JPH0619751B2 - Command transfer method to multiple devices - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Outline] A command transfer method to a plurality of devices in a data processing system including a management unit such as a host computer, a plurality of peripheral units, and a common transmission path connecting these units, In order to enable efficient command transfer without increasing the number of signal lines and the hardware for transmission and reception compared with the conventional method, start execution unique to each peripheral unit. A code is assigned, and when the management unit transfers a command to any of a plurality of peripheral units, it creates a command in which the execution start code of the transfer destination peripheral unit is added to a series of instruction codes and creates this command. It is sent to all the peripheral units via the common transmission line, and each peripheral unit uses its own execution start code in the received command. A station is configured to execute its command only when specified.

[Industrial application field]

The present invention relates to a command transfer method for transferring a command from a host computer or the like to a peripheral device or the like in a data processing system to a plurality of devices.

As a command transfer method, it is necessary to transfer commands quickly and efficiently without increasing the hardware scale of the system.

[Conventional technology]

As a system for transferring a command from a host computer to a peripheral device, there is a system as shown in FIG. 7, for example.
This is one in which the host computer 1 and each peripheral device 2 to 2 are connected by different command cables 3 to 3, respectively, and the host computer 1 only via the command cable to the peripheral device to which the command is to be transferred. Command transfer is performed individually.

This method has the advantage that the control is simple and the transfer speed is fast, but on the other hand, when the number of peripheral devices 2 increases, the number of command cables 3 increases, and as a result, it is difficult to pull down the cables. This leads to problems such as an increase in the number of host computer ports and an increase in the space required for cable wiring.

Therefore, as a command transfer method capable of controlling a large number of peripheral devices while reducing the number of command cables, there is a system using a bus type network as shown in FIG. This is a host computer 1 and peripheral devices 2 to 2.
The cables are connected to each other by a GP-IB type cable 4, and a command is transferred from the host computer 1 to a target peripheral device according to a predetermined communication control procedure. As the communication control procedure, for example, the data transmission method is instructed from the host computer 1 to all the peripheral devices 2 to 2: Waiting for all the peripheral devices 2 to 2 to send back an ACK signal to the host computer 1: From the host computer 1 to the peripherals Designate the source device and the destination device in the devices 2 and 2: Notification from the host computer 1 that data (command) will be sent from now: Actual data transfer from the host computer 1:

[Problems to be Solved by the Invention]

In the above command transfer method using the bus network, the communication control procedure (protocol) is complicated and it takes time to instruct the data transmission method. That is, the overhead is large. In addition, the hardware becomes complicated.

Therefore, an object of the present invention is to enable efficient command transfer to a peripheral device without using a complicated protocol without increasing the number of signal lines, the transmission hardware, and the reception hardware as compared with the related art. is there.

[Means for Solving the Problems]

FIG. 1 is a diagram for explaining the principle of the present invention, and FIG. 2 is a diagram showing a schematic configuration of a data processing system for carrying out the method of the present invention.

A method of transferring a command to a plurality of devices according to the present invention includes a management unit 11 such as a host computer and a plurality of peripheral units 1.
2 to 12 and a common transmission line 13 that connects them
And a unique execution start code is assigned to each of the plurality of peripheral units 12 to 12,
The management unit 11 includes the plurality of peripheral units 12 ...
When executing any of the 12 command commands, a variable length command in which the execution start codes of a plurality of transfer destination peripheral units are added to a series of command commands is created and transmitted to all peripheral units 12 to 12. Each of the peripheral units 12 to 12 is configured to hold the received variable length command, check the execution start code in the command, and execute the command command only when the own station is designated by the code. .

[Action]

Now from the management unit 11 to the peripheral units 12 to 12
Of these, for example, transfer the command to the unit 12,
It is assumed that the peripheral unit 12 executes the command.

As the command to be transferred, for example, the command shown in FIG. 1 [A] is used. Here, the word length of one code is 8
Bits are displayed in hexadecimal code. Among these codes, a code with "F" at the beginning such as "F *" is used as a control code for a special purpose. In this example, "FF" is a reset code and "F0" to "FE". "Is the execution start code.

The execution start codes "F0" to "FE" are assigned to the peripheral units 12 to 12 as unique codes, as shown in FIG. In this example, since it is an 8-bit code, the execution start code "F0"-
Peripheral units 12-12 to which "FE" is assigned
The maximum number of cars will be 15.

Other codes are used as command words or data. In the example of FIG. 1 [A], "D8" as an instruction code for specifying the operation type and "2" as an auxiliary code.
0 "," 4F "as the data transmission destination code, and" C0 "," D0 "as the data reception destination code form a series of instruction code strings.

The command structure is such that the reset code "FF" is first, followed by the instruction code string "D8" to "D0", and finally the execution start code "F0". This command is shared by the management unit 11. It is sequentially sent to all the peripheral units 12 to 12 via the transmission line 13.

Each of the peripheral units 12 to 12 has means for temporarily holding the sent command, and when the reset code "FF" which is the first word of the command is received, the contents of this holding means are cleared. Then, successively input instruction code strings are sequentially stored, and the content of the last received execution start code is checked.

As a result of this check, when the received execution start code specifies the own station, the operation is started to execute the instruction content held in the holding means.

After the operation is completed, the management unit 11 is notified of the completion of the operation via another signal line.

On the other hand, if the own station is not specified, the peripheral unit ignores the command and does not execute the command.

In the next command transfer cycle, each of the peripheral units 12 to 12 receives the reset code "F" at the head of the command.
Since the contents of the holding means are reset at F ″, the normal operation is always performed in each transfer cycle regardless of whether it is designated in the previous transfer cycle.

When it is desired to cause a plurality of peripheral units to execute the same command in one command transfer cycle, a plurality of execution start codes may be sent in succession as shown in FIG. 1B. In the example of FIG. 1 [B], the start code “F0” of the peripheral unit 12 and the start code “F of the peripheral unit 12”
4 "are sent continuously, which causes the peripheral units 12 and 12 to execute the same command.

Embodiments Embodiments of the present invention will be described below with reference to the drawings.

FIG. 4 shows a data processing system using a command transfer method as an embodiment of the present invention. This system is disclosed, for example, in JP-A-59-184973 and JP-A-60-
218183, Japanese Patent Application Laid-Open No. 63-113782, Japanese Patent Application No. 63-172796, and the like, which is an image processing system as a visual sensor for a robot or the like,
A plurality of image processing peripheral units 20 for performing various image processes such as image contour extraction, a host computer 40 for managing and controlling the plurality of peripheral units 20, and a connection between the host computer 40 and the peripheral units 20. The data ways 51, 52 and the control bus 53 are included. In the figure, only one peripheral unit 20 is shown.

The peripheral unit 20 includes a controller board 21, data way interface (I / F) boards 22, 23, a program load control board 24, a buffer memory board 25, a data monitor board 26, and a plurality of mapping boards 2.
7 to 27, a monitor 28, a data / control bus 29 and the like.

The controller boat 21 is a board that receives control information such as commands from the host computer 40 via the control bus 53 and controls the operation of each board in the unit.

Data way I / F boards 22 and 23 are data way 5
It is a board that takes in data from 1, 52 respectively and transfers it to each board in the unit. The mapping board is a board that processes image data by parallel calculation.
Seven boards are arranged in parallel.

The program control board 24 parallel-serial converts the microprogram given from the data ways 51 and 52 via the data way I / F boards 22 and 23 to load the load line 20.
It is a board that performs control such as loading onto the mapping boards 27 to 27 via the control unit 1.

FIG. 5 shows the detailed structure of the controller board 21. As shown, the AND circuits 30 to 33, the comparator 3
4, 35, a counter 36, a decoder 37, a flip-flop 38, a gate buffer 39 and the like.

This controller board 21 has a host computer 4
Control information such as command data having a width of 0 to 8 bits is loaded, and the host computer 4
The strobe pulse from 0 is also input. It is assumed that the command has the same format as that described above with reference to FIGS. 1 and 3. As described above, the reset code "FF", the instruction code, the auxiliary code, the data transmission source code, the data reception destination code, and the execution start code.

The instruction code is a code for designating the operation type, and for example, the memory start address of the microprogram can be designated. The auxiliary code is, for example, a code used to specify loading of a micro program, and has a meaning of specifying all mapping boards when "4F".

When the data transmission source code is "80", for example, it is a code for designating the data way I / F board. The data receiving destination code is a code for designating the program load control board when it is "E0", for example.

The upper 4 bits and the lower 4 bits of the command data are input to the AND circuits 30 and 31, respectively.
The output signals of the ND circuits 30 and 31 are input to the AND circuit 32. With this configuration, the AND circuits 30 to 32 function as a detection circuit of the reset code "FF" in the command, and the detection signal is reset and input to the counter 36 and set to the flip-flop 38, and the bus 29 Is output to each board in the unit 20 as a reset signal ST.RES.

The lower 4 bits of the command are input to the P input terminal and the output signal of the AND circuit 30 is input to the G input terminal of the comparator 34, and the Q input terminal is uniquely assigned to each unit 20. A fixed code of any of "0" to "E" has been input. When the P input and the Q input match and the G input is received, the output signal is output to the AND gate 33.

An 8-bit command is input to the P input terminal of the comparator 35, and a fixed code of "4F" is input to the Q input terminal. Therefore, the comparator 35 has a "4F" detection function and a "4F" detection function. Upon detection, the ST / ALL signal is output to the control bus 20.

FIG. 6 is a time chart of each signal in the controller board 21. The operation of the controller board 21 will be described with reference to FIG.

When commands as shown in FIG. 1A are sequentially input from the host computer 40 in 8-bit parallel and word serial, the reset code "FF" is first detected by the AND circuits 30 to 32 and synchronized with the strobe pulse. And a reset signal ST.RES is output, which resets the counter 36 and causes the flip-flop 38 to
Thus, the gate buffer 39 is opened, and thereafter, the code data at that time is sent to each board in the unit 20 via the data bus 29 via the gate buffer 39. Further, this reset signal ST.RES is sent to each board in the unit 20 through the ST.RES line of the control bus 29 to reset those boards.

When "D8" as the instruction code is next input, the counter 36 counts up and the decoder 37 obtains the pulse shown in FIG. 6 (E), which is sent to the ST.MOD line of the control bus 29. It

Next, when "20" as the auxiliary code is input, the pulse shown in FIG.
Transmitted to UX line. Next, when the data transmission source code is input, the pulse shown in FIG.
To the ST / SRC line. In particular, when the code "4F" is input, this is detected by the comparator 35, the pulse of FIG. 6 (G) is sent to the ST / ALL line, and all the mapping boards 27 to 27 are designated. The input of the data receiver code can be repeated until the input of the execution start code. Therefore, at the time of the repetition, the data of the code and the pulse of FIG. 6 (H) are repeatedly transmitted. Each of these pieces of code data will be held in the destination board.

Finally, when the execution start code is given, the comparator 34
It is determined whether or not the execution start code is the code assigned to the own station, and if it is the code of the own station, a detection signal is output to the counter 36, whereby the pulse of FIG. 6 (I) is output to the control bus. It is sent to the inside ST / START line. As a result, each instruction code received earlier is started to be executed by each board in the unit 20.

Various modifications are possible in carrying out the present invention. For example, in the above-described embodiment, the bus type network is used to send a command from the host computer 40 to each peripheral unit 20, but the present invention is not limited to this, and a ring type network is used, for example. Alternatively, for example, a command cable may be connected between the host computer 40 and the first peripheral unit 20 only, and for other peripheral units, between the controller boards of the peripheral units. It is also possible to connect to the cascade with a cable and transfer the command sent from the host computer 40 from the first peripheral unit to the other peripheral units in sequence.

Further, in the above example, since the command uses an 8-bit code, only “OO” to “EF” can be used as data and commands, and the number of peripheral devices is 1 from “F0” to “FE”.
There is a limitation that it is limited to five. Therefore, by expanding the transmission line from 8 bits to 9 bits, it is possible to eliminate the above restriction. In that case, all 8 bits as data, that is, "0 ₂ 0 ₁₆ 0 ₁₆ " to "0 ₂ F ₁₆ F"
₁₆ "can be used. Here, the subscripts 2 and 16 represent a binary number and a hexadecimal number, respectively. When the reset code is" 1 ₂ F ₁₆ F ₁₆ ", the peripheral device has" 1 ₂ 0 _{16 ".} 0 ₁₆ 255 _{"~" 1 2 F 16 E} 16 "
It is possible to connect up to a stand.

EFFECTS OF THE INVENTION According to the present invention, it is possible to efficiently transfer a command to a peripheral device without using a complicated protocol without increasing the number of signal lines, the transmission hardware, and the reception hardware as compared with the related art. become.

[Brief description of drawings]

FIG. 1 is a diagram for explaining the principle of the present invention, FIG. 2 is a diagram showing a schematic configuration of a data processing system for performing the method of the present invention, and FIG. 3 is a diagram for explaining an operation example of the method of the present invention. FIG. 4 is a block diagram showing an image processing system using a command transfer method as an embodiment of the present invention, FIG. 5 is a block diagram showing a detailed configuration of a controller board in FIG. 4, and FIG. FIG. 7 is a time chart of signals of various parts of the controller board, and FIGS. 7 and 8 are diagrams for explaining a conventional command transfer method. In the figure, 1 ... Host computer 2-2 ... Peripheral device 3-3, 4 ... Command cable 11 ... Management unit 12-12, 20 ... Peripheral unit 13 ... Common transmission line 21 ... Controller board 22 , 23 ...... Data way I / F board 24 ...... Program load control board 25 ...... Buffer memory board 26 ...... Data monitor board 27-27 ...... Mapping board 28 ...... Monitor 29 ...... Data / control bus 40 ...... Host computer 51, 52 ... Data way 53 ... Control bus 30-32, 33 ... AND circuit 36 ... Counter 37 ... Decoder 38 ... Flip-flop 39 ... Gate buffer

Claims (1)

[Claims] 1. A management unit (11), a plurality of peripheral units (12-12), and a common transmission line (13) connecting these units, the plurality of peripheral units (12-12). A unique execution start code is assigned to each of the plurality of peripheral units (1).
2 to 12), in order to execute an instruction command, a variable-length command in which execution start codes of a plurality of transfer destination peripheral units are added to a series of instruction commands is created to generate all peripheral units (12 to 12). 12), each peripheral unit (12 to 12) once holds the received variable length command, checks the execution start code in the command, and executes the command command only when the own station is designated by the code. A method of transferring a command to multiple devices configured to execute.