end n+ain Relocation is performed by MS-DO3's operating system function E When the program is running, the Data (default data segment) (assembler codeword indicating the segment value of the component) is loaded into register AX. done at the time. In the generation of executable files, the assembler and linker The address of the program is not given at the time of execution. The EXEC function is An executable program in RAM is given an appropriate value in relation to the PSP. Start.

Therefore, the actual program code of the application is changed at the time of execution. .

The EXEC function is intended to work with RAM and cannot be loaded into RAM. The executable program loaded changes. However, the executable program If the program is run directly from ROM, the EXEC function or other relocation facility is function cannot operate properly because the executable program in ROM cannot be changed. I can't make one. Therefore, an executable program executed from ROM Regarding this, relocation is not possible. Therefore, change the executable program No operating system can directly execute programs in ROM. I can't. This drawback is that, for example, ROM cards can be used instead of floppy disks. Limit the operation of portable computers that use .

Main butt standing plate The virtual network management function of the present invention is different from that of the prior art in connecting computers. Eliminate these shortcomings and create a new computer-to-computer communication system by linking data. This makes it possible to

Furthermore, the virtual network management functions can be implemented by a set of commands, e.g. These commands access the interrupt in the example and the virtual network Virtual networks controlled by management functions can be used by any user application. It can also be accessed and used from other applications. Virtual node work management function , not only access by user applications, but also virtual network management. Any user application can be given two different sets of capabilities by its functions. available.

at least one computer linked by electronic means to exchange information between computers In a computer system including two computers, according to the principles of the present invention, A virtual network management function is installed on each computer. virtual net In the first mode, network management functions manage Used for exchanging information. The first computer user application is Send a message to a second computer via the virtual network management function The user application on the second computer can connect to the virtual network. Receive messages via the network management function and send reply messages to the primary computer. can be sent to a computer.

In operation in the second mode, all but one computer in the system The computer's virtual network management functions service requests from that single computer. To do so. Therefore, one computer user application can The resources of a server computer, such as information and peripherals, can be transferred to a virtual network. It can be accessed through the network management function. This computer system The form in which one computer controls and uses other computers is called virtual control. called a computer. A virtual computer is a server computer that Common for servicing requests from connected computers over a network This is fundamentally different from other methods. Additionally, it offers a wide range of virtual network management features. A user application executed on a computer system by a function , it is possible to provide flexibility and means of communication that were not previously possible. .

Virtual networks in addition to giving new capabilities to user applications Management functions are new provide a means to do so.

Virtual network management functions are the primary means of sending, receiving, and generating messages. It has an application layer as a means. The application layer It receives commands from the user application and sends them to the second means as a transport layer. generates the first message structure to be used. The transport layer receives and sends messages. ,Occur. Upon receiving a message from the application layer, the forwarding layer generates a second mensage structure called a ket. The transport layer is - or multiple packets for each message from the main layer. Each packet After forming the packet, the transport layer transfers the packet to the data link layer for sending and receiving information. and then send it to a third means. The data link layer connects the network via the communication link. ・Connect node to other data link layer of remote computer using predetermined transmission parameters. Send using

The data link layer only sends and receives information, and the data link layer of the remote computer When a packet is received by ), it is sent to the forwarding layer of the computer in question. . The transfer layer of the two computers is determined by e.g. the transfer rate and the data sent over the transfer line. Controls the size of packets sent.

The originating transport layer generates an error code for the information in the packet and adds this information to the packet. Attach error code. The receiving transfer layer records the error code of the received packet. Compare the error code that occurred and the error code received with the packet do. If the error codes differ, the receiving transport layer sends the error to the originating transport layer. - message and the packet is retransmitted. Add a predetermined amount to the transmitted packet. If a forwarding error occurs, the originating forwarding layer Adjust the speed or both.

Therefore, unlike traditional systems, the virtual network management function is The transfer parameters can be changed at any time after a predetermined number of transfer errors. You can adjust the size of the data, node size, and transfer speed. The principle of the invention According to this, the originating and receiving computers are defined by the user. virtuality Network management functions control which functions are installed on each computer. can also be done.

The primary message structures generated at the application layer are headers and Contains a data area that stores information transferred between data. one implementation In the example, the header (i) specifies the virtual network that originates the message; (ii) performed by the computer receiving the message; fields that specify the behavior to be performed, i.e., commands from the user application. and (i i i) the direction of the message, i.e. if the message is A field to specify whether it is an action request message or a reply message, (iv) a field specifying the remote computer to receive the message; The application layer, which contains the messages itself, Build a unique message structure for each remote action you want to send.

The second message structure generated at the transport layer is the header at the application layer. a different header and at least part of the information of the message of the first structure and an error code. contains the code.

To facilitate the transfer of information between computers, the application layer Compress and decompress information exchanged over links.

At least two computers linked via a data link according to the invention The information transfer method between Information is sent and received using data. In one example With the given parameters, it is possible to send and receive the largest network size at the highest speed. is selected as follows. predetermined transfer) The following processing is used.

After the initial connection, a predetermined number of packets of information are sent over the data link. Information is transferred at a given transfer rate and block size until an error is detected. . Once a predetermined number of errors are detected, the block size is reduced and the transfer is performed again. It will be done. If the transfer is successful, the new block size and same transfer rate will be used. used. If the transfer is not successful, the block size is reduced again.

This step is repeated until the transfer is successful or the block size is the minimum. It will be done. If the block size is the minimum and the transfer is still not successful, the transfer speed The block size is maximized. Again, block until the transfer is successful. The book size is continuously reduced. Both transfer speed and block size are minimal. , If the transfer is not successful, the transfer is aborted. The transfer speed is fixed for some reason. If specified, the block size only changes once a certain number of errors are detected. be changed.

In accordance with the principles of the present invention, programs that can be executed directly from read-only memory (ROM) A computer program that runs executable programs, especially for virtual networks. A novel loader and method for configuring a computer is provided. Method and loader , overcomes the limitations in the prior art regarding relocation.

The loader of the present invention allows random access memory (R AM). When the loader is loaded into RAM, the loader Configure the environment necessary to run an executable program directly from ROM.

In particular, in one embodiment, the initialization means includes an executable program in a ROM. Initializes the computer for direct execution from ROM. Initialization is Contains the formation of a data area in RAM for executable programs in ROM .

Once the necessary environment settings for running an executable program in ROM have been made, the The data is preferably evacuated from the RAM by RAM release means. ROM fruit Executable programs use the data area created in RAM by the loader. can be executed directly from ROM.

Accordingly, in one embodiment, the loader includes (i) an executable program in ROM; means for initializing a computer system to directly execute from ROM; ii) having means for releasing RAM; Of course, in this example, Provide the loader with a means to release random access memory by the loader itself. However, even without a means of releasing the RAM used by the loader, the loader is It is fully operational.

However, in these embodiments, even after the loader is no longer needed, some of the RAM remains is occupied by the loader, which deteriorates the efficiency of RAM usage.

In other embodiments, data area initialization means and application existence confirmation and setup means. The means for initializing the data area is (i) How to form a working data area in RAM for an executable program in ROM and (ii) to directly execute the executable program in the ROM from the ROM. and means for initializing variables in the data area necessary for the data area. operating The working data area, if the system creates a suitable working data area. A forming means is no longer necessary.

The means for confirming the existence of the application is in the ROM of the computer system. Verify that an executable program is physically present. Application When the existence confirmation means of the ROM detects an executable program in the ROM, the loader continues the operation by the setup means. However, the computer system If there is no ROM executable program in the system's ROM, an error message will be displayed. The error message is given to the user by the termination means, and the loader The operation will stop. The means for verifying the existence of the application further includes the execution of the ROM. Check if an executable program is already loaded on your computer system. Contains means to check.

The setup means includes data segments required by the executable program in the ROM. sets the register and other registers and vectors. Cent up means The operation depends on the nature of the executable program in the ROM, as described later. Become. Generally, the setup means will reset the RAM at the request of the loader itself. The executable program in the ROM takes over control of the computer. Set The top-up means converts the executable program in the ROM into a resident program (TSR). have the means to do so.

After initialization by the loader, the executable program in the ROM is will be executed accordingly. Generally, an executable program is (i) a TSR application; and (ii) non-TSR applications. Non-TSR apps In applications, control is pulled to an executable program in ROM after initialization. The program is inherited and the executable program in the ROM is executed. When the execution is finished, R The OM executable program is terminated, usually the MS-DOS operating Return to system.

For TSR applications, the setup means are ROM Let the executable program be a TSR program. Also, the ROM's executable The program (i) sets up the necessary interrupt vectors, and (i) Executes the code in ROM that forms the application as a TSR program ( ii) It is also possible to make it resident while releasing unnecessary code/data It is. In this embodiment, the executable program in ROM is the TSR program. Since it functions as a ROM, it is not executed directly from the ROM after initialization. To the present invention An important feature is that the loader executes the executable program in ROM directly from ROM. It is important to have the necessary structure to carry out the work.

According to the principles of the present invention, an executable program in ROM can be executed directly from ROM. A method is provided. This method uses a working data array for an executable program. steps for forming the program in RAM and executing an executable program in ROM. The method includes the step of initializing variables in the data area of the RAM. preferred In particular, after the initialization step, the RAM used for initialization is released.

times! breast milk FIG. 1 is a diagram showing a computer system according to the prior art.

FIG. 2A shows the two prior art uncoupled computer systems of FIG. FIG. 2 is a diagram illustrating information transfer between systems.

FIG. 2B shows the two prior art computers of FIG. 1 connected by a modem. FIG. 2 is a diagram showing information transfer between computers.

FIG. 2C shows the two prior art components of FIG. 1 connected by a direct link. FIG. 2 is a diagram showing information transfer between computers.

FIG. 3 is a diagram showing a network of the prior art computer system shown in FIG. be. FIG. 4 is a generalized block diagram of the virtual network configuration of the present invention. Ru.

FIG. 5 is a more detailed block diagram of the virtual network configuration of the present invention.

FIG. 6 shows a general message structure including a header and data area according to the principles of the present invention. FIG.

Figure 7A shows the message structure for the functionality of the OAh C0NNECT service. FIG.

Figure 7B shows the mensage structure for the functionality of the OAh LISTEN service. This is a diagram.

Figure 8 shows the mensage structure for the functionality of the OAh DISCONNECT service. FIG.

Figure 9A is a diagram showing the message structure for the functionality of the OAh5END service. It is.

Figure 9B shows the message structure for the OAh RECEIVE service functionality. FIG.

Figure 10A is a diagram showing the message structure for the functionality of the OAh MAP service. It is.

Figure 10B shows the reply message structure for the OAh MAP service functionality. This is a diagram.

Figure 11A shows the functions for the 09h 5ELECT DTSK service OEh. FIG. 3 is a diagram showing a message structure.

Figure 11B shows the functions for the 09h 5ELECT DISK service OEh. FIG. 3 is a diagram showing the structure of a reply message.

Figure 12A shows 09h GET FREE DISK 5PACE service 3 FIG. 6 shows a message structure for the 6h function.

Figure 12B shows 09h GET FREE DISK 5PACE service 3 FIG. 6 shows a reply message structure for the 6h function.

Figure 13A shows the function of 09h MAKE DIRECTORY service 36h. FIG. 2 is a diagram showing a message structure for.

Figure 13B shows the function of 09h MAKE DIRECTORY service 36h. FIG. 3 is a diagram showing a reply message structure for.

Figure 14A shows the 09h REMOVE DIRECTORY service 3Ah. FIG. 3 is a diagram showing a mensage structure for a function.

Figure 14B shows the 09h REMOVE DIRECTORY service 3Ah. FIG. 3 is a diagram showing a reply message structure for the function;

Figure 15A shows 09h CHANGE D'IRECTORY service 3Bh This is the time to show the message structure for the function.

Figure 15B shows the 09h CHANGE DiRECTORY service 3Bh. FIG. 3 is a diagram showing a reply message structure for the function;

Figure 16A shows the function for the 09h DELETE FILE service 41h. FIG. 3 is a diagram showing a message structure.

Figure 16B shows the function for the 09h DELETE FILE service 41h. FIG. 3 is a diagram showing the structure of a reply message.

Figure 17A shows the function for 09h CREATE FILE service 3Ch. FIG. 3 is a diagram showing a message structure.

Figure 17B shows the function for 09h CREATE FILE service 3Ch. FIG. 3 is a diagram showing the structure of a reply message.

Figure 18A shows the menu for the 09h 0PEN FILE service 3Dh function. FIG. 3 is a diagram showing a tsage structure.

Figure 18B shows the return for the 09h 0PEN FILE service 3Dh function. FIG. 2 is a diagram showing a communication message structure.

Figure 19A shows the menu for the 09h CLOSE FILE service 3Eh function. FIG. 3 is a diagram showing a tsage structure.

Figure 19B shows the return for the 09h CLOSE FILE service 3Eh function. FIG. 2 is a diagram showing a communication message structure.

Figure 20A shows the menu for the function of 09h WRITE FILE service 40h. FIG. 3 is a diagram showing a tsage structure.

Figure 20B shows the return for the 09h WRITE FILE service 40h function. FIG. 2 is a diagram showing a communication message structure.

Figure 21A shows 09h READ FILE service 3Fh READ FILE FIG. 3 is a diagram showing a message structure for LE functionality;

Figure 21B shows the return for the 09h READ DISK service 3Fh function. FIG. 3 is a diagram showing a communication message structure.

Figure 22A shows the message for the 09h LSEEK service 42h function. It is a figure showing a structure.

Figure 22B shows the reply message for the 09h LSEEK service 42h function. FIG.

Figure 23A shows the message for the 09h CHMOD service 43h function. It is a figure showing a structure.

Figure 23B shows the reply message for the 09h CHMOD service 43h function. FIG.

Figure 24A shows the function of 09h GET DIRECTORY service 47h. FIG. 3 is a diagram showing a message structure for.

Figure 24B shows the function of 09h GET DIRECTORY service 47h. FIG. 3 is a diagram showing a reply message structure for.

Figure 25A shows the menu for the function of 09h FIND FIR3T service 4Eh. FIG. 3 is a diagram showing a tsage structure.

Figure 25B shows the return for the 09h FIND FIR3T service 4Eh function. FIG. 3 is a diagram showing a communication message structure.

Figure 26A shows the menu for the 09h FIND NEXT service 4Fh function. FIG. 3 is a diagram showing a tsage structure. .

Figure 26B shows the return for the 09h FIND NEXT service 4Fh function. FIG. 3 is a diagram showing a communication message structure.

Figure 27A shows the message for the function of 09h RENAME service 56h. FIG.

Figure 27B shows the reply message for the 09h RENAME service 56h function. FIG. 3 is a diagram showing a sage structure.

Figure 28A shows 09h GET/SET TIME/DATE service 57h FIG. 3 is a diagram showing the message structure for the function.

Figure 28B shows 09h GET/SET TIME/DATE service 57h FIG. 3 is a diagram illustrating a reply message structure for the function of FIG.

FIG. 29 is a diagram illustrating a status structure in accordance with the principles of the present invention.

FIG. 30 is a diagram showing the calling structure of the service function of the present invention.

Figure 31 shows the status event specified by the status address in Figure 30. FIG. 2 is a diagram showing the configuration of an agent handler.

FIG. 32 is a block diagram of the transfer layer function TSEND.

FIG. 33 shows a packet according to the principles of the present invention generated by the transport layer function TSEND. FIG.

FIG. 34 is a block diagram of the transmission operation of the Paquette 170S12 in FIG. 32.

FIG. 35 is a flowchart of the transfer layer function TRECEIVE.

FIG. 36 is a flowchart of the receiving operation of the I70R6 (bucket in FIG. 35). .

FIG. 37 is a flowchart of the sending operation of reply 170R13 in FIG. 35.

FIG. 38 is a flowchart of the link function DRESET.

FIGS. 39A and 39B illustrate a 5ET CONNECTION according to the principles of the present invention. This is a flowchart of the 5PEED function.

FIG. 40 shows RECEIVE CHARACTER5TAT according to the principle of the present invention. It is a flowchart of the US function.

FIG. 41 shows the flowchart of the RECEIVE CHARACTER function according to the principles of the present invention. It is a low chart.

FIG. 42 shows a 5PEED RESYNCHRON IZA according to the principle of the present invention. It is a flowchart of the TI ON function.

Figure 43 shows the flow of the SET LISTENSPEED function according to the principle of the present invention. It is a chart.

FIG. 44 is a flowchart of the data link function DSEND according to the principles of the present invention. It is.

Figure 45 shows the flow of the 5END PACKETSIZE function according to the principle of the present invention. It is a chart.

FIG. 46 is a flowchart of the GET PACKETSIZE function according to the principles of the present invention. It is a chart.

FIG. 47 is a flowchart of the data link function DRECEIVE according to the principles of the present invention. It is a chart.

FIG. 48 illustrates the flow of interrupt service routine 180-INT in accordance with the principles of the present invention. It is a low chart.

FIG. 49 shows the flow of the interrupt subroutine 180-RECV according to the principles of the present invention. It is a chart.

FIG. 50 is a flowchart of the data link function DCONNECT according to the principles of the present invention. It is a chart.

Figure 51 shows the MAIN INITIALIZATION function according to the principles of the present invention. This is a flowchart.

FIG. 52 shows the C0NNECT/LT5TEN INIT according to the principle of the present invention. 3 is a flowchart of the IALIATION function.

Figure 53 shows the flow of the CHECK LINESPEED function according to the principle of the present invention. It is a chart.

Figure 54 shows the flow from the data link function DLISTEN according to the principles of the present invention. It is a chart.

FIG. 55A shows a computer including a ROM executable program loader according to the present invention. 1 is a block diagram of a computer system.

FIG. 55B is a block diagram of a loader embodiment of the present invention.

Figures 56A and 56B are flowcharts of one embodiment of the loader of the present invention. Ru.

Figures 57A-57C illustrate the RAM memory allocation associated with the loader of the present invention. FIG. 7 is a block diagram of another example.

Figures 58A to 58C show the functions for the library of functions stored in the ROM. 1 is a flowchart of one embodiment of a loader according to principles of the present invention.

Figures 59A to 59D show the loading by the loader of Figures 58A to 58C. ROM (75 applications) that operates with programmed functions and resident programs. 1 is a flowchart illustrating one embodiment of a loader according to the present invention for loading.

Note! Hoshi R-ri In accordance with the principles of the present invention, at least two computers can be combined into a single virtual computer. A novel computer configuration for configuring as a data system is provided. In particular, virtual The network configuration is attached to each computer in the system. new virtual A network is a network in which a first computer performs normal operations and a second and other computers The computer function only services requests from the first computer, and therefore the server -Can perform functions called computers. The configuration of this computer is called a virtual computer. Therefore, according to the principles of the present invention, the first computer In addition to the computer's main memory and secondary storage, the Connect to other computers' computers through a virtual network configuration called a network. It is also possible to directly access in-memory, secondary storage, and peripheral devices. in fact , the first computer accesses all the resources of the server computer.

The secondary storage of the server computer is the low-speed secondary storage of the first computer. It will be placed. Here, slow is the first computer of information on the server computer. refers to the access time relative to that of computers and similar information. . Access times for information on server computers are in addition to normal access times. , the time to receive the information request and the time to send the information to the first computer.

Therefore, the server computer's access time is less than the first computer's normal access time. access time, and the resources of the server computer are The speed is slow compared to the data resources. Information access time is the first computer's Although slower relative to access time, information transfer is faster, as explained in more detail below. The flexibility of transfer is greater compared to traditional transfer methods and, more importantly, , virtual network configurations enable new computer applications. It is to provide a virtual computer system.

The virtual network configuration according to the present invention provides user applications and computer interposed between the computer's operating system. In one embodiment, the virtual The network has read-only memory in the computer and is detailed below. Uni ROM Loaded as part of BiO2. In other embodiments, virtual The network is loaded as a terminal and resident program.

The virtual network configuration provides two functions to user applications. No. One feature is that in the first embodiment, a user application can connect two virtual networks. between networks, one computer and the other computer as a server. It is possible to form a virtual network by using the following data. The first function is , not only allows connection with other computers (virtual network system) system status and to initialize the virtual network. Enables sending and receiving messages. Virtual networks allow user applications to The second function provided to the application is that the user application An operation that uses the resources of the server computer when executed by a computer. The ability to execute rating system commands.

One embodiment of the invention is shown in FIG. first computer 100, first user application 101, virtual network management function 102, Operating system 103 is contained in main memory 117. second The memory 217 of the computer 200 stores user applications 201, A virtual network management function 202 and an operating system 203 are provided. It will be done. In this example, the user application 101.201 is functionally , and the virtual network management function 102.202 is functionally the same. The operating systems 103.203 are functionally identical.

Based on the description below, a person skilled in the art will be able to identify computers with different operating systems. The principles of the present invention can be used to perform virtual network management functions on a computer. can be used. These embodiments describe different embodiments of virtual network management. Requires functionality and user applications for each operating system . However, the functionality and general operation of the user application is as follows: This will be explained along with the configuration of the virtual network.

Computer 100.200 (FIG. 4) has secondary storage devices 118.218 and C PU110.120 and video displays, input/output devices, other components, Contains common parts found in black and mini computers. these elements are well known to those skilled in the art and are included for clarity in Figure 4. not present. (P, Norton and R, Wilton) , Wilton), “New Beta for IBM PC and PS/2 -See “Two-Tone Program Guide” Micro Press (1988)) in Figure 4. As shown, CPUll01210 is connected to serial line 150. Siri The Alline 150 has serial ports (not shown) for each computer 100 and 200. ). Each Syrian boat is constructed according to principles known to those skilled in the art, Coupling, i.e. serial port and CPU computer interface connected to the computer's CPU by hardware components in the Ru. These elements do not form part of the invention. The present invention is suitable for serial connection or gives users new communication possibilities through other data links. Main implementation Although serial connections are used in the examples, those skilled in the art will may have any number of physical or It is also possible to use electronic connections. User Application 101.2 01 and operating system 103.203 are standardly known to those skilled in the art. The computer 100.200 is loaded using the techniques described above. less than Further details user application 101.201 is a virtual network Contains a set of commands that interact with management functions 102.202, including (i) Installing virtual network management function 102.202 on computer 100.200 (ii) Activate the virtual network management function 102.202. let them make (i　i　i)　Computer 100.200 as virtual computer 300 configure, (iv) Execute actions specified by the user on the virtual computer; (v) Send and receive information between user applications 101 and 201.

-g, the user interacts with the user application 101.102. , computer 100 or computer 200 is selected as the first computer. Since the computer is selected, it is configured as a virtual computer 300. first computer is selected, the other computer by definition Become a server. A method for exchanging information between computers in the prior art. This is different from a system that allows users to input information even from computers that are connected to each other. When the virtual computer 300 is created, the server computer computer and supports user human power from the server computer. It will not work.

The computer 100 is a first computer, and the computer 200 is a server. When the virtual computer 300 is created, the user application 20 1 only supports the exchange of information with the computer 100, and the user application The application 101 controls the virtual computer 300. Therefore, the user The virtual computer 300 is run by a user application on the computer 101. Manipulate. As a result, the application 101 is executed on both computers 100. 200 resources are available.

User application 101 commands virtual network management function 102 When the command is sent, the virtual network management function 102 processes the command. less than As further detailed, the virtual seven network management functions 102, at times commands and It has a set of commands called, and controls the operation of the virtual network management machine tm 102. Control. Additionally, any user application can manage the virtual network. A set of commands can be used to access the management functions 102. follow Therefore, unlike conventional systems, any user application can Virtual network configuration features can be used.

A command from the user application 101 is sent to the server 20 along with the operation. 0 information is identified, the virtual network management function 10 is further detailed below. 2 sends a command to the virtual network management function 202 to manage this virtual network. The network management function 202 uses the resources of the computer 200 to system 203 (and sometimes user application 201). Execute the specified action. Upon completion of the requested operation, the virtual network 202 , a message indicating that the requested information and/or requested action has been completed. is sent to the virtual network management function 102, and this information and/or message is sent to the virtual network management function 102. The page is provided to the user application 101.

A command from the user application 101 is sent to the server computer 20 If the 0 operation and information are not specified, the command is local rather than remote information. Since the virtual network management function 102 is related to Processes commands in conjunction with operating system 103. In either case Also, the user application 101 is configured to send commands to the virtual network. It does not matter which of the management functions 102.202 processes the process.

A user application 101 is connected to a computer via a serial port 150. does not provide information suitable for direct transfer to 200. Therefore, the user application The virtual network management function 101 uses the virtual network management function 102 to perform specific operations. Generates a command towards. Generally, a command specifies an action and some type of information. , the user application 101 stores a pointer to the buffer used as information. give ta.

The virtual network management function 102 provides commands and, if necessary, user Convert information in the buffer specified by the application 101 into a message This message is then sent by computer 100 via serial line 150. converted into packets to be forwarded. Similarly, the information in the packet is sent to the computer 20. 0, the virtual network management function 202 sends the packet information to Convert to messages, then convert messages to buffers and commands.

After building a buffer of information from forwarded packets, virtual network management The processing function 202 executes the specified command using the information in the requested buffer. go When the operation is complete, the virtual network management function 202 returns the buffer. the reply message is transferred to the virtual network management function 102. Convert to sent packet. The virtual network 102 sends the received packets to Convert reply messages to buffers, convert reply messages to buffers, and use buffers. the user application 101.

In the above description, the user application 101 is assumed to be accessing or using. However, more details are provided below. As described above, virtual network 102 also serves as a hub for computer 100 information. command processing.

In one embodiment, the virtual network management function 102.202 may 120 and 130 (Figure 5). The first function 120.220 is means for forming a virtual computer 300 from a computer 100.200; provides a means for transmitting information between computers 100 and 200.

Accordingly, the virtual network management function 102.202 operates in two modes.

Operation in the first mode is such that the user application 101.201 Using the first function of the network management function 102.202, the computer 1 Information is transmitted and received between 00.200 and 200. In the second mode of operation, the controller Computers 100 and 200 are configured as virtual computers, and one The computer is in server mode, and the first function 120.220 and the second function 130 are Resources on both computers can be accessed by user applications on other computers. used for accessing

The operation of the virtual computer 300 using the virtual network is shown in the system shown in FIG. Formed through real lines. Virtual network management function 102.202 is , first function 120.220 and second function 130.230, transfer layer 170.1 80 and a data link layer 180.280.

(As used herein, when two reference numbers are included after the name of the object, The functions of the objects shown are the same. ) Each function is - boat fishing (including some services). I'm reading.

Each virtual network management function 102.202 layer communicates only with adjacent layers. . For example, application layer 160.260 is user application 10. 1,201 and the transfer layer 170, 270. As shown in Figure 5, the series Alline 150 also has a data link connected computer 100.200. This is an example. The principles of the present invention apply to I can do it. Virtual network management function data link 180.280 electronic data If not selected to allow connection with link and transport layer 170.270 No.

In one embodiment, a user first accesses user application 101. A command is given to start up the virtual network management function 102. Depending on the user's instructions, The user application 101 sends a command to the first function 120 to temporarily The virtual network management function 102 is reset. This command is for local operation. command, the function 120 is a virtual network management command of the computer 100. The management function 102 is activated. The user performs this process as temporary P, = Q node work management. Iterate on computer 200 to activate function 202.

When the virtual network management function 102.202 is activated, the user Create a virtual network through applications 101.201 and The computer 100.200 is made to function as the virtual computer 300. Therefore, this In an embodiment, a user may send user application 201 to a computer. 200 to be configured as a server. Function 220 is a user app In response to commands from the application 201, the virtual network management function 202 is the server, and the virtual network 202 uses the connection command of the serial line 150. receive the code. When the virtual network management function 202 receives the connection command, As discussed below, computer 200, sometimes referred to as a node, - commanded by application 101 via virtual network 150 Service commands.

The user then sends the user application 101 to the computer 101. 200 connections. The user application 101 is a user application 101 that In response, a connection command is sent to the function 120 of the application layer 160.

Connection is one of the services supported by function 120. Application The application layer 160 generates a connection message and calls the transfer layer 170. below The application layer supports each service supported by the functionality, as detailed in generates a unique message. However, the application layer A set of messages is a structure containing a message header and a data area following the header. Become. The message structure also includes first and second functions 120 and 130 for each service. The details are explained below. Also, as described below, the message data area is Sometimes it is compressed to minimize the information that must be transferred to remote nodes.

Transfer layer 170 is used to accurately transmit information to the server computer. The connection message generated at the application layer 160 is broken down into packets. Transfer layer 1 70 processes each packet sequentially. In particular, the transfer layer 170 A CRC-16 checksum is added to each packet to to the computer 200'. .

The transfer layer 170 waits for a reply from the transfer layer 270 indicating that the packet has been received. De The data link 280 receives the packet 280 and transfers the received packet to the transport layer 27. Send to 0. The transport layer 270 checks the transmitted packets to detect errors. and sends a reply to the data link 180 via the data link 280. child The reply is sent to the transport layer 170.

If the packet is forwarded without error, the next packet is transferred to forwarding layer 17. Transferred to 0. However, if an error occurs in the transfer, the transfer layer 1 70 retransmits the packet. This indicates that the forwarding layer 170 successfully forwards the packet. or the transfer layer cannot perform the transfer at the current transfer rate and block size. This will continue until a decision is made. Here, block size means the number of bytes in the packet do.

In this example, the maximum transfer rate is 11 per second, as described in detail below. 5,200 bits, and the maximum block size is 4 kilobits. -touch, I In a computer equivalent to BM-PC or IBM-PC, the UART is It is the interface between software and hardware in a computer.

UART defines a maximum transfer rate.

Typical UART usage specifies a minimum baud rate divisor of 6, so It is 19.2 kilobits. However, if you set the UART baud rate divisor to 1, By setting 115,200 bits per second, it is possible to obtain 115,200 bits per second.

Transfer layer 170 allows information to be transferred in large blocks without errors. If the forwarding layer determines that there is no Divide into block sizes that are half the block size. Smaller block size If the transfer is not successful, the transfer layer 170 determines the block size of the packet. We try to reduce the error by further halving. be transferred The size of the packet is 256 if the transfer is successful or the block size is the smallest. It is split in half until a byte is reached.

If the error rate does not decrease by decreasing the block size, the forwarding layer 170 data link layer 180 to reduce the transfer rate and maximize the packet block size. command to repeat the error reduction operation. The transfer speed is determined at the time of connection. Unlike prior art systems in which transfer parameters are , i.e. the block size and/or transfer rate of the packet is Adjustments can be made at any time when a predetermined number of errors occur. Therefore, the original Advanced virtual network management capabilities can respond to changes in forwarding conditions.

When the packet is successfully received by the data link 280 and the transport layer 270, the transport layer 2 70 sends the message to application layer 260 and sends the message to application layer 26 0 processes the message given from the transfer layer 270 and receives the connection command. determine what happened. Therefore, the application layer 260 Sets the management function to server mode and receives information from the virtual network management function 102. believe

Application layer 270 completes the actions specified in the received message. Once completed, the application layer 260 forms a reply message and sends the reply message. Messages are sent to transport layer 270. Transfer layer 270 is described with respect to transfer layer 170. It performs several operations as described above and sends a reply message to the forwarding layer 170. transfer layer Upon receiving each packet from transport layer 270, 170 forms a reply message. , upon completion of reception, transfers it to the application layer 160. Application 160 disassembles the reply message and uses the returned information to the user application. 101.

Complete set of commands and commands provided to the virtual network management function 102.202 The functions of the virtual 7 software management function 102.202 are explained in more detail below. Ru.

Once the virtual computer 300 has been created as described above, the user can Specifies remote drives that the application 101 can use. In one embodiment As will be explained in more detail below, the remote drive, i.e. the computer 2° drive is aliased as the drive of computer 100. In general, Computer 100 has drives identified by letters AE. Therefore , remote drives are local drives F, G, 8. Closed and aliased. Therefore , when the user application specifies one operation for drives F, G, . , virtual network management function 102, as described in more detail below. Send the request over the network.

A virtual computer 300 is thereby executed on the computer 100. The configuration is such that requests from the user application 101 are serviced. below As will be explained in more detail, server computer 200 has first function 12. 0 and the second function 130.

In another mode of operation, the user application 2.1 runs on the computer 2. 00 is not set to server mode that uses the server service of the first function. this If so, the user application 201 sends a message via the first function 220. send and receive messages. In this mode of operation, the user application 201 is a user application rather than a server service for the first function 220 101 commands are processed. Therefore, the virtual network management function 202 , the user provides processing commands to the server service of the first function 220; Processing is performed via the first function 220 of the virtual network management function 202, and the by giving the user an application that sends the user's own messages. It is possible to perform any operation such as providing a specific service.

In one embodiment, the services of the virtual network management function 102.202 include: , is called using interrupt 66h.

The first function 120 is called by the user application 101.201. interrupt 66h, executes function OAh, and the second function 130 is executed by the user application. Called by application 101.201, interrupt 66h, function 09h Execute. The selection of specific interrupts and specific features is illustrative and does not reflect the invention. is not intended to be limited to the particular embodiments described below. This opening Based on the illustrations, those skilled in the art will be able to implement the functionality of the present invention using other operating systems. point, or connect these services to an existing operating system. The present invention can be implemented by executing the system.

Therefore, in this example, the user application 101.201 is virtual network management function 102.202, function 09h or function OAh and Called by executing interrupt 66h that specifies the service of the specified function. vinegar. Details of the services supported by function 09h and function OAh are below. Explain in detail.

Here, for example, AX (sometimes called two registers AH and AL), B X (sometimes called two registers BH and BL), CX (sometimes called two registers BH and BL), registers CH and CL), DX (sometimes two registers DH and D ), SP, BP, 31. Registration of DI, C3, DS, SS and ES The name is 1A of Intel Corporation in Sunta Clara, California. Family of PX86.88 and 1APX186.188.286.386.486 Lee's microprocessor system.

However, these examples are for illustrative purposes only and do not characterize the scope of the invention. It is not intended to be limited to the specific embodiments. Based on this description, those skilled in the art The virtual network management function of the present invention can be used by other micro processors and operators. can be used with a

The functions 120.220 of the application layer 160.260 services are as described above. As described above, it is called by interrupt 66h, function OAh, and service.

Table 1 lists the nine application layer services of the function OAh according to this embodiment. I'm on strike. In other words, the service of interrupt 66h, function OAh, is provided by the user application to access the control layer 160. Must be.

A register that must be initialized by the user application 101.210 Register information returned to the host and user applications 102.201. Shown in Table 1.

The services described here are only for detailed explanation of the present invention and are It is not intended that the invention be limited to particular embodiments. For example, function OA h can provide only the set of services in Table 1, and can also provide the user with special needs. It is also possible to support additional services as required.

Table 1 Application function OAh service LISTEN, 5ERVICE OO At h entry AL: 0OH CX: Line baud rate/100 (0 = default) Dx: Port selector (0 = COML L = com2.,,,) SI: Timeout C0NNECT, 5ERVICE At O1h entry AL: 0IH Bx: Remote computer node specification CX: Line baudley)/100 ( 0 = Default) Dx: Boat selector (0 = C0M1.1 = C0M2 ．． ,,,)AX:SID or negative error code BX2 Total number of available remote drives DISCONNECT, 5ERVIC E　02h construction time AL: 02H BX: SID session close 5END, 5ERVICE 03h At the time of factory entry ^L: 03H BX:5ID CX: Length of sent message DS: DX: Pointer of message to be sent RECEIVE, 5E RVICE 04h construction time AI: 04H BX:5ID CX: Buffer length DS: Dx: Receive buffer pointer SI: Timeout Upon return AX: Message length or error code 5TATIJS, 5ERVICE 　05h work entry time AL: 05H BX: SID or all Qi 1 CX: Buffer length DS: DX: Status pamphlet pointer RESET, 5ERVI CE 06h construction time ^L: 06H CX2 Work buffer length DS: BX: AX: Error code of working buffer supplied by user de MAP, 5ERVICE 07h BX:5ID CL: Drive number of local alias drive (0=A, 1=Bet c,) CH: Drive number of remote drive (0 = A, 1 = B etc. ) DX: O.Arias release, doorArias generation, 2.Return to Matupdeple If DX is 2, DS:CX: Buffer for receiving drive map table 5EPVE, 5 ERVICE 08h construction time AL: 08H CX: Line baud rate/100 (0-default) DX: Pointer select Data (0' = C0M1.1 = C0M2.,,,) SI: Timeout BX:DI: NuL if there is no address or hand in the callout structure L (0:0) Upon return Ax = Error code Errors in function OAh are returned to the user application as a negative value. machine Successful completion of the function is indicated by a positive (generally zero) return value. Possible occurrence Table 2 shows the return values from the possible functions OAh.

Table 2 Function OAh results belief value 〇　−Function completed normally -1- Poor connection -2-Unknown node ID -3- Incorrect line speed -4= Sending/receiving error -5- Timeout error -6- Unable to connect to specific node -7- Invalid buffer address =8 - Buffer too small -9 - Invalid session ID -10- No sessions available -11-CRC error -12- Unexpected message -13=Illegal remote drive -14- Drive is not local and not aliased to remote drive =15 - Re-entry function request rejected -16- Service of function 10 is called and returns before calling RESεT was done =17 - The service was terminated at the request of the user. The second function 130 provides special services via the user application 101. It is started by calling function 09h, including interrupt 66h. mentioned above As mentioned above, in this embodiment, the operating service selection for MS-DO3 is It can be thought of as a nearly orthogonal set of services. Orthogonal means these services can be used to achieve the full range of interrupt 21 services. It means.

The set of selected services is shown in Table 3.

Table 3 register AL value: Service to be executed OEh Disk selection reason) 36h Obtain free disk space 39h Directory creation 3Ah Delete directory 3Bh Change directory 3Ch file generation 3Dh file oven 3Eh File close 41h Unlink 42h Lseek 43h Cb+1+od 47h Current Direction Acquired 4Ch Processing completed (locally processed) 4Eh Head detection 4Fh Next term detection 56b File close 57h Get/set file data and time To facilitate the use of the virtual network configuration of the present invention, interrupt 66h, function The service number 09h is the same service number as the MS-DO3 interrupt 21h service number. I have a service number. Furthermore, it is supplied when interrupt 66h is called. The required register values are determined by the MS-DO3 function in register AH and parameter AL. function code is stacked by the user application before interrupt 66h. Other than that, the service is the same as the corresponding interrupt 21h service of MS-DO8. .

According to one embodiment, the global interrupt 66h handler is called Before returning to the program, release the pressed AX parameters and use the user program. The program needs to adjust the stack after calling interrupt 66h function 09h. I try not to.

The use of the MS-DO3 interrupt 21h aspect is determined by the application program. Interrupt 66h and Interrupt 21h virtual network configuration, since the concatenation of To facilitate the realization of In practice, as explained in more detail below, interrupt 66h Function 09h normally interrupts interrupt 21h to efficiently perform the requested service. use.

The operation of the virtual network management function is generally illustrated in FIG. Other examples In , the global interrupt 66h handler is The operation starts when interrupt 66h is issued. User application interrupts 66 Before issuing h, the user application calls the interrupt function 66 Set the parameters required by h. In this example, all For functions other than function 09h of interrupt 66h, register AH is The service number is set in register AL.

In the case of interrupt 66h function 09h, the user application must register AX Set the appropriate value for the service and push that value onto the stack. You The user application then sets register AH to 9 and interrupts 66h. occurs.

When interrupt 66h is received, the global interrupt 66h handler first re-ends. If a relevel is performed and this entry point is not a reentry point, the stack is Switch. The current value of the CPU register is saved, for example when using a screen segment. Various parameters such as color and associated colors are determined. global interrupt The 66h handler then hands over control to the main dispatcher. This may The dispatcher determines whether the interrupt 66h feature is supported. . In this embodiment, the main dispatcher has a function number below OAh. It is determined that the period is up to 30 hours. If the feature is not supported, the glow A computer loaded with the global interrupt 66h handler and virtual network management functions. In response, the computer system performs the following actions.

Pending application by John Bee, Fairbanks, et al.

Title of the invention, assigned to applicant by Fairbanks et al. 1. Portable Low Power Comp U.S. Patent Application No. 07/375,721, filed by J. The disclosure relates to low power On portable computers, if the feature is not supported, Interrupt 66H hand that becomes operational when the network management function is loaded. La is called. When returning from the handling routine, the control is passed from the main dispatcher to the global picture interrupt 66h handler. It will be done. The re-entry level is determined and the stack is restored to its state at the time of entry. , if the handling routine returns an error code in register AX The carry flag is set and control is returned to the calling application.

Therefore, virtual network management functions are transparent to the calling application. becomes.

In other embodiments, on a computer other than a low-power portable computer, the global If the function passed to the interrupt 66h handler is not supported, the The local dispatcher globally assigns error messages indicating parameter failures. The global interrupt 66h handler returns to the interrupt 66h handler, and the global interrupt 66h handler re-levels the interrupt. lowers the value, sets the carry flag, restores the stack to its initial state, and return the error code to the user application.

If the function passed to the interrupt 66h call is supported, the main The spatcher compares the requested service with the highest service number for that feature. Determine whether the service you are using is legitimate. If the service is legitimate, the main The spatcher loads the address of the function it handles and then returns control to the handler. processing function. When returning from a function, the main dispatcher takes control Pass to global interrupt 66h handler, global interrupt 66h hand , as mentioned above, lower the re-entry point and carry depending on the returned code Set the flag, reset the stack, and return to the user application . As a result, a series of operations of the virtual network management function are processed.

However, the servicing of interrupt 66h function 09h is handled somewhat differently. be done. The maximum service number for function 09h is 1. Therefore, function 09h is If called by service O, the main dispatcher will Load the VNA dispatcher, which will be described in detail. If function 09h is called other than O If the service is unsupported, the a function is 09h, so the control Control is passed to the VNA dispatcher. When the service of function 09h ends, The VNA dispatcher takes saved registers off the stack and puts them on the stack. control from the global interrupt 66h handler. Returned to user application. Service is not supported and does not work is not 09h, what to do if the specific feature mentioned above is not supported. The same process as the process is performed.

The operation of application layer services for functions 09h and OAh is shown below. This will be explained in detail below. While the global interrupt 66h handler is doing other processing, Legitimate functions other than functions 09h and OAh do not form part of the present invention and will not be described in detail. I will omit the details.

In one embodiment of the invention, the functionality of the application layer is primarily in the C language. The interrupt handler and dispatcher are written in assembly language. , VNA dispatcher and function 09h service and main dispatcher There is an interface written in assembly language between OAh and the application layer function OAh. A set of routines is needed to interface the These interfaces Sroutines move from a register-based assembly environment to a stack-based “C” environment. parameters for adjusting the parameters in a manner well known to those skilled in the art.

As mentioned above, the VNA dispatcher is interrupted by the main dispatcher. When it is determined that 66h has been issued with the requested function number 09h, the control is receive. The VNA dispatcher first runs the virtual network management functions. Determine whether the work is in progress. If the virtual network management function is not working , the VNA dispatcher performs local functions. In particular, the requested service If the service is the termination function of MS-DO3, the virtual network management function The re-entry point is reset to O, and a flag indicating whether the VNA is in operation is set. indicates that the VNA is not operational. The register is set before the start of interrupt 66h. , the value of register AX is drained onto the stack, and at this point, Interrupt 21h is executed. When returning from interrupt 21h, processing is performed by VNA. Returned to the dispatcher and then returned to the main dispatcher.

If the virtual network management function is running, the VNA dispatcher will Check the service number and confirm that the virtual network management function is directly supported in function 09h. whether the requested service number is listed in the table of service numbers to judge. If the service is not in the table, the service will be added to the virtual network Management@functions are performed locally as if they were not active. On the other hand, the service If the interface functionality is will be called. The interface function decomposes the parameters and actually hacks the control. passed to the function to handle.

Each handling function uses parameters given by the user application. Determine the bus, handle, or other identifier that identifies the desired behavior do. Based on this information, the handling function determines whether the function should be executed locally. Set the flag indicating whether it is a kimono or not, return it to the VNA dispatcher, and then return it to the VNA dispatcher. increases VNA re-entry points and provides specific functionality, as described in more detail below. , and click the flag indicating whether the function should be executed locally or not. and return to the VNA dispatcher.

When returned from the handling function to the VNA dispatcher, the VNA device to determine whether the dispatcher is required to perform a function locally. The flag can be checked for If the flag is sent, the function is executed when it is not running. If the flag is not set, V The NA dispatcher reduces re-entry points and, as described above, Return to spatcher.

In the following, each user application layer's support for functions 09h and OAh is explained below. services are explained. Some services are performed locally, others are processed remotely is executed. Here, local means the computer that performs a specific operation; Remote refers to other computers accessed as a result of local operations. virtual net For each remote operation performed through a workpiece, the application layer function Forms a message that is provided to the transport layer by a function call. Transfer layer function For each call, the application layer provides the information shown in Table 4 and the transport layer function and receive the information shown in Table 4. The called transport layer function This will be explained in more detail after the explanation of the main layer. The transport layer functions coexist with the data link functions. to call. This data link function will be explained after the transport layer function.

The message structure of the application layer (Figure 6) consists of function OAh and function 09h. The service is the same. In Figures 6 to 28B, the message The first byte of each column is shown to the left of the column. Also, a value or multi-value descriptor is given to each byte of these drawings. Each mensage 160-M has 8 bytes It consists of a header 160-MH and a data area 160-MD. message's The part after the header is the data area, but in most cases the data area is several blocks. Contains information characterizing the site's data and actual data. Therefore, the data area area is the storage area for information transferred across a virtual network in a message. It is. Messages can contain up to 64 kilobytes of information, including an 8-byte header. can contain information.

8 bytes of message formed by application layer 160.2609 The header has a fixed structure. In this example, the application The first 3 bytes from 00h to 02h of the header of the message layer indicate the source of the message. used for identification. Therefore, in Figure 6, "VNA" is the message is determined by the application layer 160 (FIG. 5) of the virtual network management function 102. This indicates that this has occurred. The fourth byte is a space, and this Pace can also be considered as part of the message source identifier.

Figure 7A~! In Figure 28B, spaces are the ASCII code for spaces. 20h''.The fifth byte indicates that the message packet is an operational message. used to indicate whether it is a message packet or a reply message structure. , thus indicating the transfer direction. If the fifth hide is 00h, the message package A packet is an operational message packet. If the 5th byte is Olh, The message packet is a reply message packet. The sixth Hyde is Metsu Used to identify the service that generated the message. The sixth byte is Indicates a command specified by a user application. For example, function 0 For a 9h service, the sixth byte corresponds to the decimal number given in Table 3. It becomes what it is. The seventh and eighth bytes of the message packet are identification (SID), which will be explained in more detail below.

The remainder of the message provides specific information on the function being processed. one Generally, in the operational packet, the rest of the information is passed to the user application. Therefore, we have the given data. In the reply packet, the message The remaining portion corresponds to the information and/or actions specified in the outgoing message. is typically provided by the application layer of the user application. Ru.

The service C0NNECT of function OAh is the remote node specified in register BX. It measures the formation of a sensation with. (Here, a node is a virtual network means a single computer connected to a network. ) In one embodiment of the present invention Therefore, the application 160.260 and the transport layer 170.270 are distributed over multiple servers. - computers, and the virtual computer 300 is the first computer 1 00 and multiple server computers. As stated below, C LI 5TEN in response to 0NNECT returns session identification (SID) . provided to the transport layer 170.270 by the application 160.260 C0NNECT message 160-C is shown in FIG. 7A.

In the second embodiment, the node identification and session identification are the same. but However, in general, node identification is unique to each computer in a single virtual computer. used to identify data. SID prevents interference and confusion between tasks. identify tasks that are performed over a virtual network to For example, you A server application can request server computer information, and this The request has a SID of 1. Next, the user application is 5TDI other information before the information associated with it is provided by the server computer. may be requested. Therefore, the second information request is 5ID2. You The server application uses the information provided by the server computer using the STD. to identify. Similarly, in a multitasking environment, the SID is associated with each task. can be attached. In order to use multiple SIDs, the data link layer described below is required. Development is required.

CONN E CT message 160-C (Figure 7A) is based on the principles of the present invention. This shows the structure of a typical message.

The fifth byte is "FFh". SID goes low in response to C0NNECT. Since it is generated by I5TEH, C0NNECT message 160 -The seventh and eighth hates of C indicate that no SID is generated. It's ooh. In this example, the C0NNECT message 160-C is: It is 16 bytes long. Ninth and tenth Hyde to transfer data indicates the maximum working buffer size. 11th and 13th byte is used as the first identification field, and the 13th and 14th bytes is used as the second identification field.

In one embodiment, the virtual network configuration is a low-voltage cellular network, as described above. This portable computer handles all interrupt 66h calls. Then, the interrupt 66h is processed, for example, by the virtual network management function l○2 (No. It has a global interrupt handler 66h that passes it to an appropriate handler such as Figure 5). . Therefore, the first identification field is the barge of the global interrupt 66h handler. the second identification field is used to identify the virtual network management Used to identify the version of function 102. Global interrupt 66h If the handler and/or virtual network management function 102 is changed, the version of global interrupt 66h handler and/or virtual network management Functions may not be compatible.

The 13th and 16th bytes of C0NNECT message 160-C are has a derived language. The message derived language is the application layer This is the version number of the message structure used in the message. For example, the seventh The cents of the message structures shown in Figures A to 28B are derived from message-derived languages. It is considered to be 0. Subsequently, if - or multiple mensage structures are changed , the message derived language of the new message structure set is, for example, 1.

The operation of the virtual network management functions is explained in more detail below. In relation to the message structure, different message derived languages can be deciphered in both message derived languages. Does not function in virtual network management functions that do not have the capability. It is shown in Table 1. The user application 101 (Figure 5) is the C0 of the application layer. (i) Node ID of the computer connecting to the NNECT service, (ii) Transfer transmission rate, (ii) the time period, typically 1 second, waiting for a reply from a remote node; give the timeout value, and (iv) the serial port to connect to.

The handling function of service C0NNECT is the function VC of the application layer. Call ONNECT and pass this information to function VCONNECT. entry At this point, the function VCONNECT is the first virtual network management function to operate. Determine whether or not the device is in operation. If the virtual network management function is not operational , an error is returned to the user application.

If the virtual network management function is operational, the function VCONNECT is , calls the transport layer function TCONNECT, described in more detail below. machine The function VCONNECT is the node ID given by the user application, Provide the transfer rate, timeout value and port information to the function TCONNECT. machine If the function TCONNECT operates normally, the function VCONNECT will be set as the reply value. and receive the SID.

In this example, for a SID to be valid, the SID must be greater than zero. must be less than the specified maximum value. The actual return value used is arbitrary. It is. An important feature is that the return value in a normal case is different from the return value in an error case. That's true.

If the function TCONNECT operates normally and returns a valid SID, the function VCO NNECT stores the message header in the local message buffer as described above. form and give it to the function VRESET. The function VREST is explained in more detail below. do. Here, if the M function forms the header of a message, the function Initializes the message data header within. Message Ba Nfa Size (Part 1) Bytes 08h and 09h in Figure 7A) are set to the local mensage buffer size. and defines the function VREST, as described in more detail below. identification by is set to the local version identifier. In this case, the local area is C0NNE This means the version used on the computer that started CT. Mail After forming the message, the function VCONNECT calls the function TSEND of the transport layer. to send a message. The function TSEND is described below with respect to FIG.

If the function TSEND returns an error, the function VCONNECT Call function TD I 5CCONNECT to remotely compile the local computer. disconnect from the computer.

The function V CON N E CT sends the return value from the 2 function TSEND to the user account. application 101.

On the other hand, if the function TSEND completes its operation normally, the function VCONNECT The transfer layer function TRECE IVE is set to the current SID, reception buffer length, Called with receive buffer pointer and timeout value. Function TRE If CE IVE returns an error, function VCONNECT returns function TD l5CO. NNECT and returns the error to the user application 101.

If the function TRECE IVE is successfully received, the function VCONNECT will Select the smaller of buffer size and remote buffer size, and set the working buffer size to Press the selected value. In the following explanation, the local mensage buffer is Refers to the area of the working buffer used for forming application layer messages. do. The function VCONNECT uses the message 160-L and save the rest of the reply message. Function VCON NEC provides the total number of SIDs and remote drives to the user application 101. return.

Service LI5TEN of function OAh is commanded by user application. When the C0NNECT request is set up for a virtual computer, LISTEN Waits for a signal to start the session. Service LISTEN terminated normally. 5ID (06h and 07h in Figure 7B) and disk drive counter. (Hides OAh and 0Bh) respond to C0NNECT by LISTEN. returned to the node that responded. In particular, disk drive counts are Information that can be accessed using the system's disk drive services Identify the number of storage drives. Also, when LISTEN ends normally, the node RECE IVE, as described in more detail below. S.I.D. Used in all consecutive operations to identify the session.

Message 160- generated by LISTEN in response to C0NNECT L is shown in Figure 7B. The first 8 bytes are C0NNE in Figure 7A. CT message and the fifth Hyde message is a reply message. is the same except that it is changed to Olh to indicate OAh and 09h. Bytes is the local message buffer size, OCh and ODh are the local global These are two identifiers: the global interrupt 66h handler and the virtual network management function.

10h and llh are server types. The first server type is LIST The second server type associated with EN is described in more detail below. 5ERVE. The server type has the two behaviors mentioned above. It corresponds to the mode. Bytes 12h and 13h are local virtual network management It shows the message derived language of the physical function.

When a user application calls service LI5TEN, LI The handling function for 5TEN calls the function VL I 5TEN . Transfer rate, timeout value and port are passed to VL I5TEN. en At three points, the function VLISTEN first operates the virtual network management function. Determine whether the state is the same or not. If the virtual network management function is not operational , an error is returned to the user application.

When the virtual network management function is in the operational state, the function VLISTEN is Wait for a connection request by calling the transport layer function TLISTEN, which will be explained in more detail. Two. When the function TLISTEN returns an error, the error is returned by the user application. will be returned to.

However, if the function TL I5TEN completes its operation normally, the SI D is set to the SID value returned from the remote node. Receive message buffer is set to the point of the local message buffer, and the transfer function TRECE IVE is , called to receive an incoming connection message. length of received message is zero or the sixth byte of the message is not set to -1. , the function TDISCONNECT is called and the length of the received message is Returns the original user application.

If the length of the received message is non-zero and the sixth byte is -1, and If the buffer size is larger than the buffer size of the receive message buffer, The working buffer size is set to the buffer size of the received message. Samo Otherwise, the working buffer size is set to the local buffer size. selected The buffer size specified will be included in the reply message. Global interrupt 66h identifiers for the controller and virtual network management functions are used for local messages, The identifier of the C0NNECT message is saved. Similarly, reply message 1 Other information at 60-L is provided to a local message buffer. Transfer layer function T SEND is called to send a reply message.

If the function TSEND returns an error, the function TDISCONNECT The result of TD I 5 CONNECT is returned to the user application. It will be done. When the function TSEND completes successfully, the SID is sent to the calling user app. returned to the application.

The handling function of Service DI 5CONNECT is call VDISCONNECT in response to a call from the application. Functional VD l5CONNECT calls other application layer services and connects to remote nodes. Close the session on the board. An error message indicating an abnormality in the session or function. The code will be returned. All areas specified by the service map shown below The drive becomes invalid and the service that referenced the previously aliased drive Returns an invalid drive error. DI 5 CONNECT ME, Sage 160-D C is shown in FIG. The first 8 bytes have the general structure described above. The sixth byte is “FFh” to indicate DISCONNECTION. ”.

Hyde of 08h and 09h has an optional parameter that is not used at this time. There is.

The function VD I 5 CONNECT is passed to the SID to be disconnected. Function V DI 5CONNECT first determines whether the virtual network management function is operational. Determine whether If the virtual network management function is not operational, an error message will appear. is returned to the user application.

If the virtual seven network management feature is operational, the remote node will As you can see, it is in server mode, and all the activities the server has are live files are opened and closed. DISCONNECT message is formed in a local buffer, and the transport layer function TSEND sends the message to the server. Called to send. The server status is inactive and closed. Each drive associated with a SID to be It is deleted from the drive map table by attaching a mark. drive pine The map table is described in more detail below with respect to the service map. S.I.D. It is reset to zero and the transfer function TD l5CONNECT is called. function The result of TD I 5 CONNECT is sent to the calling user application or or returned to the calling function.

Service 5END of function OAh is used to used to send messages to the receiving remote node. 5END is a user Returns zero indicating that the message was successfully sent to the application. Metsuse Returns a negative error if the page was not sent. 5END is as follows A message structure 160-3 (Figure 9A) is generated, which will be described in detail.

RECE IVE is used when waiting for a message to be received. meso sage Upon receipt of the RECE IVE, the calling service is informed of the actual length of the message. return it.

5END generates a general message 160-5, as shown in FIG. 9A, RECEIVE sends a general reply message 160-R as shown in Figure 9B. generate. Both messages have a general structure with respect to the first eight hides mentioned above. The sixth Hyde is “FCh” to indicate that it is a general mensage. cents. The 08h and 09h bytes are members starting from the OAh byte. Indicates the length of the data area of the page.

The user application 101 calls service 5END and The handling function sets the function VSEND to SID, message length, and message bag. Called by the file pointer. Upon entry, the function VSEND first temporarily Determine whether the virtual network management function is operational. virtual network management machine If the function is not operational, an error is returned to the user application. .

When the virtual network management function is in operation state, the function ■SEN is shown in Figure 9A. A general message 160-5 is formed as shown. The maximum message size is 2 bytes used for message data length from message buffer size The length is calculated by subtracting the first 8 highs and the first 8 highs. User application 101 The message length given to 5END by If the message length is set to the length of the local message buffer, 08 The hide lengths of h and 09h are also set to the length of the local message buffer.

Next, the general message provided by the user application 101 is Copied to the local message buffer and the transport layer function TSEND is called. . The information given to the function TSEND and the operation of the function TSEND are as follows: Explain in detail. Function VSEND is the completion command supplied by function TSEND. code to the calling user application.

In general, a user application: +- rules the service RECEIVE; RECE IVE handling function is SID, message length, timeout value and call the function VRECEIVE with the message buffer pointer. .

At the entry point, the function VRECEIVE activates the virtual network management function. Determine whether or not the device is in operation. Virtual network 7) When the work management function is not operational , an error is returned to the user application.

When the virtual network management function is in operation state, function ■5END is the transfer function. Call RECE IVE to send a general message to the local message buffer. receive. TRECEIVE has a length greater than zero in bytes 08h and 09h When a message with length is received, the length of these bytes is saved. save If the saved length is greater than the size of the local message buffer, the saved The length set is set to the length of the local message buffer. local message bag Messages copied from the message buffer to the message buffer are sent using the function VRECEIV. Passed when E is called. The function VRECEIVE is the length of the message received. to the calling user application.

5TATUS is used to obtain virtual network information. status The information is given to the user application provided by the user application. . The structure of the buffer 160-3TATUS is shown in FIG. 5ID( 7) The count field (bytes 02h and 03h in Figure 29) is currently active. Indicates the number of sessions indicated in the buffer provided. SID count fee The session is repeated for a SID count time of 18h bytes. A status indicator is given. In one embodiment, the STD count is It is 1.

As shown in Figure 29, the session status indicator is remote node ID (bytes 06h and 07h), send hide (high) 08h to 0Bh), reception hide (high+-och to 0Fh), reception error ( Hyde 12h and 13h), transfer rate (byte 14h and 15h), and the above C Contains the identification information (bytes 16h to 19h) described for 0NNECT. Ru. 5TATUS is hooked up, i.e. by a pointer in the call of 5EPVE. It can be called from the part of the software provided to 5ERVE.

User application 101 uses 5TATUS as SID and status buffer long, called with a pointer to the status buffer. Then 5TATUS The handling function calls the function VSTATUS and The transport layer features are explained in more detail below with the parameters specified in Table 4. function TSTATUS. The function TSTATUS is the status of the virtual network obtain. If TSTATUS returns an error, the error code is returned to the application. However, if no error code is detected The global interrupt 66h handler and virtual network management functions are running If the version number of the session is set to the remote interrupt 66h handler and Set to the version number of the virtual network management function. Message derived language is set to zero, and finally, a successful return code is sent to the user application. is returned to the section 101.

The function OAh service RESET is the function of the virtual network management function 102 (Figure 5). Used for initialization/startup or stopping operation. The virtual network management function 102 The user application 101 assigns a valid buffer point to the service RESET. buffer length and effective buffer length. Virtual network management functions 101 is caused by an invalid pointer and/or bad buffer length (preferably zero). The operation is stopped by calling the service RESET.

The handling function of the service RESET calls the function VRESET. Ba The buffer length and buffer pointer are given to the function VRESET. first function VRESET displays virtual network status, server status and SI Set D to zero.

If both the passed length and buffer pointer are zero, the virtual network If the ending offset and ending segment are zero, normal termination occurs in the calling program. returned to the program. However, the ending offset and ending offset of the virtual network If the completion segment is non-zero, the program segment prefix (PSP) An address is obtained. If the PSP address is non-zero, the given buffer The file's PSP pointer is set to the PSP ending address, ending offset and Segment is set to zero. Then ``normal termination is specified by the user application. returned to the section. Also, if the PSP address is zero, normal completion is returned. .

The given buffer length is the predetermined minimum buffer length, 256 in one embodiment. If the buffer length is smaller than kilobytes, the calling user returned to the application. If the given buffer pointer is zero and the given buffer If the buffer length specified is non-zero and greater than the given minimum size, the buffer A pointer error is returned.

If both the given buffer length and buffer pointer are non-zero, the local buffer The file size and starting position of the local buffer are set. current drive, The total number of faulted drives and locally available drives is obtained. The default disk transfer area (DTA, ) and number of disks are obtained, The current DTA is set to the default DTA. Local global interrupt 6 6h handler and virtual network management function version numbers are obtained. Rotation The transport layer function TRESET is called to reset the transport layer.

If the function TRESET returns successful completion, a PSP address is acquired. PSP a If the address is non-zero, the PSP pointer for the given buffer is at the PSP end address. It is set at the point of the dress. If the PSP pointer is not equal to the 10th PSP , the ending offset and segment are reset to zero, and the PSP pointer is 10 PSP cents. Next, the status of the virtual network management function changes. The result of i-function TREsET is set to active state and the result of i-function TREsET is returned to the application.

The service MAP of function OAh is: (i) service RESET is activated, it aliases the local drive as a remote drive, and (ii ) Release the remote drive from Arias, or (II) virtual network Investigate the management function regarding the current state of the map. When the drive is aliased , all references to the aliased drive are, as explained in more detail below, Redirected through a virtual network.

MAP service message 160-MAP is shown in FIG. 10A. first The 8 bytes are the general header structure mentioned above, and the 6th hide is It is set to "FDh" indicating MAP. Part-time jobs for 08h and 09h are Arias. remote drive. The reply message 160-RMAP is shown in Figure 10B. and the fifth byte indicating the reply message and the error code. The messages other than 08h and 09h are the same as the MAP message. to remote drive Assignment of local drive aliases is arbitrary.

The service MAP holds a drive map table.

The drive map table has 26 entries.

Each entry in the table contains (i) the remote drive number for this entry (0= A, 1=B, , , , ) and (i) the SID associated with this entry. have. The zeroth element of the drive map table is for local drive A. These are linked entries. Information about virtual drive Q can be found on the drive map. This is entry 16 of the table. Entry 16 is the remote drive number, Q is the aliased drive and is the SID of this alias.

To alias the drive, user application 101 uses service M AP to SID, local drive identifier, remote drive identifier, alias parameters Call by. If the alias parameter has the value 2, the user alias The 26 word array given by the application is stored in the user drive map. Returned as a table. The user drive map table is a drive map table. It is a subset of the table. Each element in the table has no SID value and has no SID value. The drives mapped by the shown as not being The handling function of MAP service is these Provide parameters to function VMAP. The function VMAP first Determine whether the network management function is in operation. Virtual platform management function is operational If not, an error is returned.

If the virtual network management function is in the operational state, the function VMAP is Determine whether the parameter of the parameter is zero. If the parameter is zero, Drive Map Table Drives associated with aliased local drives The identifier may be set to a value of -1, for example, to indicate that the aliased local drive is free. Indicates that the Successful completion is returned to user application 101.

If the alias parameter has a non-zero value, the map message header data is created in the local message buffer. Message bytes 08h and 09 h is sent to the remote drive identifier and given to VMAP when the function VMAP is called. available. After forming the message 160-MAP, the function VMAP call the transport layer function TSEND to send the message.

If function TSEND returns an error code, function VMAP returns function TD of the transport layer. Call l5CONNECT to end the session. Otherwise, the function VMAP calls the transfer function TRECE IVE to retrieve the mapped driver. A reply message 160-RMAP (FIG. 10B) is received with an indication of .

If TRECEIVE does not return an error code and the returned drive is If the drive does not indicate that it is unavailable, the drive map table The aliased drive is set as a remote drive and the SID indicator is displayed on the SrD. cent. A success code is returned to the user application.

If no remote drive is available, a drive failure error code is displayed in the user app. returned to the application. When TRECEIVE receives an error code, the function VMAP calls the transfer layer function TD I 5 CONNECT to complete the session. finish.

Function OAh's service 5ERVE uses a computer (node) as a server. Called by user applications to configure. Service 5E PVE is service 5. Once EPVE is successfully executed on a node, service is an application layer command from a remote computer that controls a virtual computer. Because it responds to commands, it is actually a higher level service than the other services mentioned above. This is Bell's service. Therefore, the basic features of service 5ERVE are as follows: After explaining the service of application layer function 09h, service 5 ERVE shall be explained in more detail. SERVICE 5ERVE IS FUNCTIONAL VLIS C0NNECT message from remote computer by calling TEN 160 -C (Figure 7A) and performs function VD I according to the request from the remote computer. Call 5CONNECT. User application 201 (Figure 5) In 5EPVE, general message handler, status event handler, user callout structure 160-C3TRUC with the address of the user exit handler Give the address of T (Figure 30). Entries to the callout structure are additive 0:0 is set for unused entries. callout structure The formant is shown in Figure 3Q.

The general message handler is used when SEPVE receives a general message. Always called by the server. Status event handlers are When a request is received from a computer and before sending a reply to the request, called by the server. The general mensage handler is as shown below: Length of message data in register CX, message buffer length in register BX Depending on the size of the buffer and the address of the message data in register DX:AX. will be called.

A general message handler services requests and stores message buffers in new the length of the updated message buffer or the error in register AX. -Return code.

If register AX is zero, no reply message will be sent.

If register AX is -1, the server does not send a reply and returns function VD I5. Call CONNECT.

The status event handler uses the status event structure of register DX:AX. call the address. Status event structure 160-3TATSTRUC T is the current status structure (see Figure 31) and the remote address of the current message. has a Status event handlers allow the server to perform common sending methods. (Figure 9) and the server Called immediately before sending a message other than a standard reply message (Figure 9B). Ru.

General message handlers and status event handlers handle requests from remote nodes. The service of the virtual network management function 202 is used unless communication with the virtual network management function 202 is requested. be permitted to do so. Application layer functions requiring remote access 09h A reference to the service (Table 3) with the carry flag set and Returned with registered register AX (access denied). Requires remote access To refer to the service (Table 1) of function 10 of the application layer, register aX is returned with an error code of -15.

User exit handler terminates on server (after any necessary cleanup processing) used by the user application to command the server , when a remote function request is serviced, and in the waiting state of a remote function request, e.g. For example, in one embodiment, after a predetermined time interval of every 15 seconds, the user termination Call Nora. The user exit handler should be used when the server should not be exited. Returns zero in register AX indicating that it is not working. When a 1 is returned from register AX, the The end of the server is commanded.

The handling function for the disk selection service OEh of function 09h is the SID and function 5EL by drive specification and local message buffer pointer. Call DSK. First, the functions S E L D S K are shown in FIG. 11A. As shown, set up the Message 1.6O-SDSK 8-hide Rada. Ru. Message hides 08h and 09h in the call of function 5ELDSK Set to the given drive. Function 5ELDSK is the transfer layer function TSE ND to send messages to the node indicated by the SID via the virtual network. transfer the page.

Disk selection message 160-3DSK (Figure 11A) is sent to function TSEND. Therefore, when the transfer is successful, the function 5ELDSK is transferred to the transfer layer function TRECE I. Call VE and receive reply message 160-R3DSK. Disc selection The reply to selection message 160-3DSK is shown in Figure 11B. Di The occurrence of the screen selection reply message is explained in the explanation of the operation of the server function below. I will clarify. When the function TRECEIVE completes successfully, the function 5ELDSK returns M. Message 160-R3DSK bytes 08h and 09 Obtained from h. The drive letter of the returned disk is set in register AX, Processing returns to the user application as described above.

When TSEND or TRECE IVE returns an error flag, the carry - flag is set and the reply error value is stored in register AX. In this case too , processing is returned to the user application as described above.

The handling function of the free disk space acquisition service 36h of function 09h is as follows. The function is determined by the SID, drive specification, and local message buffer pointer. Call FREESPACE. The operation of the function FREESPACE is as described above. It is the same as function 5ELDSK. Message 160-FS in Figure 12A is a local message. The transfer layer functions TSEND and TRECE are formed in the transmission buffer. is used as described above. Reply received by function THECE IVE Message 160-RFS (Figure 12B) is classed as Hyde 08h and 09h. The number of sectors per data, the number of free clusters of bytes OAh and OBh, the number of free clusters of bytes OEh and and the total number of clusters on the disk of OFh. Function FREES fee ACE , these four values are stored in registers A, X to DX, respectively.

Error handling and replies are explained in the description of function 5ELDSK.

Function 09h Directory creation service 39h, Directory deletion service 3A h, directory change service 3Bh and file deletion service 41h The handling function also calls the function DIRFUNCS. Each service Function DIRFUNC3 contains the SID, service identification, drive, and bus for the service. and a pointer to the local message buffer. Service 39 h (Figure 13A), service 3Ah (Figure 14A), service 3Bh (first 5A) and the message of the file deletion service 41h (Fig. 16A) are the same. It has the same basic structure and differs only in the sixth byte that identifies the service. ing. Similarly, Figures 13B, 14B, 15B, and 16B are They have the same structure.

Therefore, the function DIRFUNC3 first stores the information in the local message buffer from each function. Generate message headers based on the information provided. message header raw After treatment, bath salt is added after the drive salt to coat the local message buffer. is copied. The drive salt is a dry salt given by the function DIRFUNCS. determined by the block identifier. As mentioned above, the transfer layer function TSEND is It is used to send messages over a virtual network, and the transport layer function T RECE IVE is the reply message from the server specified by SID. used to receive. Functions TSEND and TRECEIVE completed normally. Then, register AX is given to pi I-OA, h and OBh of the reply message. Sets the length of the returned reply message. Reply message bytes 08h and 0 If the 9h error code indicates an error, the carry flag is set. It will be done. If function TSEND or function TRECE IVE generates an error, The carry flag is set and the operating system net is stored in register AX. network error code is set. Processing is done by the user application as described above. returned to the application.

Metrics for file creation service 3Ch and file open service 3Dh Sage 160-CF and 160-OF are shown in Figures 17A and 1.8A. I'm here. Related reply messages 160-RCF and 160-ROF are the 17th It is shown in Figure B and Figure 18B. The structures of these messages are identical; Both services access the function CROP. Handling these services The functions are SID, service identifier, drive identifier, file access attribute, File bus pointer, local message buffer pointer function CROP give to The function CROP opens or creates a file and converts it to 16-bit hardware. The user name is a set of integers called bundles that are used consecutively to identify files. user application.

However, the function CROP opens files on remote nodes and The rating system is not aware of this behavior. Therefore, local operating The system generates a handle on the local file that is identical to the handle on the remote file. there's a possibility that. To avoid this handle collision, the function CROP is The local operating system is used to first create another handle each time the Create and local operating system handle count, oven, and the number of generated files. Therefore, the first of the function CROP The operation is locally the file handle "S t dou t" (I console output) is to use the operating system to duplicate. This creation This ensures that you get a unique file handle.

Function CROP then stores the handle generated by duplication in register AX. and check whether the handle was generated normally. an error occurs , the carry flag is set and the function CROP transfers control to the user as described above. user application.

Checking the unique handle, the function CROP is pointed to by the call @NOCROP. Generates a transfer to the local message buffer for the specified service. Metsuse After generating the message header, the path name is added to the drive salt and the local message buffer is The colon is copied to The drive salt is the drive identity given to the function CROP. Determined by Besshi. As mentioned earlier, the transport layer functionality supports virtual networks. The function TRECE IVE is used to send message through 3 It is used to receive reply messages from the server shown in 1D.

The transfer of the function TRECEIVE was successfully completed, and the reply message was returned at 08h. If 09h and 09h do not indicate an error, register Aχ already has an appropriate value, as described above. The handle is cent. Therefore, the table of file handle structure is updated by the current SID, the handle is the reply message's hide OAh and Set to the value of OBh. Locally duplicated file handle closed handles to keep them unique. Processing is then performed by the user, as described above. returned to the user application.

However, the function TRECEIVE or @functionTSEND returns an error code. In this case, register AX contains an operating system network error. is set and the carry flag is set. In addition, locally duplicated The file handle that was created is closed, the handle remains unique, and the processing , is returned to the user application as before.

The handling function of the file close service 3Eh is the function NETCLO3E Call to close the remote file. Function N E T CL C) S E contains a pointer to the local mensage buffer and the hand of the file to be closed. is given. Function NETCLO3E given from handle table Obtain the SID of the handle and store the message shown in Figure 19A in the local message buffer. page 160-form the CFI.

As mentioned above, the transport layer function TSEND sends messages via the virtual network. The function TRECEIVE is used to transfer the Used to receive reply messages from the bar.

When the transfer of function TRECE IVE is completed successfully, the reply message is stored in register AX. Bytes OAh and OBh of Tsage 160-RCFI (Figure 19B) are sent. It will be done. Reply message 160 - Error code in bytes 08h and 09h of RCFI If the code indicates an error, the carry flag is set and the Processing is returned to the user application. The error code indicates an error. If not, the local file handle of the table of file handle structures is is erased. The file handle and SID in the file handle table are set to zero. Set. Processing is then returned to the user application as described above. Ru.

The handle function of the write file service 40h is Iafn NETWRITE. The function NETWRITE, which calls , the length of the information and a pointer to the information to be written, and the local message buffer. A pointer is given. Message 1 generated by function NETWRITE 60-W is shown in FIG. 20A and includes a reply message 160 from the server. -RW is shown in Figure 20B.

Before considering the operation of the function NETWRITE, it is important to understand that the function NETWRITE Consider the compression mode used. Information written via virtual network Compression improves network performance. However, the function NETWRIT E compresses information without compression or by compression methods other than those described here. You can also write by In addition, other message sequence information, e.g. The general transmission message described above can also be processed using the following compression method. Ru.

In this example, information compression is performed using run-length limited (RLL) aspects. It will be done. The compressed buffer will have the following format:

Length (# of targets in the buffer) □　[Target 1]　□　(First target) □　[Target 2]　□　(Second target) □　[Target n]　□　(Final target) Each object i, where i=1 to n, has one of the following formats.

compressed target IRI Tag indicating target type (bytes) Length # Compressed bytes (words) Data Uncompressed data (word) What to uncompress ゛゛゛ Tag (byte) length indicating target type Size of uncompressed target (workload) ) Uncompressed data with the number of bytes equal to the data length When a compressed target occurs in the data stream and adjacent bytes are not identical, In the data flow, uncompressed objects are generated. For example, a database with the following data: If we use Intel's byte order for the buffer, Given that, the compressed buffer is 'R' 050001 'L' 06 00020304050607'R' 040002 have.

Sending or When writing, the data must be restored (extended) in order to read or receive. Must be. This mode of restoration (expansion) is the opposite of the mode of compression, i.e. It reads out the numbers of the objects and performs the appropriate action on each object according to the given flag.

Returning to the operation of the function NETWRITE, the code for the function NETWRITE is The handle given in the handle is used to obtain the SID of the write operation. Used as an index for the dollar table. The data length of the information buffer is The buffer offset may be larger than the length of the original message buffer. A pointer called a pointer tracks the data portion of the information buffer being transferred. first , the buffer offset is set to zero and the pointer is centered at the start of the data. will be carried out. After each transfer, a buffer offset, sometimes called the buffer length, is is incremented by the length of the data area of the local message buffer.

Therefore, if the buffer offset is less than the length of the information being written, The function NETWRITE determines the length of the local message buffer and the remaining information buffer. Compare the length of write data. The length of the write data remaining in the information buffer is , is greater than the length of the local message buffer, the length is unchanged.

However, the length of the remaining write data is less than the length of the local message buffer. If the length is also smaller, the length is set to the remaining data length.

Once the buffer length for the transfer is determined, write message 160-W 8 high header for the message is generated in the local message buffer. handle of the given file The dollar value is set in bytes 08h and 09h, and then the data transferred is as described above. compressed as shown. The length of the compressed data is the message pie) OAh and OB. h, and compressed data is given after OCh. Transfer function TSEND , is used to forward messages to remote nodes, and is used to forward messages to remote nodes when function TSEND is successfully Once completed, the transfer function TRECE IVE is called.

The function TRECE IVE completes normally and the reply message 16 shown in FIG. 20B is sent. When 0-RW is received, register AX contains the reply message bytes OAh and OBh is set.

If bytes 08h and 09h of the reply message indicate an error, the cache The leaf flag is set and processing returns to the user application as described above. It will be done. If no errors occur, the resulting value is compared to the buffer length.

If the result is less than the buffer length, the operating system When the memory becomes full, writing will stop and an error message will not occur, so the remote data vise is full. If such an error occurs and the buffer length If the result is greater than the result, the result is added to the current value of the buffer offset. This sum is set in register AX, and processing is performed by the user application as described above. returned to the section. Therefore, the user application must determine the returned length and the length of the information. Compare the values to confirm the occurrence of an error.

If the buffer length and the result of the reply message are the same, set the buffer length to the buffer offset. The buffer offset is incremented by adding it to the buffer offset, and the transfer process Iterates until the offset matches the length of the information. If the buffer offset is Once equal to the length of the information, the buffer offset is written to register AX and processed. control is returned to the user application in the same manner as before.

The function TRECE IVE fails to receive the reply message or the function If TSEND fails to send a message, the carry flag is set. The error code for the operating system network error is registered. Written to AX. Processing then returns to the user application as described above. be returned.

The handling function of read file service 3Fh is based on the function NETREAD. to call. NETREAD, a function similar to NETWRITE, has read A buffer pointer and buffer that stores the handle of the file to be read and the information to be read. length and a pointer to the local message buffer.

The message 160-READ generated by NETREAD is shown in Figure 21A. The reply message 160-RREAD from the server is shown in 21B. As shown in the figure.

The handle given in the call to the function NETREAD is the S It is used as an index of the handle table to obtain the ID. information The data length of the buffer may be greater than the local message buffer length. Therefore, a pointer called the buffer offset is the data of the information buffer being transferred. track the data part. Initially, the buffer offset is set to zero and the point The data is matched to the start position of the data. After each transfer, the buffer offset is Increased by the length of the data area of the local message buffer, which is called the buffer length. divided.

The 8-byte header of read message 160-READ is stored in the local message buffer. generated in F. The given file handle is Hyde 08h and 09h. The length of the data set and read is the message hide OAh and OB Set to h. The forwarding function TSEND forwards the message to the remote node. When the function TSEND is successfully completed, the transfer function TRECE I VE is called.

The function TRECE IVE completes normally and the reply message 16 shown in FIG. 21B is sent. 0-RREAD is received and bytes 08h and 09h of the reply message contain errors. If it does not indicate an occurrence, the reply message data will be restored to Information Hanwha. It will be done. Subsequently, the buffer offset is set to the restored data length. next , the more flags in bytes OCh and ODh of the reply message are examined and the remote node It is checked whether there is any more data to be transferred from the host. More flag is set If so, processing returns to the transfer function TRECIE. Moahu If Lag is not set, all data has been received and the register is A buffer 7 offset indicating the read data length is set in AX. Processing , returned to the user application as before. Function TRECEIVE If the reply message fails to be received, an attempt is made to receive it again.

However, if the reply message indicates that an error has occurred, register A In X, the received length indicated in hides OAh and OBh of the reply message is set. will be cut. This means that the error code is sent to the remote node's operating system. This is the case where this occurs. The carry flag is sent and processing is the same as above. , and then returned to the user application. Similarly, the function TSEND If the message transmission fails, the carry flag is set and the An error code indicating a network error in the system is sent to register AX. processing is then returned to the user application as before.

The handling function of Lseek service 42h is coded with the function NETLSEEK. file. The NETLSEEK function includes the handle and offset of the file to be accessed. Relation indicators and local mensage buffer pointer to show set value and seek mode. Inter is given. Hand given in call of function NETLSEEK is the index of the handle table to obtain the SID of the seek operation. It is used as

Next, the 8-hide message header of message 160-3EEK is the local message. A dibuffer is generated. The given handle is a byte of the message buffer. set to 08h and 09h, the given file offset is byte OAh is set to ODh and the given ration is to Hyde OEh and OFh. cent. Therefore, message 160-3EEK shown in Figure 22A is generated. It will be done.

Transfer machine n T S E N D sends messages to server node 160-3EE Called to send K. Upon successful completion of the function TSEND, the server To receive the reply message 160-R3EEK (Figure 22B) from the node. The transfer function TRECEIVE is called for this purpose. Transfer function TRECEIVE is correct It always ends and there is an error in the error code (hyde 08h, 09h) of the reply message. If indicated, register AX contains the reply message bytes OAh and The returned error in OBh is marked and the carry flag is set. Processing , returned to the user application as before.

However, if no error occurs in the seek operation, register A Hyde OCh and ODH of reply message 160-R3EEK are sent to X. The bytes OEh and OFh of the reply message are set in register DX. Ru. OCh through OFh in the reply message are new file offsets.

Processing is returned to the user application as before.

The handling function of service Chmoc43h is based on function NETCHMOD. to call. The service Chmod is used to set or get attributes of remote files. used. Therefore, the function NETCHMOD contains the SID, attributes, and Flags to set or get attributes, remote drive indicators, files The pathname of the file and a pointer to the local message buffer are given.

Using the information given, the function NETCHMOD creates a local message buffer. Message 160-CHMOD shown in FIG. 23A is generated. SID, Ji sex , function code (given flag) and bus length are in bytes 08h and 09h, O Set to Ah and OBh, OCh and ODh, respectively.

The path is copied to the local message buffer and based on the given drive indicator. It is then attached to the drive salt and cologne.

The forwarding function TSEND sends the complete message to the remote server node. used to send messages.

When the function TSEND completes normally, the transfer function TRECE IVE is called. and receive reply messages from remote server nodes. Function TRECE I When VE successfully receives the reply message shown in Figure 23B, register AX is , hides OAh and OBh of the reply message are set. Remote node behavior Upon successful completion of the operation, attributes are stored in these bytes. Samona (ba, this These bytes contain an error code. Error code byte o8h of reply message and o911 indicates an error, the carry flag is set; Processing is returned to the user application as before. On the other hand, these bars If the write does not indicate an error, register Cx is set to the value of register AX. processing is returned to the user application as before.

The function TRECE fails to receive the reply message, or the function TS If END fails to send the message, the carry flag is set; Error code of operating system network error in register AX is set and processing is returned to the user application as before.

The function 09h of the current directory acquisition service 47h is to specify a remote drive. This is used to acquire the current directory as the current directory. Therefore, Ka Handling function of rent directory acquisition service is functional NETGETD I When you call R, the SID, remote drive indicator, and remote drive path salt are displayed. , and a pointer to the local mensage buffer are given to the function NETGETD IR. available. Using the given information, the function NETGETD The message message 160-GDIR is shown in Figure 24A on the 7th floor. generate da. Mensage Hyde 08h and 09h are given remote dry The display indicator is set. The reply pointer is initially placed at position OEh in the reply buffer. Set. Function NETGETDIR Transfer Jig(7)TSEND:1- , sends a message to the server node.

When the function TSEND ends normally, the function NETGETD IR is transferred to the transfer function TRE. Call CEIVE. TRECETVE replies message 160-RGDI Upon successful reception of R (Figure 24B), register AX contains bye 1-OAh and O. Bh extended error code is set. Reply message error code If bytes osh and 09h indicate an error, the carry flag is set. , processing is returned to the user application as before. On the other hand, these If the byte does not indicate an error, the path salt of the reply message is as described above. The path salt given when calling the function NETGETD IR is copied like so: Ru.

If the function TRECEIVE fails to receive the reply message or the function TSE If the ND fails to send a message, the carry flag is set and the operator The error code of the operating system is set in register AX. Processing is as described above is returned to the user application as well.

Session termination service 4Ch is processed locally, so the mensage structure is Not necessary. The handling function of the session termination service is the function NETTER. Call MI NATE and this function NETTERMI NATE is In the case of an operational connection with a remote node, the above-mentioned function VDI 5CONN Call ECT. After calling the function VDISCONNECT, the virtual Work management functions are identified as non-operating and are handled by the operating system. Return to

Head detection service 4Eh detects the first file that matches a particular file specification. put out This service is used by the operating system to The structure defined as DTA is used. Therefore, all computers In order to properly perform this function, first check the operation of the virtual network management function. Before starting the local interrupt 21h, the head detection service must be executed. Detects when something is missing.

This operation effectively initializes the DTA and makes it available for use.

The handling function of the head detection service is based on the function NETF INDFIR3T. to call. The function NETF INDF IR3T has SID and file attributes. , an indicator of the remote drive, a path and a pointer to the local message buffer are given. It will be done.

Function NETF INDF IR3T sends message 160-F shown in FIG. 25A. F, start detection service in the same way as described above for the 09h service of other functions. Creates a standard 8-hide header for services.

Once the message header is formed, pass salt is added to the drive salt to create a local message header. The colon is copied to the page header.

The drive salt is the drive identifier given to the function NETF INDF IR3T. determined by File attributes given to function NETFINDFIR3T The bus length is sent to hides 08h and 09h of the local message buffer. is sent to the hides OAh and OBh. As mentioned above, the virtual network TSEND is used to send a message via If it is finished, the function TRECEIVE is called and the service indicated by SID is A reply message 160-RFP (Figure 25B) is received from the bar.

If the function TRECEIVE successfully receives a message, the return value is stored in register AX. expanded into bytes OAh and OBh of communication message 160-RFP (Figure 25B). The error code is set. Reply message error code byte 08h , 09h indicates an error, the carry flag is set and processing continues. , returned to the user application as before. On the other hand, these bytes If no error is indicated, the result of 28h bytes of the reply message is displayed in the calendar. The data is copied to the client DTA and processing returns to the user application as before be done.

If the function TRECE IVE fails to receive the reply message or the function T If SEND fails to send the message, the carry flag is set and the operation The rating system error code is set in register AX. The processing is Returned to the user application as before.

Next item detection service 4Fh detects the next file that fails with specific file specifications do. This service is generally used after the head detection service 4Eh.

Therefore, the search process has already been initialized. For this reason, the following detection service When the function NETF INDNEXT is called, the function NET F INDNEXT contains only the SID and local message buffer pointers. Given. The function NETF I NDNEXT sends the message shown in Figure 26A. 160-FN in the same manner as described above for the 09h service of other functions. Forms a standard 8-byte header for head detection services. To the message Once the leader is formed, a copy of the DTA returned by the head detection service is Copied to the sage data area. The operation after message formation is as described above. This is the same as the operation of the function NETF INDF IR3T after the occurrence of a message. the above The operation of is incorporated herein as part of the description.

The rename service 56h changes the file name of a file on a remote drive.

The handling function of the rename service calls the function NETRENAME. . The function NETRENAME contains the SID and remote name of the file to be renamed. interval drive identifier, file name change file bus salt, modified file The name path and a pointer to the local message buffer are given. Function NETRE NAME is added to the message 160-REN shown in FIG. In the same way as described above for the service, the standard 8. Form the hide header. Once the message header is formed, the path salt dries. A colon is added to the buffer and a colon is copied to the local message buffer. Pass salt is It is assumed to end with zero. Drive salt given to function NETIIIENAME determined by the drive identifier of the file whose name you want to rename. child The length of the path salt is located in bytes 08h and 09h of the local message header. be done. Bath salts for renamed files are added to drive salts and colons. is copied to the local message buffer after the first pass. No bath salt It is assumed that it ends with . The drive salt is a file given to the function NETRENAME. Determined by the drive identifier of the renamed file. this path The length of the salt is set to the hides OAh and OBh of the local message buffer. .

As mentioned above, sending the message 160-REN over the virtual network If TSEND is used for routers and TSEND completes normally, ECE IVE is called and reply message from the server indicated by SID 160-RREN (Figure 27B).

Function TRECErVE successfully receives message 160-RREN (Figure 27B). If successful, the byte OAh of reply message 160 RREN is stored in register AX. and OBh is set to 1-. Error code byte of reply message 08h, 0 If 9h indicates an error, a reply message (27th B) is sent to register AX. Bytes OAh and OBh in the figure are set. carry flag is cent, Processing is returned to the user application as before.

If the function TRECE IVE fails to receive the reply message or the function T If SEND fails to send the message, the carry flag is set and the operation The rating system error code is set in register AX. The processing is Returned to the user application as before.

Function 09h, get/set file data and time 57h, is used to obtain/set file data and time of a remote file. Acquisition of data and time for and cent of data and time to remote files It is used for people.

The handling function of service 57h calls the function 'NETDATATIME'. do. File handle, function indicator to distinguish between get and set operations , data and time for cent operations and local message buffer pointers are function is given to the function NETDATAT IME by the handling function. Function N ETDATATIME is the same as described above in the 09h service of other functions. The message 160-GT shown in Figure 28A for service 57h is marked as form a standard 8-byte header. File handle using the given handle A SID is obtained from the dollar table. Once the message header is formed, the handle Dollars are cented at 08h and 09h, and function indicators are set at OAh and OBh. and the time is set to OCh and ODh, and the data is set to OEh and OFh. is set to

As mentioned earlier, the TSE is used to send messages over a virtual network. If ND is used and TSEND is successful, the function TRECEIVE is Return message 160-RGET from the called server indicated by SID (Figure 28B) is received.

Function TRECEIVE receives message 160-RC; ET (Figure 28B) was successful, and the error code high l-08h and 09h of the reply message indicates an error. If not indicated, register AX is set to zero. In register DX The data of bytes OCh and ODh of the reply message are set. Register C X is set to the time of bytes OAh and OBh. The process is as described above. , the error code bytes 08h and 09h of the user application cage reply message are If an error is indicated, the reply message (Figure 27B) is stored in register AX. Bytes OAh and OBh are set.

The carry flag is set and processing is done by the user application as before. will be returned to.

Function TR, ECEIVE fails to receive reply message or function TS If END fails to send the message, the carry flag is set and the operator The error code of the programming system is set in register AX. Processing is done before is returned to the user application as described above.

The handling function of the 5ERVE service of the function OAh is the application layer Call the function VSERVE. The handling function is the communication with the server node. speed, timeout value indicating the time interval to wait for a reply from a remote node, virtual network The boats used to form the work, the data segment register values, and the The register values of the other segments and some message handlers and steps as described above. A pointer to a callout structure, including a task event handler and a user exit handler. give data. Bytemap 160-C3RTRUC for callout structure T is shown in FIG.

The function VSERVE first indicates that the virtual network management function is operational. Check. If the virtual network management function is not operational, the flag is returned to the user application and the virtual network management functionality is not working. It shows that something is wrong.

After checking whether the virtual network management function is running, The status is set to operational and the function VLISTEN described above is called. , waits for a connection request from a remote node. When a connection request is received, the function VLISTEN provides the SID and VSERVE to the value given by the function VLISTEN Cent SID to equal value. If SID is not greater than zero, then The server status is set to non-operational and the SID is set to user application. returned to the application.

If the SID is greater than zero, the function ~5ERVE first set amp and intercept. Serious error handler For those skilled in the art The function will be stopped in a well-known manner. For example, R. Duncan : (W sound collection)) rMS-DO3 Encyclopedia (The MS-DOS Ency clopedia) J Phycrosoft Press, Washington, D.C. See pages 390-398 (1988).

Function VSERVE then calls function SERVERMAIN. Function SE RVERMAIN is described in more detail below. The function SERVERMAIN is , performs one of the following termination operations.

(i) When an error other than a timeout is received (i) Either transmission When an error occurs (i i i) the user exit handler returns a non-zero value. When I did (iv) Any server function, including general sending and receiving, returned an error value. Time Function V from function SERVERMAIN to reset the severe error handler When returning to SERVE, the server status is set to inactive. S.E. The results of RVEMAIN are returned to the user application.

Function VSERVE allows function SERVERMAIN to contain SID, data segment register values, remainder segment register values, and callout structure A pointer is given. The function SERVERMAIN is the function disk of the server. Puncher is available, and this function is performed by puncher from the first computer which is a remote node message, locally performs the action specified by the message, and sends a reply message. generate a page and send it to a remote node. Therefore, the function SERVERMA■N is When function VSERVE is called by user application 201 Then, a reply message is generated for each service of function 09h and function OAh mentioned above. Ru.

At the time of entry of function SERVERMAIN, the current DTA is acquired. , the variable stored in that address. Kolua A status event with a callout table defined and a callout table defined. If the transaction handler has a non-invalid entry, in this example, the command The address of the status event handler in the callout table contains the event code. Luout is set. The event callout register DS is The event callout register ES is set to the register value of the data segment. is set to the value of the given remainder segment. Steps according to one embodiment The byte structure of the data event handler is shown in FIG.

The function SERVERMAIN then calls the user in the callout table (Figure 30). Make sure that a user exit handler is defined. A user exit handler is If the function SERVERMAIN is defined, the function SERVERMAIN is Call La. If the user exit handler returns zero, then Processing continues. User exit flag if a non-zero value is returned is set, and processing moves to server error termination processing. Server error termination process performs the cleanup operations required to close the server and clean up the forwarding layer. Call function TD I 5CONNECT. Function TD l5CONNECT Once completed, an error is returned to the user application.

If no user exit handler is defined or the user exit handler is If it returns zero, the function SERVERMAIN returns the function TRECE of the transport layer. Call IVE. Function TRECE IVE generated a timeout error Then, the function SERVERMA IN handles the process defined by the user exit handler. The process returns to the step of determining whether or not it is true. Function TRECE IVE is another error If an error occurs, the process moves to the server error termination process described above.

Function TRECE When the IVE successfully receives a message, the first 4 bytes are checked and the identifier is It is confirmed whether or not it indicates that the message was sent. appropriate identifier is not defined, processing is determined whether a user exit handler is defined or not. The process returns to the determination step. Otherwise, the function SERVERMAIN is Call the application layer function VSTATUS described above. Function SERVE After RMAIN calls the function VSTATUS, the status event hand Once the controller is defined, the status event will be handled by the address of the status buffer. Handler is called.

If the sixth function number of a message does not indicate a partial message, The server function corresponding to the sixth hide of Tsage is called. these The functionality of is explained in more detail below. When the function completes, the called server The length of the reply message generated by the It will be stored as soon as possible.

If the 6th byte of the message indicates a partial message, it will not be remembered. message length is set to zero. Some messages in callout structure If a handler is defined, the function SERVERMAIN handles general messages. The message handler is set to register CX, the message data length, and the local register BX. Message buffer size, register vs. DX: AX message data address Call by giving a response. General message handlers service requests. updates the local message buffer with the new data, and Returns the length of the message buffer or the error code in register AX. user Some applications must provide message handlers, so the actual The actions taken by the user application or some message handlers are defined by the However, the general message handler The message of Figure 9B is decomposed to generate information for the message of Figure 9B.

The length returned by the general message handler is stored for processing described below. will be paid. If the length is greater than zero, the generated reply meso set of 8 hides. The message header (Figure 9B) is formed in the local message buffer and the message header (Figure 9B) is The length is set in hides 08h and 09h of the partial reply message. remembered Length is updated with the message length.

Memorized after processing by a specific server function or some message handler If the length specified is greater than zero, the function SERVERMAIN calls the handler is defined, call the status event handler again. C After calling the controller, the function TSEND of the transport layer is called and the function TSEND is returned to the remote node. send a message. If the function TSEND returns an error, processing: The process moves to the server error termination process described above.

Either no status event handler is defined or the function TSEND is returns no error, and the stored length is less than zero, Processing moves to the server error termination process described above. On the other hand, if the memorized length is If the value is greater than is returned to the check step.

Each function of function 09h (Table 3) is called by SERVERMAIN. It has Untabert. Generally, these features separate sent messages. The information in the message is used to set the register to the appropriate value, and the local Generates an interrupt 21h to the operating system. The function is interrupt 21h. The information thus supplied is used to generate the function's reply message. especially, Select disk OEh, secure free space 36h, create directory 39h, disk Directory deletion 3Ah, directory change 3Bh, file deletion 4th, mode change additional 43h, ]5eek42h, file name change 56h and data and time acquisition 5 7h fails in this case. Other functions 09h services are explained below. requires additional action.

The functions of the server serving Creation 3Ch and Oven 3Dh will be Disassemble the message and start the appropriate interrupt 21h service. 8-byte return mail After creating the message header (Figures 17B and 18B), the operating The results returned from the messaging system are set in bytes OAh and OBh of the reply message. will be cut. If the operating system sets the carry flag In this embodiment, the error code bytes 08h and 09 of the reply message are h is set to 1. If the carry flag is not set, the local flag is An empty slot is detected in the file handle table. The SID of the empty slot is This is set to the SID of the incoming message, and the slot contains the operating system. The result returned by the system, ie the handle, is set.

The size of the reply message is returned to the function SERVERMAIN.

Server function close 3Eh service is creation 3Ch and oven 3Dh service. The function is similar. The difference is that this function corresponds to the SID of the incoming message. The process is to search the local file handle table for the handle to be used. to S.I.D. If the corresponding handle is a legal handle, both the SID and the handle is set to zero.

The server function is 3Fh service, the data to be read comes in a long time. OAh and 0Bh (Figure 21A)) of the message sent are the data of the reply message. It is somewhat more complicated because it can be larger than the area. Therefore, this feature Parses the incoming message to obtain the file handle and the length of the data to read. Ru. If the lead length is greater than the reply message data length, the lead is The read data of all reads except the read of the reply message data area Split into a series of leads to fill length. operating system The length of the read returned by is added to the length of the previous read, and the stored value and the returned value are Comparison of read lengths from sent messages now indicates that the read is complete. will be saved. The read data is compressed and localized as described above. stored in the image buffer. The length of the compressed data is the byte of the reply message. Set to l-0Ah and OBh.

The operating system uses a carry flag to indicate an error on a read. In this example, when the reply message (Figure 21B) is set, Bytes 08h and 09h are set to non-zero values. operating system If the system has no errors and there are more bytes to read, the more flag is set. set, but the more flag is not set if there are no more bytes to read. do not have. Next, if a handler is defined, the status event handler is called. After calling the handler, the transport layer function TSEND returns the Called to send a message to a remote node. Function TSEND is error If the error is returned, the process moves to the server error termination process described above. function If TSEND does not return an error, this process continues until all bytes have been sent. The principle is repeated. Once all bytes have been sent, the zero length function SERVERM Returned to AIN.

The 40h service (server functionality) is used to respond to incoming messages (Figure 20A). ), restore the data as described above, and start writing interrupt 21h. . The function then sends the reply message (Figure 20B) to the local message buffer at 8. form the header of the byte and operate on the bytes OAh and OBh of the reply message. Sets the result given by the testing system. operating system If the carry flag is set by Code is set to a non-zero value. The length of the reply message is determined by the function SERV. Returned to ERICEMAIN.

The server's function directory acquisition 47h service is to handle incoming messages ( Figure 24A) is disassembled and the directory acquisition service of interrupt 21h is started. The function then forms an 8-byte header of the reply message (Figure 24B), High 1-OAh and OBh of the reply message are given by the operating system. Cent the results obtained. Carry flag by operating system is sent, the error code of the reply message is set to 1. , path length (high) OCh and 0Dh) are set to 2 in this embodiment.

If the carry flag is not sent, the bus is sent to the local message buffer. , and the path length is set to an appropriate length. The length of the reply message is Returned to function 5ERVERI CEMA IN.

The first detection 4Eh and next detection 4Fh functions of the server are the server functions described above. The function is similar to that of interrupt 40h, the difference being that the appropriate interrupt 21h is called and Shown in the sixth hide of the incoming message (Figures 25A and 26A) This is the point at which the service is executed. Also, as mentioned above, DTA It is copied to the data area of the message (Figures 25B and 26B).

Regarding the service of function OAh, the general message handler is V SEND and VRECE IVE (7) Process the message. VD I 5 In response to the CONNECT message (Figure 8), the opened file is Checked against the message's SID. Opened file is SID , the file is closed. Next, the function TD of the transfer layer I 5CONNECT is called and zero length is function SERVERMAIN will be returned to. A connection has already been made, the status is local command only, Since no messages are sent over the network, the server Does not respond to ISTEN and VSTATUS. VRESET simply resets and the server does not respond because nothing is promoted over the network.

The function MAPDR ■ on the server for service map FDh is page (Figure 10A). Accessing a remote drive on a virtual computer Two problems can arise when accessing. The first is the requested driver. Eve is the case when the remote node is not physically present. Second, remote node The MNO operating system prohibits the operation of virtual network management functions. This means that there is a possibility that procedures will be taken. For example, originally, most IBM PCs A comparable microcomputer only has one floppy disk drive. but the drive gives address as drive A or drive B I was able to do it. In this case, drive B does not physically exist. Drive Live A used and the command is given to drive B, the operating system inserts a disc into drive B and requires a keystroke to continue. Ru. Obviously, such a request of a remote node will limit the operation of the virtual computer. Ru.

Therefore, the virtual network management function can be used to manage remote drives that are not physically present. Do not allow pings.

Upon receiving the incoming message 160-MAP (Figure 10A), the server function MAPDRV sends the 8-byte header of reply message 160-RMAP to a local message. message buffer, and bytes 08h and 09h of the reply message are invalid. Set it in the drive. Function MAPDRV then sends message 160-MAP The specified drive physically exists and cannot be removed on the remote node. Make sure that.

If these two conditions are met, bytes 08h and 09 of the reply message h is set in the requested drive. Similar behavior occurs when the drive can be removed. This is also done if the drive is enabled and the drive is present. reply message length is returned to the function SERVERMAIN.

The transport layer 170 (FIG. 5) supports functions 120, 130 and 130 of the application layer 160. or by a function of the data link layer 180. In this example In this case, the function that calls the transfer layer function is called the transfer layer function and °C” format from the right. Interface via stack frame with parameters transferred on the stack to the left - Face. (Here, lcl is B, Kernig. han) and Dee, Lisochii (D.

Ritchie) "C Programming Language" 1st edition, Prentice Hall, New – Eaglewood, Chassis, as shown in Criso, (1978). It shows the C programming language. ) The parameters pushed into the stack are shown in Table 4. listed. The transport layer 170 is 64 km from the application layer 160. It can process messages that are smaller than a byte. The transfer layer 170 The message from the application layer is broken down into multiple packets. Transfer layer 170 uses a CRC-16 checksum, as described in more detail below. authenticate each packet to the remote node indicated by the session identification number (SID). Confirm accurate transfer. In particular, the transport layer prefixes each packet with a header and a CRC - Abend 16 checksum. The packet is then sent to data link layer 1 80 for forwarding to a remote destination.

Transfer 1170 detects errors in the communication link and handles them as necessary. retransmit the packets that are present. Forwarding N170 allows you to send large messages as needed. Split packets into sentences. That is, if a predetermined number of errors occur, or the message is larger than the maximum size that can be sent in a packet. This is the case. In one embodiment, the message is a packet of 4 kilobytes or less. It is divided into two parts. In general, the maximum packet size determines how accurately the packet can be sent. It must correspond to an error checking scheme to confirm that the

If an error occurs while transmitting a packet, the forwarder N170 retransmits the packet a predetermined number of times. Send. When a packet cannot be sent without an error within the specified number of transmissions In response, transport layer 170 halves the size of the packet. of the packet If the size is 256 bytes, which is the minimum size in this example, the data Link layer 180 reduces line speed and packet size to the maximum in this embodiment. Increase to 4KB. In addition, if a transfer error occurs, the packet size Transfer layer 170 aborts the transfer process and returns the caller's machine to Returns an appropriate error code.

The transport layer 170, as well as the data link and application layers, It can be designed to be able to handle various motions. But long In one embodiment, the data link layer 180 supports multiple sessions. To facilitate further development, the maximum number of sessions is limited.

In this embodiment, the transfer layer 170 includes TL T5TEN, TCONNECT, T DISCONNECT, TSTATUS, TRESET, TSEND and TRE It supports a set of seven functions called CE IVE. depending on the caller's capabilities. the input information that needs to be provided to each transport layer function and the results returned to the caller. are shown in Table 4.

Table 4 Transfer layer function TLISTEN At entry Speed: baud rate of the line divided by 100 (to allow selection of optimal speed 0) Timeout: Number of clocks to wait for data (0 = infinite) Port: 0 = C OM1. l=C0M2. ．． ．． .

Upon return SID or negative error code returned TCONNECT At entry Node ■D: Identification of remote node Speed: baud rate of the line divided by 100 (to allow selection of optimal speed O) Timeout: Number of clocks to wait for data (0 = infinite) Port: O, C O gate L1, C0M2.　,,.

Upon return SID or negative error code returned TD　I　5CONNECT At entry SID: Session identification Upon return An error code indicating session or failure is returned.

TSTATUS Length: supplied buffer length 5tatbuf: When the pointer of the supplied buffer returns The buffer supplied by the calling function is filled with virtual network data structures. TSTATUS returns zero or an error code.

Virtual network data structure The ID number of the local node in the active session array of the SID data structure (each one in the middle session) SID data structure Session ID, number of sent bytes, number of received bytes, number of sending errors, number of receiving errors, Current baud rate, SAI version, VNA version, application Msg derived language TRESET At entry Length: Length of supplied working buffer Buffer: Supplied working buffer pointer 5END At entry SID: Session identification Length: Message length Message: Message pointer Upon return return error code THECE IVE At entry SID: Sension identification Length: Received message length Buffer: Received message pointer Timeout: Number of clocks waiting for data (0 = infinite) Transfer layer function error is negative Returned as a value to the user application. Successful termination of the transfer layer function is positive ( indicated by a return value (usually zero). Examples of return values are shown in Table 5. Ru.

Table 5 Return value of the result of the functional operation of the transfer layer function〇〇　−Function completed normally -1- Breaking the connection -2-Unknown node ID -3- Incorrect line speed -4- Sending/receiving error -5- Timeout error -6- Unable to connect to specified node -7- Buffer too small -8 - Invalid buffer address -9- Invalid session ID -10- No sessions available Transfer layer function TL I 5TEN, TCONNECT, TD I 5CONNEC T, , TSTATUS and TRESET are virtual null operations, and these functions Each of these is simply a passage between the application layer 160 and the data link layer 180. and it works. In general, application functionality co-opts corresponding transport layer functionality. If a call is made, the transport layer function calls the corresponding data link layer function. subordinate Therefore, each transport layer function supports the information required by the application layer functions. must be For example, TSTATUS adds a virtual Stores status information regarding network management functions. given to the buffer The structure is the application layer status data structure 160-3TA described above. It is equivalent to TSRUCT (Figure 29).

The above functions act as a path, but the transport layer functions TSEND and TRECE The IVE calculates the number of packets, the size of each “packet” and the forwarding of each packet to the remote node. control the transfer rate used by the data link layer 180 for transmission. In particular, applications If the function of the transfer layer calls the function TSEND of the transfer layer, the function shown in Figure 32 is The specified action is performed.

The transfer layer function TSEND (TSEND) 170-3 first transmits S ID check 170-31 is performed. SID check 170-3l is TSEN Check whether the SID given in the call to D is valid. S.I.D. is non-zero, less than the maximum allowed SID, and one of the SIDs is related to Sension. If the SID is attached, the SID is valid. Given to TSENDl 70-5 If the received SID is invalid, processing moves to SID error return 170-32. In this example, the error return code is set to -9 and the application is returned to the calling function in the application layer 160 (FIG. 5).

If the SID check 170-3l determines that the SID is valid, the point Message point given when Tachek 170-55 calls TSEND Determine whether the data is defined, that is, whether it is null. The message pointer is , the starting point of messages generated by the application layer 160 (FIG. 5). identify. Therefore, the message pointer is TSEND 170-3 (third (Figure 2) must be defined in order to detect and then process messages. stomach.

Pointer check 170-35 determines that message pointer is not defined In this case, the process moves to pointer error return 170-34, and in this embodiment If the error return code is -8, the application layer (Figure 5) ) is returned to the calling function.

Pointer check 170-35 (Figure 32) indicates that a message pointer is defined. If it is determined that there is, the process moves to the reset counter 170-36. less than In more detail, the message failure error counter will affect the transfer speed and packet size. used when determining the block size of the cut. Therefore, the reset counter 170-36 resets the message bad error counter to zero.

After resetting the message bad error counter, the number of packets 170-37 is the highest. First, use the message length given when TSENDl is equal to 70-5. Calculate the number of packets to send to the remote node. Here, from the application layer The message length is up to 64 kilohids, but as mentioned above, transport layer 1 70 (FIG. 5) is capable of transmitting packets of up to 4 kilobytes. Pake The number of packets 170-37 (Figure 32) is calculated by dividing the message length by the current packet length. and round the result up to an integer value. Therefore, the length of the message is 64 kilobytes, and the current packet length is the maximum 4 kilobytes. , the number of packets 170-37 means that 16 packets must be sent. calculate.

Buffer offset initialization 170-38 directs processing beyond the local message buffer. Set the buffer offset to zero to start from the beginning. buffer off Set the offset to the local message buffer data for the next packet. shows. After buffer offset initialization 170-38, check for available packets. A check 170-39 is performed to determine if there are other packets available. Ru. When all packets have been successfully sent, available packet check 17 0-39 transfers processing to return 170-3IO and transfers control to TSEND 17 0-3 (Figure 32), the application returns a return code indicating normal completion. The function returns to the calling layer 160 (FIG. 5).

Available packet check 170-39 determines that the packet is usable for transmission. If so, processing moves to packet formation 170-Sll. Packet creation 170 - The Sll first creates a header for the packet. In one embodiment , the header 170-MH is 14 hides long. first 2 Hyde is a basic sign such as "■1 Pa".The third and fourth Byte has the identity of Mensage. The transport layer uses a small Since at least two blocks are transferred (sending and receiving), the identity of the message is This is used to identify which packet the sent block is. In one example In this case, message identification is performed continuously each time TSEND processes one packet. increased. The fifth and sixth bytes are of the packet source node. The seventh and eighth hides are the identification of the node to which the packet is sent. It is. All packets forming the message calculated with packet number 1'70-37 The number of packets transferred is stored in the ninth and fourth bytes of the header. The count is stored in the first and thirteenth bytes of the header. header tenth The third and fourth bytes indicate the length of the data included in this transfer packet. used for transfer.

Once the packet formation 170-3ll completes the formation of the header, the data for the packet is data is copied from the local message buffer to the transmitted packet, i.e. packet 17 It is stored in the storage area of 0-MD. Once the data is copied, packet formation 17 0-311 calculates CRC-16 checksum, calculated CRC-16 Add checksum to the end of the data. , in the second embodiment, inverted CR A C-16 generation function is used to obtain the CRC-16 checksum.

J. Campbell, [C Program for Serial Communication] Logging Gaitoco Chapter 19, Howard W.

Sam's & Company, Indianapolis, Indiana, 535- See page 550 (1987). Therefore, the packet creation 170-3ll is header and part of the message of the application layer 160 (FIG. 5). It consists of a data field and an attached CRC-16 error code.

When a packet is formed in packet creation 170-3ll (Fig. 32), the packet is The packets are provided for packet transmission 170-312 as detailed below with respect to FIG. It will be done. Generally speaking, packet transmission 170-S12 is performed at the data link layer. Calls the function DSEND of , and provides the packet to be sent. data link layer 180 sends a normal completion code or an error code to the packet transmission 170-312. return. Error checks 170-313 indicate whether an error code was returned or sent. Determine whether the process ended normally. If the transmission ends normally, check for errors. 170-S13 transfers processing to the available packet 170-39 and processes the next packet. or the application layer 160 (FIG. 5) also returns. However However, if the error check 170-513 (Fig. 32) detects an error, , processing moves to menosage defect check 170-314.

The error returned from packet transmission 170-312 is a message failure error. If so, processing moves to count check 170-315 and a message bad error is detected. - Determining whether the message failure count value shown on the counter is smaller than 5. will be held. If the message failure count value is less than 5, the count check Check 170-315 transfers control to bucket number 170-37 and re-executes the transfer process. go

Count check 170-515 determines that the message bad count is 5 or more. If the packet number set 170-518 is set, processing is moved to the packet number set 170-518, and the packet number set 170-518 is Set the number to 1.

This branch modifies the transfer parameters before attempting a retransmission. meso message If not defective, message defect check 170-314 and packet count set 170-316 transfers processing to hardware error check 170-316. .

Any connection broken or bad node ID sent packet 170-312 If returned by , hardware error checking 170-516 processes is moved to error code return 170-319. However, the connection is broken. If not, and no bad node ID is returned, processing proceeds to block size check. Moving on to check 170-320.

If the block size is larger than 256 bytes, processing will change the block size. Moving on to 170-324, the current block size is halved. block rhinoceros If the block size is 256 bytes, block resizing 170-320 is Move to speed change check 170-521 to determine whether the transfer speed can be reduced. Ru. If the virtual network is formed through a modem or the lowest speed In this case, the transfer speed cannot be changed, so the speed change check 170-S21 is not processed. Transmit/receive error returns 170-322 and send errors to the application. return to the calling function in the control layer. If speed can be decreased, check speed change 1 70-321 calls speed reduction 170-323, explained in more detail below. The data link layer function DRESET is used to reduce the transfer rate.

Both speed reduction 170-323 and block size change 170-324 are controlled by is moved to retransmission 170-325. Retransmissions 170-325 are buffer offsets is reset to the starting position of the packet in which the error occurred, and the process returns to packet creation 1. Return to 70-3 liters. Therefore, the transfer parameters have been changed and the transfer has no transfer errors. It is restarted from the point where it occurred. Changing forwarding parameters while a message is being forwarded is a feature not available in prior art systems. In the conventional technology , the transfer parameters are determined at the connection point, and these parameters are used for all transfers. was used.

Packet transmissions 170-312 (Figure 32) are shown in more detail in Figure 34. Ru. Error counter initialization 170-3P1 first initializes the transmission error counter and data. Set the Link Transmit Error Counter to a selected value, typically 3. sending error After counter initialization 170-3PI, the process is data link transmission 170-3P2 , and the function DSEND of the data link, described in more detail below, is called. It will be done. Function DSEND forwards the packet to a remote node. Data link transmission 1 70-3P2 moves the processing to data link transmission error check 170-3P3. , check whether a data link transmission error has occurred. Data link transmission error If no error has occurred, processing moves to data link reception 170-3P9. , if an error occurs in data link transmission 170-3P2, The transmission error counter is decremented by the transmission error counter decrement 170-3P4. Ru. Hardware error check 170-3P5 indicates a data link transmission error. Determine whether it is a connection disconnection error or a bad node ID. these errors In either case, processing is transferred to error return 170-3P6 and send 1 Returned to 70-312.

However, hardware error check 170-3P5 shows neither error. If not detected, the data link transmission error counter It is decremented by Decrement 170-3P7. Connection check 170-3P8 is error - Determine whether the counter value is zero. Count check 170-3P If 8 determines that the error count is zero, the process functions as described above. Return to error return 170-3P6. However, the count check If it is greater than zero, processing returns to data link transmission 170-3P2. , the packet is retransmitted.

If the data link transmitter 170-3P2 transmits the packet without error, , as described above, the data link transmission error check 170-3P3 Data link reception 170-3P9, header, return code and CRC- Wait for reply message from remote node with 16 checksum. data When the link receiver receives a reply message, error check 170-3PIO resets. Analyze the turn code to determine whether a data link reception error has occurred . When a data link reception error is detected, the error check 170-3PIO Branch to error return 170-SP15, set error code, and read data. The link is cleared and processing is moved to the decrement counter 170-3P19. But long If the error check 170-3PIO does not detect an error, the process Move to RC sum check 170″-3PII.

CRC sum check 170-3PII is CRC-16 checksu of header Calculate m and use the calculated checksum and data link reception 179-3P9. and compare the received CRC-16 checksum. CRC-16che If the cksums are not the same, processing returns CRC error 170-3P1. 6, clearing the data link is called, and the error code of the CRC error is displayed. is set, and processing then moves to counter decrement 170-3P19. CRC service If system check 170-3PII detects no errors, processing Go to page ID check 170-3P12.

Packet to which message ID was transferred from data link reception 170-3P12 If the message ID does not match, the process returns a return defective message 170. -3P17, where the error code of the defective message is set. Finally, Turn code 170-5P13 is return from data link reception 170-5P70 If the code is smaller than zero, the process returns an error code. 170-318, and then process the counter decrement 170-3. Move to P19. Error check 170-3PIO117〇-3PII, 170- If 3P12 and 170-3P13 do not detect an error, the process returns 1. The process moves to 70-3P14, and a return code indicating normal completion is returned to the sender 170-3. .

When control is received, counter decrement 170-3P19 decrement the data link error counter. Decrement the counter and transmission error counter and check the control count 170-3P Move to 20. If the error count is greater than zero, the process is data link transmission 170-312, and the transmission process is repeated. On the other hand, count check 1 If 70-SP20 detects zero, processing returns to error return 170-3P21. and returns the error code to the transmitter 170-3. In one embodiment, process 1 70-3P9 to 170-3P1B are packets after error check 170-33. Included in the separate functions called by Send 170-312.

As mentioned above, TSEND responds to the remote node's TRECE IVE. T The RECEIVE function is similar to the TSEND function and is shown in Figures 35 to 37. has been done. As shown in FIG. 35, the function TRECE IVE of the transfer layer 170 is , first perform SID check 170-R1 and pointer check 170-R5. Ru. These check operations are carried out by the SID check of the transmission 170-3 (Fig. 32) described above. This is similar to check 170-3l and pointer check 170-35, and these The description is incorporated herein by reference. Pointer check 170-R5, processing is transferred to the packet receiver 170-R6. Regarding packet reception 170-R6, the third 6 will be explained in more detail below. Packet receiving 170-R6 receives data. Call the link function DRECEVE. bucket) reception 170-R6 is the reception data data length, header, CRC-16 checksum and errors detected. Provide error code in case. Error check 170-R7 is for packet reception 170-R6 determines whether an error has occurred. If no errors are detected Copy data 170-R12 copies the received data to the receive buffer. , the process is moved to reply transmission 170-313.

Reply sending 170-R13 will be described in more detail with respect to FIG. Briefly explained Specifically, the reply sending 170-S13 sends the packet of information to the data link receiving 17. 0-SF3 (Figure 34). Reply transmission 170-R13 (Figure 35) If an error occurs Error check 170-R14 processes, packet reception 17 whose function was explained earlier Pass to 0-R6.

If no error has occurred in reply transmission 170-R13, error check 1 70-R14 passes control to the final packet chenok 170-R16 If the point has been processed, transfer the processing to return 170-R18 or Transfer to Furuchesoku 170-R17. If the receive buffer is full, processing Moved to buffer error return 170-R19, buffer full error is changed to app is returned to the calling function in the application layer 160. if the buffer is not full Ha nofa full check 170-R17 process packet reception 170-R6 Also get the next packet from TSEND.

In the above description, the error check 170-R1 is the packet receiving 170-R. It is assumed that no errors were received from 6 onwards. Error check 170-R7 detects an error, processing is transferred to increment counter 170-R8, and the receiver After incrementing the error counter, control is passed to the hardware error check 170-R. 9, packet reception 170-R6 indicates bad node ID, disconnection or time If the output error is returned, hardware error check 170-R9 restarts. Processing moves to turn error code 170-R15. Otherwise, the hardware The error check 170-R9 processes the defective message to the 170-RIO. move. If no message failure error occurs in packet reception 170-R6, Message Bad Check 170 - RIO controls packet count set 17 0-R11, set the number of packets to 1, and control the packet reception 170-R. Move to 6. TRECEIVE is the step after packet reception 170-R6. The final bucket check 170-R16 returns the loop to the calling bagogram. Continue until.

Packet receiver 170-R12 (Figure 35) is shown in more detail in Figure 36. Ru. At the time of entry of packet reception 170-R6, data link reception 17 0-RPI is called data link function DRECEIVE. Function DRECE The IVE is called by data link send 170-3P2 (Figure 34). The packet sent from the DSEND is received. The received packet contains As mentioned above, it includes the header, data area and CRC-16 checksum. It will be done. Error check 170-RP2 (Figure 3'6) 0-Determine whether the RPI has generated an error. Data link reception 170-R If an error code is generated by the PI, processing is performed by the data link clip. The process moves to step A 170-RP3, and the data link function DFLUSH is called.

Then, the error returned by the selected error check 170-RP4 is disconnected. Determine whether it is an error, timeout error, or bad node ID error. error is one of these selected errors, processing returns an error code. Move on to code 170-RP7 and also error check 170-R7 (Figure 35). Doru. If any other errors occur, processing will proceed by sending a reply as described in more detail below. Move to communication 170-RP5. Reply transmission 170-RP5 generates an error packet , sends this bucket 1- to data link transmission 170-32 (FIG. 34). return When the communication transmitter 170-RP5 (Fig. 36) completes transmitting the error packet, the processing moves to return error code 170RP6, then error check 170- Return to R7 (Figure 35).

If error check 170-RP2 (Fig. 37) does not detect an error, , the process moves to CRC error check 170-PR8. CRC error check , generates a CRC-16 checksum of the received data bucket and checks this Ch ecksum is compared with the received CRC-16 checksum. CRC-1 If 6checksum are the same, CRC error check 170-RP8 is processed. is transferred to the packet error checker 170-RP2. Packet error check is Compare the number of packets with the number of packets received and if they are not the same, take action. Transfer the control to Datalink Clear 170-RP3, and refer to the details below for Datalink. Calls the function DFLUSH, described in . Data link clear 170-RP3 passes the processing to Reply Send, which is explained in more detail below, and finally returns the message to Bad. Error 170-Return to RP5. If the packet numbers are the same. , packet error check The check 170-PH10 transfers processing to the data return 170-RPI6 and The length, header, and CRC-16 checksum of the received data are sent to the packet reception. return.

The above explanation indicates that CRC error check 170-RP8 did not detect any errors. I'm assuming. However, the CRC error check 170-RP8 shows an error. When detected, the processing is data clear 170-RP3, reply transmission 170-RP5 and Return CRC error code 170-Go to RPII. Other processing 170-R P3.170-RP5 and 170-RPll are 170-RP3.17 described above. Equivalent to the processing of 0-R, R5 and 170-RPI5, their explanations are here. Use it as part of the explanation.

The reply transmission 170-RP5 (Figure 36) is shown in more detail in Figure 37. . At the time of reply transmission 170-RPI entry, reply packet creation 170-3RI is the transport layer header, the given return code and the CRC-16 checksum Generates a packet with . After generating the packet, counter initialization 170-3 R2 initializes the error counter to zero. Data link transmission 170-3R3 calls the data link layer function DSEND to create a packet 170-3R forwards the packet created by 1 to the remote node that sent the received packet. send Error check 170-3R5 detects an error in data link transmission. Determine whether it has occurred. If no error occurs, process is passed to return 170-3R6, so control is transferred to packet receiver 170-R6. Passed to the next operation.

However, if data link transmission 170-3R3 generates an error, the error Check 170-3R5 moves to processing WP counter check 170-3R7. workman If the error counter is less than 5, counter check 170-3R7 takes control. Pass counter increment 170-3R4, increment the error counter, and change the processing to data redirection. link transmission 170-Return to SR3. On the other hand, if the error counter is 5 or more, the counter Data check 17o-3R7 transfers control to data link error 170-3R8, Return to bucket receiver 170-R6 with data link error code.

Therefore, the transport layer functions TSEND and TRECE IVE of the data link layer Call functions DFLUsH, DSEND and DRECEIVE. Furthermore, this In addition to these functions, the 7'-tan') link N180 has NNECT.

Contains DDISCONNECT, DSTATUS and DRESET. mentioned above As mentioned above, each of these functions is responsive to a corresponding function of the transport layer. data link The possible return codes from the check function are shown in Table 6 below.

Data link layer 180 sends a single packet across a network (i.e., a physical layer), along with supporting services. At the data link layer, There is no accuracy check. If the transfer layer 170 requests a change in speed, the data The tarlink layer 180 changes the speed if it is possible to do so. Data link layer 1 80 is a low level. This means that the layer is hardware dependent. Ru. The embodiments shown in Microfiche Appendix A are part of the disclosure herein. It is incorporated as a section. Data link layer 180 includes IBM personal computers and It is equivalent to an 18M compatible personal computer. However, the data link Through proper use of layers, the forwarding layer 170 can function independently of hardware. becomes possible.

Data recovery function DRESET (DRESET)4. t, data link try used for initializing and re-initializing the server. Transfer layer function TRESET or DRESE All other functions that call T perform a normal reset or change the transfer rate. Pass a parameter indicating this to DRESET.

One embodiment of the function DRESETI 8O-R5T is shown in FIG.

Reset check 180-R3TI is a parameter given to function DRESET 180-R3 and initialize the process if a normal reset is indicated. Pass it to T2. Initialization 180R3T2 is the change used in the data link function. Cents the number to the selected value. In particular, data link layer functions are When using timeout variables and other timeout variables to wait for an action to occur. monitor the time. These variables are ``timeout variables'' and ``intercharacter It is distinguished as a "timeout variable".

Both of these counts are set to zero when the flag is used to indicate connection status. will be cut. A zero value for either timeout variable indicates that the variable is used to monitor elapsed time. It means that it is not used.

In this embodiment, the connection state flag is used to indicate a non-connection state or a successful connection state. set to zero to indicate that a connection or listen is being attempted. Set to 1 and set to -2 to indicate that a speed adjustment is being attempted. . Once the flags and variables are set, initialization 180-R3T2 sets the transfer delay variables. For example, set a predetermined value of 1, move the processing to return 180R3T3, and call Returns successful completion to the original function.

Reset check 180-R3TI checks the parameters given to function DRESET. If the data indicates a speed change, the process is changed to parameter set 180-R3T4. Move to. Parameter set 180-R3T4 first sets the timeout variable to 30. seconds, then call Speed Sera) 180-3S (not shown) to transfer Set the speed to 300 po. Speed set 180-3S handles UART speed. function to the index given.

The speed cent 180-3S first searches the UART baud rate divisor table. , loads the divisor corresponding to the given index. The state of the UART is The send operation is monitored until completion, after which the UART baud rate divisor is accessed. is allowed. The loaded divisor is sent to the UART. At this time, the lower The data is transferred in order from byte to higher byte. Next, the divisor of the UART baud rate access is prohibited and processing returns to the calling function. In this case, the parameter data set 180R3T4. In the following, Speed Serra l-180-3S is It means this series of operations.

When the speed is set to 300 po by 180-3S, the baud rate The index is set to indicate speed. Next. Parameter set 1 80-R3T4 moves to processing and connection speed set 180-C3. Connection speed set 1 80-C3 sets the transfer rate with the remote node. Connection speed Cera) 180-C The operation of 3 is detailed below with respect to FIGS. 39A and 39B. connection speed cera ) When the operation of 180-C3 ends normally, an error flag is returned and an error check is performed. The block 180-R3T5 transfers processing to the delay count set 180-DC. delay Count set 180-DC first loads the baud rate index. No. One variable, loop count, is used to indicate relative machine speed.

In this example, the loop count is determined by the function DCONSECT. Defined as the number of times the register pair DX:AX can be incremented in the time interval between clocks. be justified. One clock is 55 milliseconds in this embodiment. Baud rate is 1 If greater than 9200, the delay count is multiplied by 8 and divided by the baud rate divisor. The number of cents is set to a value equal to the loop count. On the other hand, the baud rate is 1920 If less than zero, the delay count is set to zero. Delay count setting After the call, the delay count set 180-DC returns processing to the calling function.

In the following, the delay count set 180-DC represents this set of operations. Therefore , when the delay count set 180-DC is completed, DRESETl 80-R3 T returns processing to the calling function via 180-R3T3.

Details of connection speed sensor l-180-C5 are shown in Figures 39A and 39B. Ru. In general, in diagrams, circled characters mean connections within the same drawing; Letters within the shape indicate connections with other drawings, and also indicate connections with similar Rahel in the drawings. The elements marked are equivalent, and multiple references to the same element in a drawing are It is indicated by Rahel with capital letters.

The first operation of connection speed sensor t-180-C5 is connection state sensor) 180-C3I and the connection state flag is set to indicate that speed throttling is occurring. It will be done. Connect character transfer 180-C32 then transfers the key to the specified remote node. Send a character. In this example, the connecting character is "ENQ" . As used here, the connection character indicates that the functionality is through a virtual network. means that it is called to send a character. This feature transfers given a character and a delay factor. The character transfer function is Waits for the specified number of clocks and transfers the UART until the transfer holding buffer is empty. Test the condition of.

The given character is sent and processing returns to the calling function. below "Character transfer" means this series of operations.

After connected character transfer 180-CD, timeout check 180-C33 is , waits two clocks and checks the timeout variable mentioned above. Timea If so, processing moves to timeout return 180-C34. Sets the return value on a timeout error and returns the caller before returning to the calling function. Cent the leaf flag. If it has not timed out, check the timeout check. The check 180-C33 transfers processing to the character state reception 180-RC3.

Character status reception 180-C5 is described in more detail below with respect to FIG. Ru. Character status reception 180-RC is (i) remote in response to connected character transmission. receive a character sent by a node or (i) an error code; occurs, or (ii) indicates that the character is not in the waiting state , transfers processing to character check 180-C35. When the character is on standby If there is no character or character reception is not confirmed, check character check 1. 80-C35 returns processing to character transfer 180-C52.

Here, the reception confirmation or character reception confirmation is a character such as "A CK". It is indicated by. Once the reception of the character is confirmed, processing begins with timeout set 1. 80-C36 and set the timeout variable to 5 seconds. 'Processing is as follows Then, the process moves to character reception 180-RC. Character reception 180-RC is as follows This will be explained in more detail with reference to FIG. Character reception 180-RC function or error to the calling function. Therefore, error check 180- When C37 receives an error, it transfers control to error return 180-C39. vinegar. Error return 180-C39 sets the carry flag and codes processing. return to the original processing.

If no error is returned, processing continues at the start check of the baud rate sequence. 180-C38. Character reception is the star of the baud rate check If no error is indicated, processing transitions to Transmit/Receive Error Return 180C310. , set the return value to send/receive error, and set the carry flag before returning. Set.

Once the start of the baud rate sequence is received, processing begins with character reception 18. 0-RC-A and the low byte of the remote node's speed and the remote node's speed. Receive the upper byte of . If an error occurs in receiving these, please As mentioned above, the error check 170-CS7-A is the error check 180-C. Works before 37.

If no error is detected in error check 170-C37-A, the necessary Velocity request check 180-C3II compares the received rate with the requested rate. Here, the key requested speed means the speed given in the call to connection speed set 180-C3 do. If the receiving speed is lower than the requested speed, the process changes speed 180-C31. 2, and set the requested speed to the receiving speed.

If the speeds are the same, or if the speed change 180-C312 is complete, processing: Proceed to effective speed check 180-C313 to determine whether the speed is valid. do. If the speed is greater than or equal to the maximum allowable speed or less than 300 points, the processing Move to speed error return 180-C322 and change the return value to invalid line speed. and returns after setting the carry flag.

If the speed is legal, the character receive 180-Rt,-B is sent from the remote node. Receive the next character to be sent.

Error check 180-C7-B shows that no error has occurred as described above. If so, control is passed to baud rate string termination check 180-C314.

If the received character does not indicate the end of the baud rate sequence, processing Transmit/Receive Error Return 180 - Passed to C3IO. however. Character If the baud rate indicates the end of the baud rate sequence, the process proceeds to baud rate transmission 180 -C315 (Figure 39B).

Baud rate transmission 180-C3-15 transmits the selected speed through the virtual network. This is a series of operations for sending a message.

In this embodiment, the baud rate transmitter 180-C 315 waits eight clocks; Next, a character indicating the start of changing the baud rate is transmitted. After sending the character , character reception (FIG. 41) is called. For character reception, character or return an error. If an error is returned, the carry flag is sent and The turn value is set to time error and processing is returned to the calling function. Ella If the character is not returned, the character reception is confirmed. is checked to determine whether or not. Character reception not confirmed If the carry flag is set and the return value is detected as a transmit/receive error. is called and processing is returned to the calling function.

If the reception has been confirmed, the start of the baud rate sequence will be sent two clocks later. Transmitted following the end of the speed low byte, high byte and character transmission.

This completes the baud rate transmission operation, and the process continues with character reception 180- Move to RC-C.

If an error is returned, the error check 180-C3-16 will Pass to error return. Received character received 1III recognition 1 If it is not a character, reception confirmation check 180-C317 sends the processing. Transmit/Receive Error Return 180 - Pass to C3IO. Once the acknowledgment is received, the Degree set 180-3S is called and performs the operations described above.

Subsequently, character reception 180-RC-D is called and error check 180 -C316-A and acknowledgment 180-C3-B operate as described above. Character Once the reception of the session end character is confirmed, the end of session character is 80-C19 and the character received 180-RC-D is sent by the remote node. used to receive characters from.

If an error is detected, error check 180-C3-16-B times the process. Out error return 180-pass to C34. The received character is set If it is an end-of-session character, processing ends with session end check 180-C. S20, then passed to return 180-C321, otherwise send /Receive error 180 - Passed to C3IO.

The character status receiving function 180-RC3 is shown in FIG.

This function accepts a single character in register AL, or If the data is not ready, set the Z flag. First, the character The data reception function 180-RC3 performs a start test 180-RC3I. vinegar The tart test 180RCS1 checks whether the connection is made and the framing error. Determine whether or not a problem has occurred. Framing errors occur when two nodes This occurs only when transmitting at different baud rates. connection error or If a framing error occurs, start test 180-RC3 I rate resynchronization 180-3R, the process of which is described in more detail below with respect to FIG. Give it to E. Speed Resync 180 - SRE adjusts the connection speed so that both nodes are the same Communicate at high speed.

If speed resynchronization 180-3RE returns an error, error check 180- RC35 returns the processing to return 180-RC34, clears the Z flag, and returns the command. The function returns to its original function. On the other hand, if speed resynchronization did not return an error; error check 180-RC35, process error code cent 180-RC6 and set the error return value to speed change error. Processing then returns 180-RC4, clears the Z flag, and returns to the calling function.

Character reception function 180-RC (Figure 41) stores one character in register AL. Receive Kuta. On entry, timeout check 180-RCII Determine whether the timeout variable is 1 or not. If the variable is 1, processing is an error return. Go to step 180-RCII, set carry flag and return time error. . On the other hand, if the timeout variable is not zero, the UART status 180- RC2 reads the UART line status and checks the framing error. Ru.

Framing error 180-RC3 then determines whether there is a framing error. , and also determines whether the connection status flag is zero. Friends If no errors occur and the connection status flag is zero, the process then moves to character preparation check 180-RC4. No ready characters If so, processing returns to timeout check 180-PCI. On the other hand, character If the character is *a, character reading 180-RC6 is character is acquired, and processing is transferred to return 180-RC7.

If there is a framing error or the connection status flag is not zero If framing error 180-RC3 saves processing timeout 18 0-Move to RC3 and save the current timeout value. Then the process is as follows: Turning to speed resynchronization 180-3RE, which is described in more detail with respect to FIG. 42 below. speed Once resynchronization 180-3R,E is completed, timeout restore 180-RC8 restores this to the saved timeout value and speeds up processing with error 1 Transfer to 80-RC9. If an error occurs in speed resynchronization 180-3RE For speed change error 180-RC9 process error return 180-RCII Move to. On the other hand, if no error has occurred, processing is performed using the speed change 180-RCI Return to O and set the result of the function to indicate the change in speed.

The speed resynchronization function 180-3RE has the character reception function 180-RC (Fig. ) and character status reception 180-RC3 (Figure 40). This is shown in Figure 42. This feature allows timeout reset. The UART 180-3 REI first zeros out the timeout value, then the UART Boat clear 280-3RE2 reads the UART data port and clears the port. clear. Minimum speed check 180-3RE3 then speed is 300 baud Determine whether it is set. If the speed is 300 baud, the error return 180-3RE4 sets the carry flag and returns the speed error value to the caller. Return to function.

If the speed is not minimum, the minimum speed check 180-3RE3 is performed as described above. The process moves to speed set 180-3S.

In particular, the speed sensor) 180-3S sets the node speed to 300 baud and processes Move to baud rate save 180-3RE5 and change the baud rate index. baud rate. The Rinse Speed Set 180-LS is then remotely connected. Called to resynchronize speed with the board. Listening speed sevent 180-LS behavior The process will be explained with reference to FIG. Complete Su Noson Speed Cent 180-LS Then, delay count set 180-DC is called. This delay count set are as described above and are hereby incorporated by reference as part of this disclosure.

Upon completion of the delay count cents 180-DC, control is returned to the calling function. Ru.

The rinse speed cent 180-LS is shown in FIG. listen speed sensor The call to the baud rate is given as a parameter. connection Status set 180-LSI -2 after saving connection status flag cents to indicate that a speed change was attempted.

Then, the character reception 180-RC returns a character reception or an error. used for If an error is returned, error check 180-LS2 Move the processing to turn 180-LS3 ("error") and list the connection status flag. Tore and set the carry flag before returning.

If no error is returned, listen termination check 180-LS9 is set. Determine whether the end of the listened rate being If so, the processing is transferred to litter 7180-LS4. Litter 7180-LS4 is sends a reception IIg acknowledgment of the terminal character of the listened rate Restore the task flag to the value at the time of entry. Return to calling function. Otherwise When the connected character check is 180-LS5, the connected character is received. Determine whether or not. If this character is not received, processing Return to the vector reception 180-RC.

However, connected character check 180-LS5 receives connected character If so, processing is transferred to timeout set 180-LS6 and the timeout Set the out variable to 5 seconds). The process then transfers the baud rate 180-LS7 Move to. In this example, baud rate transfer 180-LS7 includes a number of operations. I'm here. In particular, the character acknowledgment is transferred and two clocks later the baud rate sequence is The starting character of is transferred. Lower hide, upper bite followed by baud rate string The terminating character of Transition. Baud Rate Save 180-LS8 saves baud from the table based on speed. Load rate index, then baud rate loaded index cent to Speed Cent 180-3S is the speed at the local machine, as previously described. Set it as instructed. An acknowledgment of the character is then sent at the new rate, Processing is transferred to character reception 180-RC.

The data link function DSEND is used to connect the remote Used to send packets to nodes. DSEND is for sending a packet. Returns zero to indicate success. If the packet could not be sent, negative An error is returned. The function to call DSEND is the length and a pointer to the message buffer to the function DSEND. DSEND 1 An example is shown in FIG.

Keyboard prohibition 180-3l turns off keyboard operations and blocks processing from packets. 180-3 PS.

Packet Size Send 180-3PS specifies the size of the packet to be sent to the remote node. Send. The operation of packet size transmission is shown in more detail in FIG.

Packet size sending 180-3PS error If it returns, the process moves to error check 180-33, and the keyboard Permission to use 180-310 lists keyboard operations with at least one digit pressure The reduced code and control are returned to the calling function from return 180-3l.

If no error is detected by error check 180-33, send Hyde increment 180-34 increases packet size since last reset. Added to the variable used to count the number of hides sent. Then the packet 180-S5 indicates whether there are additional hides available for each hide in the list. Then, the byte load 180-S12 sends the character 180-3. 13 loads this byte to send it over the virtual network do.

When a packet is sent, available bytes 180-35 are used to control the packet Is acquired 180-transfer to GPS. Packet Size Acquisition 180 - GPS receives Obtain the size of the bucket from the remote node. When an error occurs, processing , move to keyboard use permission 180-3IO. However, no error is returned. If not, packet size check 180-37 checks if the packet size is Determine whether the packet size matches the specified packet size. If the packet size is different error set 180-39 sets the return value to send/receive error. Then, the process moves to keyboard usage permission 180-310. On the other hand, the packet size is If they are the same, the normal completion sensor) 180-38 returns the return value and It is set to normal end before passing the code usage permission 180-3 to the IO.

Packet size send 180-5PS function called by function DSEND is shown in FIG. Initialization 180-3PSL is initially in the receiving operation. The flag indicated by is set to 1, and then the transfer delay is saved. The transfer delay is is set to the delay count described above. The timeout variable is set to 15 seconds. will be played. Timeout 180-3PS2 check the timeout variable , determine whether a timeout has occurred. A timeout has occurred In this case, error set 180-3PS3 sets the return value to send/receive error. 180-3PS4 sets the carry flag. then , the restore 180-3PS5 restores the transfer delay value to the saved value and performs the reception operation. After resetting the medium flag, control is passed to return 180-5PS6.

When timeout check 180-3PS2 does not detect a timeout , the transfer start 180-3 PS7 starts transferring characters to the remote node. . Character status reception 180-RC3 indicates the character or error that has occurred. , or indicates that there are no prepared characters as described above. Show. Therefore, Preparation Character Check 180-3PS9 indicates that the character is If it is determined that the be transferred. If the character is prepared and an error occurs, an error check is performed. Check 180-3PSIO returns processing to flag set 180-3PS2. Ella If no response is returned, the reception confirmation check 180-3FSII is The character returned by data status reception 180-RC3 is the reception confirmation character. Determine whether or not it was a kuta. If the character is not an acknowledgment character , processing is transferred to error set 180-3PS3.

On the other hand, when the acknowledgment is received, the packet size transfer 180-3PS12 The size of the packet yh to be sent to the remote node is transmitted. Especially packet size transfer 180-3PS12 indicates the start of transfer, the lower byte of the packet size, and the packet size. The upper byte of the data and the character indicating the end of transfer are transferred.

Character reception 180-RC receives character reception as described above with respect to FIG. receive. If the character reception 180-RC returns an error, an error check is sent. Check 1so-sp 310-A transfers processing to flag set 180-3PS4.

If the character is an acknowledgment, then the acknowledgment test 180-3PSII-A passes the processing to the restore 180-3PS5. Otherwise, reception confirmation test 180 -3PS 11-A moves the processing to error center 180-3PS3.

Get Packet Size Called by Function DSEND 180 - GPS Function is shown in FIG. Initialization 180 - CPSL sets the receive operational status flag. 1 and save the transfer delay and timeout values. Then, the transfer delay is , sent to the lag count as described above. Next, the process begins with character reception 1 Transferred to 80-RC-A.

If character reception generates an error, the error check 180-GPS2 Speed change error 180-C, returned to PS3, speed change error 180-GPS3, If the error is set to speed change error, process the character reception 1 Return to 80-RC-A. On the other hand, speed change error 180-GPS3 causes processing to transfer to GSP4, set the carry flag to cents, and restart the process. Move to I.A180-GPS5. The restore will restore the saved transfer delays and time Return the output value to an appropriate variable, set the receiving operation flag, and return the process to 180-GP. Before moving to S6, reset to zero.

If character reception 180-Rt, -A does not generate an error, error check is performed. The check 180-GPS2 moves the process to the transfer activation check 180-GPS7. Receiving If the character received does not indicate the start of a transfer, processing continues with the character received. Moving on to 180-RC-A. If the received character indicates the start of a transfer, then , Transfer Confirmation 180-GPS''8 sends a confirmation character to the remote node for processing is passed to character reception 180-RC-B.

If an error occurs here, the error check 180-C, PS9 Move to speed change error 180-GPS3 which operates as described below. An error has occurred If the transfer start character is not received, this process Transferred to character reception 180-RC-B by transmission start 180-GPSIO . Otherwise, transfer start 180-GPS], O process character reception 180- RC-C and receives the lower byte of the packet length. Similarly, the character The transmitter 180-RC-D receives the upper byte of the packet length and also sends the transfer end key. Receive a character. If an error occurs in receiving these, the processing Speed change error 180 - Move to GPS2.

Finally, the character receiver 180-RC-D does not receive the end-of-transfer character. If the transfer end check 180-GPSII terminates the process with error set 18 0-Move to GPS3 and cent send/receive error. Upon receiving the completion of transfer, Transfer confirmation sends an i-pressure signal to the remote node and clears the flag 180-CPSL2. Clear the carry flag before returning the process to the restore 180-GPS5.

As mentioned earlier, many functions of the data link are hardware dependent. subordinate Therefore, the embodiment of the data link layer 180 described here is compatible with a computer equivalent to an IBM-PC. shows the set of functions used in computers. However, based on the above explanation A person skilled in the art can define the application layer and transport layer functions as a service for these functions. It can be used with any data link layer that supports

In this example, many of the functions of the data link layer are performed on the PC bus every 55 milliseconds. hardware line I RQO is activated and interrupt 08h is executed. I am using it. The operation performed by interrupt 08h is depending on the service routine specified by the interrupt 08h vector. child By changing the address of the vector of It can perform different functions at different intervals. As mentioned above, the time interval of 55 milliseconds is , meaning clock time.

The data link layer of this invention uses two different interrupt services for interrupt 08h. are doing. The first interrupt service 180-INT is shown in FIG. The second interrupt service 180-RCV is shown in FIG. Second interrupt service service 180-RCV is used when the data link performs reception, and The link is used at all other times when the virtual network is operational. It will be done.

Both of these interrupt services are speed-critical and computer code is Optimized to run fast and reliably. In particular, interrupt service routine 18 For 0-RVC, all interrupts except interrupt 08h are disabled, and the following registers on entry to the interrupt service routine, as detailed below. A state is assumed.

Interrupt service 180-INT (Figure 48) first registers the save register upon entry. Access the register 180-INTI to save the registers needed for the interrupt.

The delay count increment 180-INT2 then increments the delay count. Action 1 8'0-INT3 to 180-10 check the two timeouts as described above. Bell. If the variable indicates a timeout, the timeout flag is set. be done. Otherwise, the timeout variable is decremented.

Next, the operation flag check 180--the operation flag received by INTII is Determine whether or not it is true. If the flag is not sent, interrupt 08 h connection 180-INTI3 is connected to the previous interrupt 08h and handles the register list. 180-INTI4. Register restore 180-INTI4 Resent the register value to the save register and process return 180-INTI5. returns the function to the calling function.

If the received operation flag is set, the process executes B103 service. Move to 180-INTI2. Other processes service interrupt 08h and these processes Processes may require long processing times to complete. thus increasing speed Therefore, the function is not combined with the previous interrupt 08h. Rather, BIOS service Run 180-INTI2 is the minimum service typically provided by BiO2. I will provide a.

The second interrupt 180-RC3I (Figure 49) is optimized for maximum speed. be done. Typically, an interrupt service routine stores the current registers for all other processing. Save before.

However, this routine only saves register AX. The routine is Furthermore, only this register AX is used.

However, since routines access memory, A segment register is required to specify the segment. Normal interrupt service The routine saves at least one segment register and then load the register into the interrupt service routine. However, interrupt support service routine 180 - All other interrupts must be disabled before enabling RCVI. , this interrupt has a segment set to the data link data segment. The data link function DRECE is only allowed in a small part of the IVE. all Save and load data segment registers are disabled because other interrupts are disabled. code is not needed since the register is already loaded with the appropriate value.

Therefore, the operation of interrupt service routine 180-RCVl is similar to save register 1. 80-R, CVl, delay count increment 180-RCV2 and internal character data Contains timeout value check 180-RCV3. Internal character timer If the out value is greater than 1, the timeout is incremented and processing ends interruptively. Move to 180-RCV5. Otherwise, internal character timeout check 1 80-RCV3 passes the processing directly to interrupt termination 180-RCV5. Interrupt end 180-RCV5 gives the interrupt end signal to the interrupt controller and registers Restore AX. Finally, return 180-RCV6 returns the processing to the calling machine. Return to Noh.

The data link function DRECE IVE forwards the message to wait for reception. used by layers. When a message is received, the function DRECE IVE is Returns the actual length of the message. DRECEIVE contains the node ID and the receive address. Corresponds to the buffer length and the number of clocks to wait for the receive buffer pointer and data. The specified timeout value is given. The function DRECE IVE is the message length or return an error code.

An example of the data link function DRECE IVE 180-R is shown in FIG. is shown. Initialization 180-R1 times out for given parameters Cent the variable and cent the timeout flag to zero. Then, the processing 180 - Move to GPS (Figure 46). receive packet size Then, the error check 180-R2 performs processing if no error has occurred. The process proceeds to reception initialization 180-R7. On the other hand, there is an error in acquiring the packet size. If an error occurs, processing is performed via error check 180-R2. Transfer to packet size 180-R3 and set the error code to packet size acquisition 180-CP. Set to the result of S.

Next, timeout cent 180-R4 is the timeout variable and internal character. Set both timeout variables to zero. Timeout error check 1 80-R5 returns the processing to error return 18 if no error is detected. Pass to 0-R6. Timeout error check 180-R5 has timeout error Error return 180-R6 is detected, processing The size transmission is shifted to 180-3PS. Packet size transmission 180-3PS is , which is the function described above (Figure 45). When reception initialization 180-R7 receives processing from error check 180-R2, , first set the receiving flag (7) and then set the timeout variable to zero. and set the internal timeout variable to 6). Reception initialization 180-R7 turn on the embedded service routine 180-RCVI (Figure 49). Then, The process moves to tJART status 180-R8 (Figure 47), and the UART Check status port. When the character is set up on the boat, Character Preparation Test 180-R9 Read Process Character 180-RIO Move to. After reading the character, check whether the buffer address is O:Oh or not. It will be judged. If the address is not 0:Oh, the character read is , is not stored by character storage 180-R12. Step 180-R1 1 can be removed from function 180-R, and is a function DCONNECT light. Line connection check operation and function Supports DLISTEN line listen check operation. used only for porting. As explained in more detail below, the re-behavior is , except step 180-R11, can be moved to the transfer layer if necessary.

After buffer address 180-R11, variable adjustment 180-13 receives character. Increment the count and set the internal character timeout variable to 3. Moaki Character test 180-R14 determines whether there are any more characters to receive. Make a judgment. If there are still characters to be transmitted, processing is performed with UA, R The process moves to T status 180-R8, and the above-mentioned processing is performed. character remains If not, update 180-R15 last resets the number of characters received. The error code is added to the number of characters received after the character is received. Set to . Interrupt restore 180-R24 is interrupt service routine 180 -Turn off RCV and turn on interrupt service routine 1801NT. The processing is , then moves to timeout set 180-R4.

The above explanation is based on the original temporary character preparation 180-R9 that transferred the character to the swordsman. It is established. However, if no character is detected, the process Move to internal character timeout 180-R16 and set the error code to internal timeout. The out variable is set, and the process moves to interrupt restore 180-R24.

However, the internal character timeout variable does not indicate a timeout If the framing error check 180-R18 detects the framing error When detected, the process moves to UART status 180-R8. otherwise, The processing is transferred to the interrupt restore 180-R24-A mentioned above, and the interrupt service is Turn on routine 180-INT and interrupt service routine 180-RCV. Turn off. Timeout save 180-R19 then saves the timeout value. save in the timeout variable and internal character timeout variable, then Speed resynchronization 180R20 works. After completing the speed resynchronization, the timeout Store 180-R21 saves timeouts and internal character timeout values. parameters. If an error occurs during speed resynchronization, the error Check 180-R22 transfers processing to timeout cell I-180-R4. one On the other hand, if no error is detected, restore 180-R23 restores all parameters. Restores the meter to the value at the time of entry and processes the function DRECE IVE180- Return to the beginning of R.

The data link function DCONNECT180-C (Fig. 50) is the transfer layer function TC Called by ONNECT. The function DCONNECT connects a specific remote node. attempt to connect to the host.

The calling function waits for the node ID, speed parameters, connection ports and data. Give the number of clocks. As mentioned above, in this example only a single node Since the node ID exists, the node ID is duplicated. Speed given to function DCONNECT The parameter is the desired baud rate divided by 100 if non-zero. Ru.

At the time of entry to the function DCONNECT180-C, the connection status The client 180-CI first sets the connection status flag to -1 to confirm that the connection is attempted. Show that you are being watched. Next, steps 180-MII to 180-MII in FIG. 18 (>Ml8) The main initialization 180-Ml is shown in more detail in -Initial membrane state make a change.

In particular, interrupt save 18C1-Mll saves the first 'current interrupt 08h vector Get and save. Next, the flag clear 180-Ml2 is set to the receive operation flag. clears interrupt set 180-Ml3 to interrupt service call to interrupt 08h. Set 180-INT. Variable initialization 180-Ml4 is DCONNE Initialize other variables of CT and variables used in other functions to required values. continue , cent rate call 18O-SS sets the transfer rate as described above. Forbidden Register clear 180-Ml6 clears all steps from the UART. UART interrupt while reading the status and data to carry the register. prohibited. Next, the velocity calculation 180-Ml7 calculates the time interval between the two clicks. In register pair DX: count the number of times AX can be incremented, The value is stored in a variable called loop count in this embodiment. return 180-Ml8 passes processing to the connect/listen initialization 180-CLI of FIG. vinegar. The operation of the Connection/Resource Initialization 180-CLI is shown in more detail in FIG. It is. In particular, the timeout check 180-CLI sets the timeout variable to Set the parameters given to function D CON N E CT 180-C. Ru. Speed check 180-CLI2 checks if the speed parameter is given to DCONNECT. If so, processing is passed to parameter return 180-CLI5. parameter Data return 180 - CLI5 returns the index of the given parameter . On the other hand, if the speed parameter is within the given value, i.e. zero is the function DCONN If given to ECT, parameter set 180-CLI3 contains the speed parameter. 180-CLI4 is set to the maximum baud rate and the return 180-CLI4 is Return the index to baud rate index 7) 180-C32 (Figure 50) .

Baud rate index set 180-C2 connects the baud rate index/ Listen Initialization 180 - Set to CLII return value. Processing is then carried out at reasonable speed. The process moves to speed check 180-C3 to determine whether the requested speed is valid.

If the requested speed is invalid, the process starts with the legal speed check 180-03 and returns to error detection. Go to bit 180-C4 and set the return value to Illegal Line Speed Error. workman The controller 180-C4 transfers processing to the connected state 1-180-C1-A. , sets the connection status flag to zero, and moves the process to restore 180-C14. Ru. The restore 180-C15 denies access to the UART divisor and Disable interrupts for T, then set interrupt 08h to the stored previous interrupt 08h. Restore to vector. Processing then moves to return 180-C15.

If the requested connection speed is legal, the legal speed check 180-C3 performs processing first. Move to 180-3S and set the required speed. then , speed change 180-C5 indicates that the specified speed is Determine whether or not it is true.

If a specific speed is set, processing is performed in character transfer 180-C1l. and send a connect/listen termination character to the remote node. 1 clock After waiting, timeout check 180-C12 determines that a timeout has occurred. do. If a timeout has occurred, error set 180-C13 is reset. The process returns the connection status 1 as described above by setting the turn value to the timeout error. 80-CI-A.

If no timeout has occurred, timeout check 180-C12 moves to the character status reception 180-RC3 process described above. Character If the character is prepared or character status reception 180-RC5 Therefore, if the received character is not an acknowledgment of the end of the transfer, processing Return to vector transfer 180-C1l. However, character test 180-C If 16 detects a confirmation of connection termination, processing proceeds to the delay counter set described earlier. Move to cut 180-DC,-A. Connected Cent 180-CIB displays the Connected Status Set the lag to zero. , the process moves to return 180-C15.

If speed change 180-C5 detects a zero value, processing changes to character set 180 Moving to C6, the connecting character is sent to the first character. Then set connection speed 180-C3 performs the functions described with respect to Figures 39a and 39b. cormorant. Connecting character CEF) 180-C6-A sets the second connecting character Then, the process moves to the connection status set 180-C1-A. connection status set Port 180-CI-A sets the connection status flag to zero. Then the delay Count cents 180-DC functions as described above and baud rate check 18 0-C7, if the baud rate is below 19200, delay the processing and set the count. Transferred to 1809-DC-A. Otherwise, baud rate check 180-C7 is not processed. The process moves to the transmit line check 180CLS, detailed with respect to FIG.

Once transmission line check 180-CLS is complete, error check 180-C8 determines whether an error was returned. If an error is returned, the error -Check 180-C8 is the connection status set 180-CI- whose processing was described above. Move to A. On the other hand, if no error is returned, error check 180-C B transfers processing to block reception 180-C9 and performs data link function DRECE. Call IVE and receive data from the remote node with a timeout period of 15 seconds. receive a block of If the block is received without error, the error check Check 180-CIO transfers processing to connection status 180-CI-A.

Otherwise, 180-CIO delays processing to count set 180-DC,-A. Transition.

At connection 180-C, operations 180-CLS, 180-C8, 180-C9 and and 180-CIO are used to verify computer line connections. other In the embodiment, these operations are moved to the transport layer to improve efficiency.

The functionality of the transmit line check 180-CLS is shown in more detail in Figure 53. Ru. Upon entry, the Block Send 180-CLSI first processes the data. ’ and then calls the data link function DSEND to set the virtual 2/2 send blocks of data across a network. DSEND does not generate an error If the function works, error check 180-CLS2 returns the process to 180. -Transfer to CLS3.

However, if the block transmission 180-CLSI is in error, , processing is slowed down by error check 180-CLS2 to 180-CLS4 be transferred. Speed reduction 180-CLS4 sets data link function DRESET to speed Called with a parameter indicating the change.

Therefore, the function DRESET returns an error as described above.

If DRESET returns an error, error checking 180-CLS5 terminates the process. Move the error return 180-CLS6, set the carry flag, and caller Return to the function. Speed reduction 180C-If no error is returned from CLS4 The error check 180-CLS5 recodes the block send 180-CLSI. Start the line check at a reduced speed by rolling.

The data link function corresponding to function D CON N E CT is DLISTEN. Yes, waits for the arrival of DCONNECT and generates a signal indicating the start of the session. . DL I 5TEN further calculates the line speed to determine the optimal line speed. Prepare to make changes. DL I5TEN is connected to speed and timeout value. Give a second boat.

An embodiment of the function DL I5TEN is shown in FIG.

Operations 180-Ll through 180-L2 are similar to operations 180-L1 and 180-L2 described above with respect to FIG. C1 to 180-C2, the descriptions of which are incorporated herein as part of the disclosure. Ru. Legal speed 180-L3 causes processing to be performed as an error if the requested speed is illegal. Transfer to set 180-L4. Error set 180-L4 returns an invalid value. Input speed error and transfer processing to connection status set 180-C3 and set the connection status flag to zero. Next, restore 180-L 6 performs the same operation as the above-mentioned Noristore 180-C4 (FIG. 50). This operation is Hereby incorporated by reference as part of this disclosure.

If the requested speed is legal, the legal speed check 180-L3 continues the process described above. 180-3S.

After completing the speed control, the connection character set 180-LS is the first connection key. 7) 180 LS functions as described above. Execute. Once the listening speed set 180-LS is complete, the connection character St. 180-LIO is set by character set 180-LS. The connection character is different from the cent. Listen speed set 180-LS If an error occurs, the error check 180-Lll performs the processing described above. Transfer status set 180-L5.

When the listen speed set 180-LS completes its operation normally, the error check 18 0-Lll transfers processing to speed change 180-L12. given velocity parameter is not zero, the speed change 180-L12 causes the control to change to the delay force load described above. (Tasen) Transfer to 180-DC-A. On the other hand, if a speed change is specified, the speed Change 180-L12 changes the processing to the delay count of the functions described above (7) 180-D Give it to C. Baud rate check 180-L13 indicates that the current baud rate is 1920. If it is less than 0, it is returned to the calling program via 180-L14. Bo - If the rate is greater than 19200, check the baud rate 180-L 13 is timeout 7) Pass the processing to 180-L15 and set the timeout value to 15. Set to seconds. After setting the timeout value, the connection confirmation 180-L16 is Call the data link function DRECETVE to confirm the connection. DRECE If the IVE generates an error, error check 180-L17 will terminate the process. Move to Imout 180-L18. Timeout 180-L18 is the timeout Check for errors. If a timeout error is detected, the error return 180-L19 returns the process with a time error, otherwise the process returned to Mout St. 180-L15.

If connection confirmation 180-L16 does not result in an error, error check 18 0-L17 returns the above processing via line speed check 180-CLS . Processing then moves to delay count set 180-DC-A. Function D The line check performed by LI5TEN can also be performed at the transfer layer. It is Noh.

DSTATUS, DCONNECT, DDISCONNECT and DFLUSH The remaining functions of the data link are unambiguous. Function DSTATUS is supplied provides the necessary status information to the buffer. In particular, the node ID is set to 1 be done. Used to track the number of bytes sent and received since the last reset. variable is included in the structure and the speed index is used to find the current speed. is used and subsequently stored in a buffer. The buffer is then returned to the transport layer .

In this example, the function DFLUSH is a null operation, so the function is simply a code. It is to return to the original function.

Function DDS I C0NNECT prohibits access to UART divisor and tJAR In addition to disabling interrupts to T, interrupt 08h is set to the previous interrupt 08h vector. and returns normal completion.

The aforementioned data link error is returned to the transport layer as a negative value. Successful completion means is indicated by the return value (generally zero). Possible occurrence in this example Table 6 shows the return values.

Table 6 Return value: Result of data link function〇　--Normal end -1 = Connection disconnected -2-Unknown node ID -3- Incorrect line speed -4- Sending/receiving error -5- Timeout error -6- Cannot change speed -7 = Internal character timeout error conventional second and second computers Unlike systems that convey information between systems, user applications , the virtual network management of the present invention is used to transfer information between two computers. You can use the physical functions. The user application simply uses the information described above. It is only necessary to provide information and commands to the application layer. less than, Two user applications that use the virtual network management functionality of the present invention Explain. One user application is "Bokent Link (Poq etLtnk) A user of the MS-DOS operating system called J - It is an interface application. The blurring link is MS-D. Features disk file management for anyone with minimal knowledge of O3 In particular, Bokent Link allows users to copy files and directories, It enables operations such as deleting, renaming, printing, and displaying files. , allows you to view, create, and delete directories. Therefore, the user application tion's Borken link and MS-DO3's C0PY, DEL, REN, PRI NT, REDIR, CHDTR, MKDIR, DIR, PRINT, TYPE Almost every command corresponds to the other.

Furthermore, the Borkent link can be attached, detached, and connected. t.

Disconnect, 5erver and conf igure commands Therefore, it is possible to operate the virtual network management function. Borkent Link Plug-in The program provides a user interface and gives The commands to be executed are converted into appropriate command strings and commands of the virtual network management function of the present invention. Convert into information.

When using Bokent Link as a user application for local processing The virtual network architecture system has Must be loaded on your computer. For remote processing, local and Remote computers are directly connected, i.e., via a virtual network architecture. R3232 and the Borkent link is connected to both computers. must be running on the computer.

To execute a Borkent Link command, the user must e.g. use cursor control. Use the Control keys to highlight the file you want to work on, then press the Menu key (File function key (FIO) and operate the cursor control keys to select the appropriate item, For example, "highlight fileco.The user then presses the return key. and again with the cursor control keys select the appropriate option, e.g. "Copy" . By pressing the enter key here, the option is executed and the user The necessary information is requested.

To configure a machine to perform these operations, the user must first configure one of the Set the Borkent link as a server (this sets the machine as a remote machine) ). to the other machine as a local machine. The user uses BorkentLink to perform the connection functions and create a logical connection with the remote machine. command to do so. Once the virtual network management function connection is formed, all remote Disk drive processing is directed to a remote machine via a virtual network will be redone.

In addition, if a virtual network connection is already formed, the connection request is executed. connection option has no effect. Otherwise, Bo Kentrin A connection request from the network generates the above-mentioned VCONNECT of the virtual work management function. do. After a successful connection, the general communications between the local machine and the server machine The remote drive's available drives are displayed by exchanging communications. In particular, the blur Links can be used using common send and receive messages as described above. Get a list of drives. Next, the server machine's current drive is mapped to an available local drive.

If there is more than one remote drive, no other remote drives will be mapped.

To map other remote drives, the user must use PokenLink's Attach option must be executed. The remote drive being attached is treated as a separate drive.

After the serve request described above is executed, the server of the virtual workbook executes the service request described above. As mentioned above, if a connection request is not received within a predetermined time, terminate the process. . If a timeout occurs, the Vokent link will serve the virtual network's request. Rerun. When the server receives a disconnect request, the disconnect is performed and the virtual The network management function will be reset and the control will display the main poker link screen. Return to the Bokent Link user application to show.

As mentioned above, the Borkent Link user application is an MS-DO Manipulate files without having direct knowledge of 3 operating systems. make it possible to carry out work. The Bokent Link user application Therefore, the user can tag (or tag) multiple files in multiple directories. It is possible to perform the desired function by attaching a mark or mark). Here, the tag is a feature of Borkent Link that allows you to copy or delete files. used to display. To tag files, users can highlight Move the bar to the target file name and press the tab key, or from the tag menu Use tag options. If you want to tag a directory name, add a tag to the directory name. The directory is opened and all files in the directory are tagged.

To delete the tag, use the Boken link to delete the '7-ri of the file. It is a function. To clear a file's tags, the user can flip the highlight bar. Move to the file name and press Shift and Tab keys at the same time, or use the Tags menu. Use the tag clear option in the menu. When clearing directory name tags The directory is opened and all tags of the files in the directory are erased. be removed.

To copy a file, the user must tag the file to be copied. , run the copy option from the file menu. The Bokent link is When , the copy destination is asked. The user has a full drive and directory bus Or you can give just the path. If no drive is given, tag The files will be copied to the current drive. This current drive , it can be changed using the change volume command in the file menu.

To delete a file, the user tags the file to be deleted and Run the delete option from the file menu.

To change the file name, move the highlight bar to the file name you want to change. and execute the rename option from the file menu.

Borkent Link uses the DO5 print command and the remote printer to print the file. It has similar print commands, except that it can print files. Remote Print is configured with Vokent Link to use a remote printer This is necessary. Files to be printed are tagged and then print option is selected from the main menu. Both files have tags If not, the currently highlighted file will be printed. .

Another BorkentLink feature is the view feature. View is MS-DO3 tie It functions like a push command, but with greater capabilities. view the file To view a file, the user must hover the highlight bar over the file name and press the Press key+enter or run the view function from the file menu. . In View@Noh, the Vokent link is the appropriate application layer mentioned above. commands to access specific information for virtual network management functions. Once the information is reread, BorkentLink displays the text on the screen. Perform the necessary coding to do so.

When viewing directory information using portent links, users opens that directory.

Bokent link first opens the root directory and then opens this directory Keep the container in the oven. Portent links list directory information and and when the intermediate directories in the list are opened, the Borkentrin inserts directory information at that location. Users can scroll up and down the list. Roll, page up or down, move to the start and end of a list, or You can blow by moving before and after resalting.

If you are trying to delete directory information rather than the directory itself, This can be done by closing the bird. Closing a directory is This means that the directory information will only be cleared from the list, but not deleted from disk. Taste. Pocket Link is open and closed, and all items in the list are Preserve directory state. Directory levels are indented. Immediately In other words, the root or parent directory occupies the first line, subdirectories are and subdirectories of subdirectories are further indented.

View subdirectory information when parent directory is closed To open a directory, the user must move the highlight bar to the parent directory and perform Open. go The user then adjusts the highlight bar to the subdirectory. Run the program. Create a new subdirectory, like \A\B. Move the light bar to the A directory and select the directory option.

If a directory is not opened, it will be inserted. Opened by pressing a key. Here you can enter the Pocketlink directory. A selection is made to create the directory. User creates a new subdirectory salt You will be asked to enter. Similarly, delete the directory using Vokeratlink. If you want to remove the highlight bar, match the directory salt and then delete option is selected.

A link provides a series of selection menus that allow the user to select a file or directory. the second menu for the selected file or directory. select menu options to perform the desired action. User selects command Once done, Vokeratolink converts the user's selection into a sequence of actions, i.e. application to perform the action requested by the user as described. Convert to layer command.

The second user application connects the remote node using Pokeratlink. Requires server mode. This second user application is then loaded on other machines and contains the server's disk directory information used to obtain.

The embodiments of the present invention described above are applicable to IBM's personal computers and IBM equivalent computers. - Designed to run on personal computers. However, this The examples are only for detailed illustration of the invention and are intended to illustrate the invention in more detail. It is not intended to be limited to. Based on the above explanation, a person skilled in the art can Extends the virtual network management capabilities of Microsoft to other computer architectures and It is easy to implement in a data programming language.

The virtual network architecture of the present invention connects user applications and intervening between the computer's operating system and the computer's operating system. In one embodiment, As mentioned above, a virtual network is a computer's free memory ( ROM) and is loaded as part of the ROM-B IOS. other fruit In embodiments, the virtual network is loaded as a resident program. Either Even in this case, after the virtual network function is loaded, it becomes a resident program.

In the conventional resident program, the executable code and data are always on the scale. It is separated into a storage part and a temporary storage part, both of which are made up of random access memory (RAM). ) (that is, it was loaded into the main memory 17B (Fig. 1). The file contains the executable code and data of the application. temporary storage The executable application code is installed in minutes. In general, follow Temporary storage resident (TSR) program initialization data and memory of the next technology Interrupt handlers stored in the resident part and unexecuted handlers left in the resident part of RAM The MS-DO3 function frees the memory used by the temporary storage portion.

According to the principles of the invention, the executable code of the virtual network architecture is , are run directly from ROM rather than from RAM as in the prior art. From ROM Since the program being executed cannot be positioned differently, the computer system A new "loader" is added to configure the system into a program executed from ROM. ” is used. "Executable program" here means "executable image" has the same meaning as Furthermore, the loader of the present invention is suitable for ROM-based TSR applications. It is not limited to applications. Generally, the loader of the present invention can be used with any professional Programs can also be built into computers such as memory card ROM, ROM-BIO3, etc. The environment of the computer system must be configured so that it can be executed from ROM, including the The purpose of this is to determine the

A computer system 1220 including a loader 1100 according to an embodiment of the invention As shown in Figure 55A. Computer system 1220 (Figure 55A) is , the input module and output module similar to the conventional system 20 (FIG. 1). Contains. However, for clarity of explanation. Related to loader 1100 Only important components are shown in computer system 1220 (Figure 55A). There is. Generally, loader 1100 provides secondary storage to computer system 1220. The data is stored in the storage 1217A as an EXE file or a C0M file. Li The code-only memory ROM 1230 is an executable memory ROM 1230 associated with the loader 1100. It has a program 1250. RAM1217B is the operating system 1203. The CPU 210 has a ROM 1230. RAM1217B and and secondary storage device 1217A through passages 1240, 1241 and 1242. Both have been contacted. Further, the passage 1243 connects the secondary storage devices 1217A and 1 217B, and the CPU 1210 operates the loader 1100. Used to move directly to RAM 1217B using the programming system 1203. be done. The operation and architecture of computer system 1220 is similar to that of conventional systems. system 20 (FIG. 1) and therefore obvious to those skilled in the art.

The loader 1100 (FIG. 55A) of the present invention has an operating system 1203. is loaded into the RAM 1217B. For example, the loader 1100 uses RAM The operating system 12 is located at addresses XχX and YYY of 1217B. By 03. , specific ROM-based executable program 1 from the user 250 in response to instructions for execution of loader 1100, according to the conventional method described above. is stored.

When the loader 1100 is loaded into the RAM 1217B, the loader 1100 loads the RO Required to run ROM-based executable programs directly from the M1230 Perform environment settings. In particular, the initialization means 1100-1 (FIG. 55B) Initialization is performed to run the M-based executable program 1250. initial The conversion means 1100-1 is a ROM-based executable program for [250] A data area is formed in RAM 1217B.

Conditions are set for directly executing the ROM-based executable program 1250 When the loader 1100 is opened, the loader 1100 preferably connects itself to the RAM opening means 1100. -2 (FIG. 55B) to remove from RAM 1217B. loader 110 After 0 removes itself from RAM1217B, the ROM-based executable A data area for the program 1250 remains in memory. ROM-based execution Possible program 1250 (FIG. 55A) can then be downloaded from ROM 1230. It can be executed directly.

In accordance with the principles of the present invention, loader 1100 according to one embodiment provides a ROM-based implementation. A computer for executing program 1250 directly from ROM 1230. It has means for initializing the data system 1220 and means for freeing the RAM. Of course , in this example, the loader 1100 is configured to perform random access by itself. The loader of the present invention includes means for freeing the access memory. It is possible to operate properly without releasing the occupied RAM. however , these embodiments require less RAM because loaders that are no longer needed occupy RAM. This obviously reduces the usage efficiency of M1217B.

Executing an application from ROM is an executable program of the application. Because the RAM size can reduce the amount of RAM available to the user, It can improve the performance of your computer. Therefore, if the user Amount of RAM available for storing information such as files and other applications is larger than if the RAM was occupied by an executable program. Ru. Furthermore, the loader 1100 can load all RAM cards or all R in the computer. Operates a portable computer using executable programs stored in OM It shall be possible to do so.

The structure and operation of the loader corresponds to a ROM-based executable program 1250. It also varies depending on the steps in which the loader 1100 is used. In either case loader 1100 ensures proper execution of ROM-based applications. Make it fruit.

Details of the loader 1100 of the present invention are shown in FIG. 56A. In this example Then, the loader checks the existence of the data area initialization means 1110 and the application. It includes means 1120 and set amplifier means 1130.　First in data area The synchronization means 1110 is a data area for a ROM-based executable program. is set in RAM.

The data area initialization means 1110 in one embodiment includes (i) a ROM-based implementation; Work data area 1110-1 (Figure 56B) for programs that can be executed (ii) ROM-based executable program to RO Contains means for initializing variables in the data area necessary for direct execution from M. I'm reading.

The application existence confirmation means 1120 (Figure 56A) is a ROM-based that an executable program physically resides in ROM within the computer confirm. The application existence confirmation means 1120 is a ROM-based program. Upon detection of the send door knob means 1130, the loader 1100 continues operation by the send door knob means 1130. Continue. However, ROM-based programs are If the error message does not exist in the ROM, the error message means 11 50 to the user, and the loader 1100 causes the abort means 1140 to Stop the operation.

Send door knob means 1130 includes data and segment registers and ROM-based Sets all registers and/or vectors needed by the executable program. to The operation of the server upload means 1130 is based on a ROM-based executable program. This will be explained in more detail below.

Generally, the load-up means 1130 is configured by at least the loader 1100 itself. Free up the required RAM and control the computer system using ROM-based Pass it to an executable program. The loader 1100 releases the required portion of RAM. After that, the data area of the ROM-based executable program remains in RAM.

Execution of a ROM-based executable program It changes depending on the nature of. Generally, executable programs include (i) TSR applications; and (ii) non-TSR applications. Non-TS For R applications, control is transferred to a ROM-based executable program. The passed and ROM-based executable program initializes the data area of RAM. executed directly afterwards.

Once execution is complete, a ROM-based executable program typically runs on MS-DOS Return to operating system.

For TSR applications, the setup means 1130 is a ROM-based An executable program can be set as a TSR program. So The details are explained below. On the other hand, a ROM-based executable program ( i) Perform the necessary interrupt vector setup, (i) TSR program (i i) Remain resident while releasing unnecessary code/data. ROM-based executable A capable program is configured as a TSR program, so it is not a ROM-based implementation. Executable programs must be executed directly after initializing the RAM data area. There isn't. Therefore, in accordance with the principles of the present invention, loader 1100 executes directly from ROM. necessary to perform the operations required to run a ROM-based executable program. It has the necessary structure.

Loader 1100 working data area for ROM-based executable programs The structure of the creation means 1100-1 of (a) is based on (i) the aspect of an executable program, that is, , executed from ROM. Depending on whether it is an EXE executable file or a COM executable file. , and in some cases (ii) ROM-based executables that execute directly from ROM. It varies depending on the size of the data area required to run the program. in general As is obvious to those skilled in the art, source code such as assembly code is machine code compiled and linked to form an EXE file containing . On the other hand 1. The EXE file is MS-D.

You can convert it to a C0M file using the S operating system. come. , for C0M files, the operating system uses 64 By allocating kilohyde, the ROM-based executable program will be 64 kilohyde. When a work data area smaller than the loader 1100 of the present invention is required shall not require the creation of a work data area.

On the other hand, once the loader 1100 is created, the ROM-based executable program The data module for is the compiled source code for Loader 1100. will be linked.

In this case, it was generated for loader 1100. EXE files are executable The E Contains the file header. Therefore, in this embodiment, work data area creation Means 1100-1 executes a ROM-based executable program from a ROM. There is no need to specially allocate a work data area for the data.

Finally, the executable program is allocated by the operating system. A larger work data area than the C0M file, that is, the loader 1100 If you require a working data area larger than the 64 kilohide allocated for In this case, the work data area creation means 1100-1 creates the necessary additional RAM. assign.

These operations are outlined in FIGS. 57A to 57C. First embodiment is first executed by the operating system when loader 1100 executes. The allocated memory is from memory address XXχ to YYY (Figure 57A). be. This allocated memory is used by the loader 1100 to further allocate memory. It does not require any action. In the first embodiment, the operating system The system stores the memory of the loader 1100 at RAM addresses XXχ to YYY (57th Figure 8). However, the loader 1100 is unable to access memory at address XXX. Only memory addresses AAA to YYY are required by the loader 110. Released by 0.

Finally, in a third embodiment, the loader 1100 is an operating system Addresses XXX to YYY are assigned by R. As an OM-based executable program requires a larger data area, Therefore, the loader 1100 uses the operating system to access the address YYY. Allocate additional memory from to BBB.

The first embodiment (57th Alffl), for example, sets the required memory size to the file header. It supports the loader of EXE files included in the header. Second embodiment (Figure 57B ), the operating system allocates around 64 bytes for an executable program. requires a data area smaller than 64 kilobytes. 00M file Compatible with the reader. The third embodiment (Figure 57C) has six executable programs. Requires a data area larger than 4 kilobytes. C0M file loader It corresponds to

Data area initialization means 1110, existence identification means 1 20, and setup means The operation of 1130 will be described in more detail with respect to specific embodiments of the loader of the present invention. .

In one embodiment, the loader (FIG. 58A) is loaded with, e.g. Pending U.S. Patent Application No. 07/375.721 by John Bee, Fe. John P, Fairbanks et al.'s "Portable FortableLow Power Computer ) J., filed June 30, 1989. It is used to configure a library of ROM-based functions in the controller. In addition, the above The disclosures of the US patent applications are hereby incorporated by reference.

The source code for this example and each of the other examples described below is written in assembly language. Microsoft Macro Assembler (Microsoft Macr) o As5cvbler) version 5.1. Compa The generated code can be downloaded using the Microsoft Overlay Linker (Microsoft Overlay Linker). ft　verlay　Linker) version 3.65. It will be done. Both the macro assembler and overlay linker are based in Washington State. It is available from Microsoft Corporation.

In this example, the data segment information of the ROM-based library of functions is , the source code of the loader l100A is linked into an executable file. , ie, an EXE file. Data segment information is stored in the ROM base. must be determined by the executable program on the base.

For most well-behaved programs, at least the TSR program The vector of the last interrupt to which control was passed is kept in the data segment information, The TSR program is linked to the interrupt.

The nature of the library functionality is irrelevant to the features of the invention.

The loader 1100A is a machine that is executed directly from the ROM included in the loader 1100A. The initialization operations required for the ROM function and described in more detail below are Supports base library functionality.

The ROM-based library function loader l100A is an operating system The operating system is 1. EXE file header Read and loader 1100A has sufficient RAM for ROM-based library functions. Allocate random access memory RAM and data segments. Furthermore, those skilled in the art As is obvious in , the operating system is the program suffix and prefix (PSP) in memory and specify the start of loader l100A. Set. PSP ordering is done by the MS-DOS operating system. It is a standard function that can be used as a standard function. Microsoft Press, Washington State, pp. 108-111. It is shown. The content of this disclosure is incorporated as part of the disclosure of the specification.

[:l-da1lOOA's executable programs are run when the operating system is R. Provide sufficient information to allocate a working data area to the OM Base Live 99 function. It is executed by the work data area creation means 1110A-1. The only operation is segment setup lll0A-1-1. Registers DS and ES are set in the data segment. In this example, R A stack of OM-based library functions is contained within the data segment.

Data area initialization 1110A-2 (Fig. 58A) is performed in four steps in this embodiment. It includes two actions. The data segment initialization means 1110A-2-1 is R Zero the AM data area. This is done by clearing the data area. It is done. The PSP storage means 1110A-2-2 stores program segments. Store the prefix address for later use. Variable initialization means 11 10A-2-3 for continuously executing a library of ROM-based functions Initialize all necessary variables. In this example, this is (i) live Stores a pointer to the previous interrupt vector used to concatenate interrupts by functions in the library. and (ii) initialization of the menu system memory allocation scheme, (ii) machine data for other ROM-based applications that use the function library. Requires initialization of the word used to store the segment. Precise initialization of variables is not an essential requirement in the present invention. An important feature is that the loader l100A Necessary for continuous execution of executable programs from AM data area and ROM. This involves initializing all necessary variables.

As is well known to those skilled in the art, the operating system loads the PSP by Loader l containing command line parameters and loader environment segment address 100A is used to track unique predetermined data. Therefore, the data error The final operation of initializing alll0A-2 is the command line lll0A-2-4 This is the process.

First, in processing command line lll0A-2-4, the PSP looksup 1110A-2-4-1 (Figure 58B), all command line acquisition of an item. If the command line argument is “■”, the command line Drain argument check function 1110A-2-4-2 commands processing Line processor ll0A-2-4-10 version number report ll ROM-based executable program version passed to l0A-2-4-3 Report the number and exit via normal exit 1'l 40 A.

If the command line argument is R”, the command line argument The client check function 1110A-2-4-4 performs processing using the command line processor 1. Pass the reset of 102A-2-4-10 to 1102A-2-4-5 and set the data error. and exits normally via exit 1140A. Finally, the command line If the input is “U” instead of “1” or R, then the command line The statement check function 1110A-2-4-6 performs processing using a command line process. Pass it to the unloader 1110A-2-4-7 of the processor 1102A-2-4-10 and load it. 1110A and the information related to the ROM-based executable program. 1140A, and typically exits via exit 1140A.

PSP lookup 1110A-2-4-1 is command line argument If not detected, the process uses three check functions 111 QA-2-4-2, Existence confirmation means 112 via lll0A-2-4-4 and 1110A-2-4-6 Move to 0A (Figure 58C).

In this embodiment, the check memory means 1120-1 is a ROM-based functional library. Determine whether the library is already loaded. Detected in interrupt vector The code that the retrieved loader 1110A receives processing is a function of the library. It is determined whether or not the person is showing the ability. In the initial configuration of the library, the code is "VNA". Therefore. Check 112OA-1 is "VNA..." If a character string is detected, error message 1150A indicates that the library function has already been activated. Normal termination with an error message indicating that the The process ends through 140A.

However, if the library function is not available, the ROM check procedure Stage 112OA-3 stores a library of functions in ROM within the computer system. To verify that.

In particular, in this embodiment, the means 112OA-3 is a string code of the library, i.e. Check "VNA...".

ROM checking means if ROM-based library functionality is not available 1120-3 passes the processing to error message 1150A, and the library function Generates an error message indicating that it is not in OM. The process then ends 11 40A, and returns the computer system to the state immediately before loader lll0A starts execution. The necessary actions are taken to restore the original state. These operations are This is a normal operation that occurs at the end of a program, and is an obvious operation for those skilled in the art. Ru.

Once the termination of the ROM-based library function is confirmed, processing begins with the ROM check procedure. Moving from stage 112OA-3 to setup 1130A (Figure 58c), the main operation is performed. In the embodiment, the memory area of the RAM associated with the loader lll0A itself will be released. Therefore, all that remains in RAM is the data area, i.e. ROM-based This is a data segment for executable programs, and this includes data from ROM. Initialized variables and P needed to directly execute a ROM-based program SP is included. Therefore, programs related to relocation are removed from the ROM base. The loader 1100A has a data area and R Set up the necessary state to perform OM-based library functions. and make it more effective.

The present invention provides a novel loader for ROM-based executable programs as described above. In other applications of the carder, it functions based on a library of some functions to ROM. Application programs are held in ROM. In this example, the The data reader l100B (FIGS. 59A to 59D) is a 1.00M file. B The programmer lll0B performs the steps required to run a ROM-based program. In addition to running ROM-based application programs as resident programs, It also includes a step to cent up as a ram.

In this embodiment, data area initialization means 11.10B and The structure of the application existence confirmation means 1120B is the same as that of the means 1110 in the previous embodiment. A and 112OA. Structural changes provide the fastest response. provided to the user. Based on this explanation, a person skilled in the art would be able to do this directly from the ROM. environment of a computer system to run ROM-based applications In addition to the basic purpose of configuring the structure of each element of the loader function of the present invention This shows that it is possible to achieve the desired speed and other functions using

As mentioned above, when the operating system executes the 00M file , a specific segment of memory is allocated, namely 64 kilobytes of memory. . Therefore, the operation of work data area creation 1110B-1 (Figure 59A) is simply Stunk to data area allocated by operating system The only thing you can do is set the knob.

In this example, each ROM-based application ensures that the RAM is The data area is set up so that it is initialized, so the data area is It is not zeroed by the controller lll0B.

In this example, the ROM-based application is a library of the above functions. is designed to operate under

Therefore, the data area initialization means 1110B-2 first performs the memory check 11. Are ROM-based library functions loaded using 10B-2-1? Determine whether or not. In particular, interrupt vectors for ROM-based library functions. Is the associated code a character string code of “VNA,,,”? checked to see if If library functions are not loaded , processing moves to error message 1150B. Error message 115 0B reports to the user that the library function is not found and the process , typically terminated via termination 1140B. End 1140B is the end 11 mentioned above. Performs the same function as 40A.

Map 11.10B-2-2 is a pending U.S. patent application assigned to applicant. John in application tube 07/375.72'1.

John P, Fairbanks et al. )'s [Portable Loin Power] Computer) J, filed June 30, 1989, and assigned to applicant. Pending U.S. Patent Application No. 07/374.691, Ian H, Varnish , “Computer System” by Ian H,S, Cullimore Information management method and apparatus in the system s for Infor + wation Management In a C computer System) J, filed June 30, 1989. Access the interrupt function in the computer's expansion BIOS to set the address segment. Set the comment EOOOh as the TRAMUBE-3U application. child The disclosures of these two applications are hereby incorporated by reference. In general, expansion BIO 3 is either BiO2ROM or ROM memory card in the computer is commanded to address the segment.

When the map 1110B-1-2 is completed, the process is passed to the confirmation means 1120B. . The application specified by loader l100B is ROM check 11 20B-1 (Figure 59B), (i.e., the string " If "PQTL" is "detected") processing continues. However, the application If no option is found in ROM, processing returns error message 1150B. will be moved to

Error message 1150B indicates that the application does not exist in ROM. is reported and processing is terminated normally.

When the ROM check 1120B-1 is finished, the operating system version John checked with Operating System Check 1120B-2. The operating system is compatible with ROM-based applications. It is confirmed that there is. If your operating system is not compatible, An error message indicating that the operating system is unable to report an error. page 1150 is generated and processing is normally terminated.

Next, memory check 1120-B-3 (Figure 59C) checks the ROM application. Determine whether the version is already loaded on the computer system. loader Similar to the I100A, ROM-based applications using a library of functions The initialized word is used to store the data segment for the Ru. In this example, the word is initialized to zero. The application is the appropriate word is set to a non-zero value. Therefore, the memory chip Check 1120-B-3 simply determines whether the word is zero or non-zero. and determine whether the application is already loaded. ROM base If the application is not loaded, processing will proceed to the sensor detailed below. The process moves to setup 1130B.

However, if a ROM-based application is already loaded Lookup 1120B-4 is PSP command line check 1120B- Access 5.

Similar to the previous embodiment, the look amplifier includes data area initialization means 1110A. include. This is because one of the structures constructed by the loader of the present invention has operational characteristics. This indicates that configuring a fixed allocation is not an essential feature of the invention. Therefore, the book Many variations of the inventive loader will be apparent to those skilled in the art.

Command line check 1120B-5 determines whether the user entered “U” or not. Determine. If the command line has a “U”, the command line process The processor 1120B=6 releases the memory allocated to the loader 1100B, and Zero the word indicating that the application is loaded. The processing is It is then normally terminated. However, if the command line does not have "U" error message 1150B if the parameter is not specified or has no parameter. reports that the application is already loaded. command line processing is supported by both loaders 1100A and 1100B. , command line processing is not required for proper operation of the loader. command Dryline processing simply provides the user with additional ROM-based executable programs. It only provides a means of control.

The setup 1130B for the second RAM-based application is described above. This is more complicated than the setup 1130A of loader 1100A. cent up The first operation at 1130B is to create a RAM data area for the application. Variable initialization 1130B-1 performs initial settings of resident variables in the application. In this example , there are two jump destinations, the alarm flag is cleared, and the interrupt vector is file is stored.

Alarm flags are used for audible alerts in ROM-based applications. Ru. The INDOS counter and disk transfer address are also initialized.

In setup 1130B, register set 1130B-2 is next registered. Data AX is initialized for jump after data memory allocation for TSR. Ru. More 1130B-3 (Figure 59D) interrupts for re-vectoring and TSR Rearrange the code to secure capacity for memory data allocation, and register AX is reset to place the revectored code. Processing is done by Regis Jump to the revectored position of data AX. TSR interrupt reset 113 At 0B-4, the TSR interrupt is reset and the interrupt vector code Allows for relocation. The entry code for ROM-based applications is encoded. This is called by the triggering means 1130B-6 to complete the initialization. Finally, open Means 1130B-7 frees up unnecessary RAM and installs the TSR program. ROM-based applications configured as RAM remain.

For example, the accuracy of the initialization of certain variables performed by Setup 1130B. This operation is not a necessary requirement for the features of the present invention. The important feature is that the ROM base The loader performs the necessary relocation and initialization for the application to work at the base. ROM-based applications function correctly as TSR programs. It is to be able to function.

Pocketlink adds ROM-based applications to your computer system. Further embodiments of the loader may be used. This loader , allocated by the operating system when the loader is executed. Requires less memory than memory. Therefore, the 20 loader is shown in Figure 57B. Free the memory allocation as shown.

Also, ROM-based executable programs can be erased by the loader, as mentioned above. Once the loader operation is complete, it is executed directly as a non-TSR application. It will be done. As mentioned above, after the loader itself is erased, the ROM-based pocket The link application's PSP and data segments remain in RAM. This fruit In some embodiments, the ROM-based PocketLink provides additional memory as needed. be assigned.

The various embodiments of the invention described above are applicable to 18M personal computers and IBM It is designed to run on a personal computer. However However, these examples are merely illustrative of the invention and are not limited to the specifics described above. The invention is not meant to be limited to the examples.

Based on the disclosure herein, those skilled in the art will be able to implement the novel loader of the present invention into other computers. architecture and can be implemented in other computer programming languages. It is obvious that it can be done.

Conventional technology Conventional technology 2-12-2 2-12-22-N Conventional technology Sun G, 17A Day G, 17th Sun G, 18A Sun G, 19A Snake G, 198 FIG, A of 2 Day G, 21A FIo, 21st 0.22A per day Mouth G, 23A Day G, 23rd Mouth G, 248 Sun G, 25A Mouth 0.25 days Sun G, 26A Day G, 278 Sun G, 28A Snake G, 29 0G, 30 Mouth 0.31 FIG 32-1 Mouth G 32-2 Japan G34-1 KEY To FIG 34 Sun G 34-2 Mouth G 39B-2 Mouth Gヰ7-1 0G 47-2 KEY To Ro G 47 Japan G 47-3 Special table Hei 5-502528 (66) Mouth G 48-1 0G 48-2 FIG. 54-1 Day G 55A 11 no no-1 Day G 558 Day G, 57A RoG, 57th Day G, 57C international search report 1MmMI+alIal^””””+1-trr/+<(ffi/n<1!R

Claims (58)

[Claims] 1. first computer means; second computer means; the first and second computers are operably coupled; means for transferring information between computer means; operating in each of said computer means (i) in said first mode of operation; between said first computer and said second computer means via said transfer means; (ii) in a second mode of operation, said first controller; forming a virtual computer comprising computer means and said second computer means; virtual network management means, The virtual network management means sets the computer means as a virtual computer. the first computer means controls the virtual computer when configured to and said second computer means receives a command from said first computer means. a user application running on said first computer means; applications to access all information stored within the virtual computer. A device characterized by: 2. The virtual network management means is a user application of one of the computer means. a first message structure that sends and receives information and includes said information in response to a request; The first means of generating operatively coupled to said first means for receiving and transmitting said information and for sending a second message; a second means for generating a page structure; The second means includes a header, at least a part of the first message, and an error code. receiving said first message structure including a code; further coupled to said second means and said transfer means for receiving and transmitting said information; and wherein said third means blocks said second message structure. via said transfer means with predetermined transfer parameters including size and transfer rate. 2. The apparatus of claim 1, adapted to receive and transmit. 3. The first means is: compressing the information before generating the first message structure, and compressing the compressed information. Claim 2 further comprising data compression means for use in said first message structure. equipment. 4. When the first means receives the compressed information, the first means compresses the compressed information. 4. The apparatus according to claim 3, further comprising restoring means for restoring. 5. The second message structure is composed of data byte information, and the second message structure is composed of data byte information. The means are the header of the second message structure and at least the inside of the first message. Generates an error code based on the disposition of each byte of information, and if any byte Claim 2, further comprising means for generating a different error code when the error code changes. equipment. 6. The second means includes (i) the size of the second message and (ii) the size of the second message. used by said third means to send said second message via a transfer means; including means for adjusting predetermined transfer parameters, including the transfer rate to be transferred; At least one of the message size and transfer rate is determined by the sending of the second message structure. The device according to claim 2 is adjusted when a predetermined number of errors occur in the communication. Place. 7. said first computer device for transferring information between said first and second computer devices; and a second computer device, in the first mode of operation, the transfer means; transferring information between said first and second computer devices via said second computer device; mode of operation, the first computer and the second computer device form a virtual computer by When said computer means is configured as a virtual computer, said first Computer means control said virtual computer and said second computer means a step for servicing commands from said first computer means and for servicing said first computer means; A user application running on a computer means is a virtual computer. The two features are that all information stored in the data can be accessed. A method of communication between computer devices. 8. said first computer device for transferring information between said first and second computer devices; and a second computer device, operating on each of said computer means, (i) said first computer via said transfer means in a first mode of operation; (ii) transferring information between the computer means and said second computer means; mode of operation, said first computer means and said second computer means A virtual network management means is provided to form a virtual computer by means of The virtual network management means uses the computer means as a virtual computer. when set, said first computer means controls said virtual computer; , said second computer means responds to a command from said first computer means. a user application executed on said first computer means; application to access all information stored within the virtual computer. A communication method between two computer devices, characterized in that: 9. the first computer means has a first virtual network management means, and the first computer means comprises a first virtual network management means; In response to each command of the commands, (i) in a first mode of operation, a first through a connecting means interposed between the computer means and the other computer means. transferring information between said first computer means and said other computer means; (ii) in a second mode of operation, said first computer means; forming a virtual computer by said other computer means; - executing an application, said first computer means executing said first user; The application generates the command by issuing a command in the set of commands. access virtual network management means; The second computer means has a second virtual network management means, and the second computer means comprises a second virtual network management means and a (i) in the first mode of operation, the second command; the computer means and the other computer means via a connecting means interposed between the computer means and the other computer means. transferring information between said second computer means and said other computer means; (ii) in a second mode of operation, said second computer means and said form a virtual computer by means of the other computer and create a second user account; the second computer means executes the second user application; the virtual application by issuing a command in the set of commands. access network management means; said first and second computer means operatively coupled to said first and second computer means; comprising means for transferring information between the computer means; said second user application to said second virtual network management means; generating a server command, said second virtual network management means said second virtual network management means; configuring the computer means as a server, the first and second computers means constitute a virtual computer, a first computer means executed on said first computer means; One user application has control over the virtual computer. A device characterized by: 10. operatively coupled to transmit and receive information via a data link A virtual network that runs on a computer system with multiple computers. network management function, Data linking using predetermined transfer parameters including block size and transfer rate. means for transmitting and receiving information via a network; means for detecting errors in the transfer of information via the data link; means for adjusting the transfer of information upon detecting a predetermined number of errors in said transfer; A virtual network management function characterized by being configured by. 11. The transfer and reception means, means for generating a message structure, said message structure including a header and said information; the message structure includes a portion of the information and an error code, and the message structure is 11. A virtual network management function according to claim 10, which is transmitted and received by the means. 12. The message structure is made up of bytes, and the transfer and reception means include: Based on the processing of information bytes of the message header and at least some of the information. has a means of generating an error code and detects an error when any byte changes. The virtual network according to claim 11, characterized in that the virtual network generates an error code Work management function. 13. The error detection means is operatively coupled to said transmitting and receiving means and configured to include headers and a header of a received message; and generating a second error code regarding at least some of the information received. , comparing said second error code with the received error code; Claim 1 detects a transfer error when the first and second error codes are different. Virtual network management function in Section 2. 14. The means for coordinating the transfer of said information includes: (i) the size of said message structure; and (ii) means for adjusting predetermined transfer parameters including the transfer rate; If a predetermined number of errors occur during the transfer of the second message structure, at least The assumption of claim 13 that adjusts one of the message size and the transfer rate. Virtual network management function. 15. operatively coupled to transmit and receive information via a data link A virtual network that runs on a computer system with multiple computers. network management function, Sending and receiving information in response to a user application on one computer means first means for generating a first message structure including said information to a second means operably coupled to said first means for receiving and transmitting said information; a second means of generating a message structure; The second means receives the first message structure and receives the second message structure. The structure includes a header, at least a portion of said first message structure, and an error code. and further includes said second means and data stream for receiving and transmitting information. third means operably coupled to the second link; The message structure is defined over the data link, including block size and transfer rate. A virtual network characterized by receiving and transmitting data using transfer parameters. management function. 16. The first means is: compressing the information before generating the first message structure, and compressing the compressed information. Claim 15 further comprising data compression means for use in the first message structure. Section Virtual Network Management Features. 17. When the first means receives the compressed information, the first means compresses the compressed information. 17. The virtual network management function according to claim 16, further comprising a restoring means for restoring. 18. said second message structure is comprised of information bytes of data; said second message structure is comprised of information bytes of data; The means of a header of said second message structure and at least an internal part of said first message; generates an error code based on the treatment of each byte of information in Claim 1 has means for generating a different error code when the error code changes. Virtual network management function in Section 5. 19. The second means includes (i) the size of the second message and (ii) the size of the second message. by said third means for transmitting said second message via said transfer means; including means for adjusting predetermined transfer parameters, including the transfer rate to be used. and at least one of the message size and transfer rate is determined by the second message structure. Claim 1: Adjustment is made when a predetermined number of errors occur in the transmission of the structure. Virtual network management function in Section 5. 20. transfer information via a data link with predetermined transfer parameters; Detects errors during transfer over a data link and detects a predetermined number of errors during transfer. through a data link that adjusts the transfer parameters when it detects A method of transferring information between at least two coupled computer systems. 21. Before the forwarding and receiving steps, a message structure is generated and the message the message structure includes a header, at least some of said information and an error card; Claim 2, wherein the message is transmitted and received in said transmitting and receiving step. 0 term method. 22. The message structure consists of bytes and includes a header and a header of the message. Based on the processing of each byte of at least some of said information, Generate a different error code when the target changes, and change the error code. 22. The method of claim 21, comprising the step of generating. 23. Said error detection step detects the received header and the received at least information. generated a second error code regarding a part of the system, and received the second error code. Transfer error occurs when the first and second error codes are different compared to the error code 23. The virtual network management function of claim 22, comprising the step of detecting. 24. The step of adjusting the transfer parameters includes (i) adjusting the message size and and (ii) adjusting at least one of the transfer speeds. Virtual network management function in Section 23. 25. header means indicating the direction of information transfer and the destination of the information; A computer-to-computer computer comprising a means for storing information adjacent to a header means; A structure used to transfer information. 26. The header means: 26. The structure of claim 25, further comprising means for indicating the origin of said structure. 27. The header means: having means for specifying an action to be taken using the information stored in said structure; The structure of claim 26. 28. Each pair of said structures is uniquely associated with one of the operations to be performed, and each pair of said structures is uniquely associated with one of the operations performed. Claim 27, wherein one structure of the pair is transmitted and the other structure of the pair is received. structure. 29. The operating system and random access memory (RAM) a loader for executable programs stored in read-only memory (ROM); and a controller to which the RAM is allocated to the loader when the loader is executed. In the computer device, the loader loads an executable program stored in the ROM. means for initializing said computer means to execute a program; and said initializing. and means for freeing the RAM, operatively coupled to the means for releasing the RAM; When the initialization of the computer device by the initialization means is completed, the A release means releases the RAM required by the loader itself, and releases the data stored in the ROM. The executable program can be executed directly from the ROM in the computer device. A loader characterized in that it is set to 30. computer to run executable programs stored in ROM means for initializing the device; operatively coupled to the initializing means; said computer system. and initialization of the computer device by the initialization means. When the loading is finished, the releasing means uses the loader itself to release the RAM and Setting the program stored in the ROM to be executed directly from the ROM An executable program stored in read-only memory (ROM) characterized by loader. 31. The initialization means has means for initializing the data area, and the initialization means has means for initializing the data area. The area initial setting means sets the data area of the RAM to the execution area stored in the ROM. 31. The loader according to claim 29 or 30, wherein the loader is set to a possible program. 32. The initialization means is for an executable program stored in the ROM. means for creating a work RAM data area, and the work data area creation means 32. The loader of claim 31, wherein the step sets the size of the RAM data area. . 33. The initialization means is operably coupled to the work data area creation means. The data initial setting means has a data initial setting means that is configured to set data in the RAM data area. Variables necessary to run the executable program stored in the ROM in 33. The loader of claim 32. 34. The initialization means includes data initialization means, and the data initialization means includes: Executes an executable program stored in the ROM in the RAM data area. 32. The loader of claim 31, wherein the loader initializes variables necessary to perform the process. 35. an application operably coupled to the initializing means and the opening means; The application has a means for confirming the existence of the application, and the means for confirming the existence of the application is Confirm the existence of an executable program stored in ROM in the ROM in the data device. The loader according to claim 31. 36. The existence confirmation means is configured to check whether the executable program stored in the ROM has already been copied. Claims having means for determining whether or not the computer device has been loaded Loader of item 35. 37. 32. The loader of claim 31, wherein said opening means comprises set-up means. Da. 38. The setup means is an executable program stored in the ROM. 38. The loader according to claim 37, further comprising means for setting the program as a resident program. 39. Read-only to execute executable programs stored in ROM data area initial setting means for setting a data area of memory (RAM); operably coupled with the data area initialization means and stored in a ROM in a ROM; An application existence verification method that checks for executable programs that have been stored in a read-only memory (ROM) characterized by being configured by Loader for executable programs. 40. The data area initialization means is configured to initialize an executable program stored in the ROM. and means for creating a work RAM data area for the work data. The area creating means sets the size of the RAM data area. Loader of section 9. 41. The data area initialization means is combined with the work data area creation means. and has means for initializing data, and the data initializing means is for the RAM data area. Claim 40: Initializing variables necessary for the ROM-based execution program within the application. Term loader. 42. The data area initialization means has a data initialization means, and the data area initialization means has a data initialization means. The period setting means sets an executable program stored in the ROM in the RAM data area. A loader according to claim 39 that initializes variables necessary for executing the program. . 43. The existence confirmation means is configured to check whether the executable program stored in the ROM has already been copied. Claims having means for determining whether or not the computer device has been loaded Loader of item 35. 44. The setup means is an executable program stored in the ROM. 40. The loader according to claim 39, further comprising means for setting the program as a resident program. 45. computer to run executable programs stored in ROM Steps for initializing the device and freeing up random access memory (RAM) Consists of step and When the initialization of the computer device in the initialization step is completed, the initialization Frees the RAM used in the step and creates an executable program stored in the ROM. A lead-on characterized in that the program is set to be executed directly from ROM. Executes an executable program stored in remote memory (ROM) from ROM. How to set up a computer device. 46. The initialization step is an executable program stored in the ROM. Claim 4 further comprising the step of allocating a data area in the RAM for Method in Section 5. 47. a step in which the initializing means sets the size of the data area of the RAM; 47. The method of claim 46, comprising: 48. The initialization step stores data in the ROM in the data area of the RAM. Contains steps for initializing variables needed to run the executable program The method of claim 47. 49. The initialization step stores data in the ROM in the data area of the RAM. Contains steps for initializing variables needed to run the executable program The method of claim 45. 50. an executable program stored in the ROM in the computer device; The method of claim 45, further comprising the step of checking for the existence of the program. . 51. The checking step includes an executable program stored in the ROM. is already loaded on the computer device. 51. The method of claim 50. 52. The executable program stored in the ROM resides in the computer device. 46. The method of claim 45, comprising the step of programming and configuring. 53. Random access memory for executable programs stored in ROM a step of creating a data area in RAM; an executable program stored in a ROM in a computer device; A read-only method comprising a step of checking the existence of Execute an executable program stored in memory (ROM) from ROM How to set up your computer equipment. 54. The creation step includes a step of setting a data area of the RAM. The method of claim 53. 55. The above creation step is performed by storing data in the ROM in the data area of the RAM. The program includes steps for initializing variables necessary to run the executable program. 55. The method of claim 54. 56. The above creation step is performed by storing data in the ROM in the data area of the RAM. The program includes steps for initializing variables necessary to run the executable program. The method of claim 53. 57. The above checking step is performed by checking the executable program stored in the ROM. the step of determining whether the program is already loaded on the computer device; 54. The method of claim 53. 58. The executable program stored in the ROM resides in the computer device. 54. The method of claim 53, comprising the step of programming and configuring.