JPH0214333A - Debug device for data flow program - Google Patents

Abstract

PURPOSE:To easily and exactly execute the development of a data flow program by obtaining the execution information of a source program with being represented in the execution order and parameter of the source program which corresponds to the execution of the data flow program. CONSTITUTION:A debug executing device 10 simulates the internal operation of a data flow information processing unit by data from an input file 20, the various types of instructions concerning the debug of a command file 30 and the various types of information in an internal file 50. Then, the executed result of a data flow graph is stored to a result file 70. On a debug display file 60, display for the three types of a node number 100 to be executed during debug execution, a processing start 210, a row 220 to be just executed and a processing end row 230 is executed. Thus, the development of the data flow program can be easily and exactly executed.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a debugging device for a data flow program, and in particular, it is possible to debug a source program by displaying each line of the source program developed into a data flow program. Regarding debugging equipment.

[Prior Art] A conventional computer is a Neumann type computer in which various instructions are stored as programs in a program memory, and a program counter sequentially specifies addresses in the program memory to sequentially read out the instructions and execute the instructions. Or most of it.

On the other hand, a data flow type information processing device is a type of non-Neumann type computer that does not have the concept of sequential execution of instructions using a program counter.

Such a data flow type information processing device follows an architecture that assumes that instructions are processed in parallel. As soon as the data to be operated on is available, instructions can be executed. Since multiple instructions are simultaneously driven by the data, the program is executed in parallel according to the natural flow of data. As a result, the time required for calculation can be significantly reduced.

FIG. 5 is a schematic block diagram showing an example of a conventional data flow type information processing device, and FIG. 6 is a diagram showing the basic structure of a data packet circulating through each part of the information processing device shown in FIG. 7 is a diagram showing a part of the data flow program stored in the program memory.

First, the configuration and general operation of the data flow type information processing apparatus will be described with reference to FIGS. 5, 6, and 7. Destination information is stored in the destination field of the data packet shown in FIG. 6, command information is stored in the command field, and operand data is stored in the data 1 field or data 2 field.

In FIG. 5, the program storage unit (PS) 1 includes a program memory, and as shown in FIG. 7, a data flow program consisting of destination information and command information is stored in the program memory. The program storage unit 1 reads destination information and instruction information by addressing based on the destination information of the data packet, stores each information in the destination field and instruction field of the data packet, and outputs the information to input buffers 2a and 2b consisting of FIFOs. .

A pair data detector (FC) 2 waits for data packets input from the program storage unit 1 via input buffers 2a and 2b, and selects one of the two data packets with matching destination information. The operand data of the other data packet is stored in a predetermined data field of the other data packet.
Output to a. Note that at this time, one of the data packets mentioned above is destroyed.

The arithmetic processing unit (FP) 3 decodes the command information of the data packet input from the paired data detection unit 2 via the input buffer 3a, performs predetermined arithmetic processing on the two operand data, and outputs the result. It is stored in the data field of the data packet and output to the branching unit 4.

Based on the destination information of the data packet, the branching unit 4 either supplies the data packet to the input buffer 1a made up of FIFO, or sends the data packet to the input buffer 5a made up of FIFO.
It is applied to an external data memory (EDS) 6 via the external data memory (EDS) 6. Data packets from the input buffer 1a or data packets output from the external data memory 5 via the output buffer 5b are provided to a confluence section 6, and the confluence section 6 supplies these data packets to the program storage section 1 in the order of their appearance.

In the data flow type information processing device shown in FIG. 5, data packets are transmitted from program storage unit 1 to data detection unit 2 → arithmetic processing unit 3 – branch unit 4 (→external data memory 5) – confluence unit 6 → As the program storage unit 1 continues to rotate, arithmetic processing based on the program stored in the program storage unit 1 progresses.

FIG. 8 is a diagram showing a schematic configuration of the program storage section 1 shown in FIG. 5. As shown in FIG. In FIG. 8, an input data latch section 11 holds destination information and operand data of data packets, and instruction information is erased. The held operand, new destination information and command information read from the program memory 13 based on the destination information are provided to the output data latch section 14 and latched. The destination information latched by the input data latch section 11 is given to the address calculation section 12, and the address of the program memory 13 is calculated from the destination information.

The program memory 13 stores a data flow program consisting of destination information and command information as shown in FIG.

[Problems to be Solved by the Invention] In the data flow type information processing device described above, an instruction can be executed as soon as the data to be subjected to an operation is prepared,
Programs are executed in parallel and asynchronously following the natural flow of data. Therefore, when a source program written to be executed on a normal von Neumann computer is expanded into a data flow program and executed, the order of execution does not necessarily match the order of execution on a von Neumann computer. If you simply follow the program, you will discover errors in the source program,
The problem was that it could not be corrected.

Further, there are problems in that it is more difficult to discover errors when executing the developed data flow program, and it is also very difficult to know the correspondence with the source program.

Therefore, a main object of the present invention is to provide a debugging device for efficiently debugging a dataflow program executable on a dataflow type information processing device.

[Means for Solving the Problems] A data flow program debugging device according to the present invention includes a program storage means, a program execution means, a state storage means, a data transition storage means, a debug information storage means, and a debug information storage means. It is composed of a display means and a control means, the program storage means stores a data flow program, and the program execution means includes a plurality of functional parts that simulate functions corresponding to a plurality of +13 component parts in the processing device. , and sequentially execute the data flow program stored in the program storage means. The state storage means stores the execution state corresponding to each line in the source program of the plurality of functional parts of the program execution means, and the data transition storage means stores the execution state corresponding to each line in the source program of the plurality of functional parts of the program execution means, and the data transition storage means is a program execution means used to execute the data flow program in each line of the corresponding source program. The debug information storage means stores debug information indicating the correspondence between each line of the source program and each node of the data flow program, and the control means stores the progress status of processing and execution in each functional part of the debug information storage means. The program execution means, state storage means, and data transition storage means are controlled according to the debug information stored in the knob information storage means. The debug information display means displays debugging information regarding each line of the source program in connection with the execution of each node of the data flow program.

[Operation] Since the debugging device according to the present invention creates a file indicating the correspondence between source programs and data flow programs, it is possible to display source programs corresponding to data flow programs that are executed sequentially. In addition, when executing a simulated program, the data flow program is sequentially advanced as each line of the corresponding source program is executed and completed, and the states and data transitions of each component of the data flow information processing device are sequentially changed. The information is stored in the state storage means and the data transition storage means, and as a result, the execution order and execution information of the source program corresponding to the data flow program are obtained.

[Example] FIG. 1 is a schematic block diagram of an embodiment of the present invention.

In FIG. 1, a debug execution device 10 is a part that simulates the internal operation of the data flow information processing device, and executes the operations of each component of the data flow type information processing device. As this debug execution device 10, a general information processing device (Neumann type computer) is used. The debug execution device 10 includes an input file 20, a command file 30, a console 40, and an internal file 50.
, the debug display file 60 , the result file 70 and the display 80 are connected.

The input file 20 is a part for inputting data necessary for causing the debug execution device 10 to execute a data flow program and perform debugging. This input file 20 includes a program file 21 and a debug information file 22. The program file 21 stores object programs of data flow programs and data used therein. Further, the debug information file 22 stores a source program list and information indicating the correspondence between the source program and the object program developed in a data flow type. The command file 30 stores various instructions related to debugging.
stores a group of commands to be executed by batch processing.

The console 40 is used for manually inputting various specifications related to debugging. The internal file 50 stores various information used inside the debug execution device 10, and includes a PS file 51, a data packet file 52, and an EDS file/EDS map file (hereinafter referred to as the EDS file) 53. The PS file 51 stores a data flow program obtained by converting an object program given from the program file 21 into a format executable by the debug execution device 10. The data packets stored in the data packet file 52 correspond to data packets that circulate through each part of the data flow type information processing apparatus shown in FIG. 5 described above.

The EDS file 53 stores information regarding the storage contents and storage area allocation of a portion corresponding to the external data memory 5 shown in FIG.

The debug display file 60 stores changes in the display during debug execution in the debug execution device 10, and the result file 70 stores the results of execution of the data flow graph by the debug execution device 10.

Therefore, the results of the arithmetic processing based on the data flow program are stored in this result file 70. The display 80 displays the contents of the debug display file 60, the contents of the result file 70, and messages.

The object program given from the program file 21 is converted into a data flow program executable by the debug execution device 10 and stored in the PS file 51. Furthermore, data given from the program file 21 is stored in a data packet file 52. The debug processing by the debug execution device 10 shown in FIG. 1 is executed according to commands input from the console 40 or commands stored in the command file 30.

2 is a diagram showing an example of a source program debugged by the debug device shown in FIG. 1, FIG. 3 is a diagram showing a flow graph based on the source program shown in FIG. The figure is a diagram showing an example of a debug execution result according to an embodiment of the present invention.

In the flow graph shown in FIG. 3, the numbers in the circles on the right side indicate execution rank values, and the numbers on the left side indicate node numbers assigned in ascending order of the execution rank values. It is assumed that the programs are executed in the order of their node numbers. Therefore, when the source program shown in FIG. 2 is executed in the order of the node numbers, the display as shown in FIG. 4 will be displayed along with debugging.

First, the display format of FIG. 4 will be explained using the case of node 2 shown at position 100 as an example. In addition,
The information displayed there shows the information when node 2 was executed.

When debugging is executed, as shown in FIG. 4, the executed node number 100. Processing start line 210° Processing line 22
0. Three types of processing end lines 230 are displayed. The processing start line 210 means a line of the source program in which data has come to one of the nodes belonging to one line. Furthermore, the line 220 being executed means a line of the source program in which data has entered all of the input nodes among the nodes belonging to one line. Note that an input node is a node that is directly connected to a node that does not belong to the row, such as a node that belongs to another row. and,
Furthermore, the processing end line 230 means a line of the source program in which all nodes belonging to that line have been executed.

All of the above three types are displayed according to the same format. Here, the format will be explained. The number 10 displayed at the leftmost position 300 indicates the line number in the source program. The position 310 to the right shows the content of the source program written on the 10th line, and the position 320 to the right lists all variables existing in that line, each (line number of output light) : (variable name).

Note that if the variable name is in parentheses, it indicates that the value of that variable has been determined, and 5source' is placed in position 320.
"" means the source of input to the program, etc. Next, 330 below indicates the line number in which the next instruction to be processed for each variable is written, and since there are multiple instructions, they are separated by symbols ",". The three types of lines 210, 220, and 230 mentioned above are all displayed in the above format.

According to the format described above, a process end line 210, an ongoing process start line 220 . Information regarding each line displayed in each of the process end lines 230 changes as shown in FIG. 4 as debugging is executed. for example,
Looking at line 10, first, by executing node 2 shown at position 100, information about line 10 is displayed in process start line 210, and then after execution of node 6, information about line 10 is displayed in line 222 being executed. is moved, and further moved to the process end line 233 by execution of node 10 and displayed. The display is updated as above for all lines, and debugging is executed.

Next, we will explain in detail how the display changes as each node is executed. As shown in FIG. 3, data is sent to node 7 by execution of node 2, so row 10 to which the node belongs is displayed in process start row 210. At this time, all variables belonging to that line have been determined, and any variables displayed at position 320 are displayed within parentheses.

Next, when node 5 is executed, data with the determined value of the variable is sent to node 11 belonging to the 12th line, so that 12th line is displayed in the process start line 211. At this time, "X", which is a variable belonging to line 12 and is displayed at position 420, is displayed without being placed within parentheses. This is because the value has not yet been determined because the 10 lines that output the value have not yet been completed.

Furthermore, by executing node 6, data is sent to all input nodes belonging to 10 rows, so those 10 rows are displayed in row 222 that is being executed.

Note that, as described above, the input node means a node that has a node that does not belong to the relevant row as an output light node, and in the case of these 10 rows, nodes 7 and 8 are shown. At this point, the output light node 2゜4 of the input two-door and the output light nodes 5 and 6 of the human-powered node 8 have all been executed, so data is coming to all of the human-powered two-door nodes 8. .

Furthermore, when the program is executed and node 10 is executed, for those 10 lines, all nodes 7.8.10 belonging to that line have been executed, and 1
Line 0 is displayed in the process end line 233. Also, as a result, the value of the variable “X” is determined, so the line 223 that is being executed
Each position 42 in line 11 and 12 displayed in
The variable “xl” displayed at 2,520 is displayed in parentheses.

As described above, each piece of information displayed changes as the program is executed. Eventually node 16
The execution of the formula on the 14th line, which is the last line of the source program, is completed, and the 14th line is displayed in the processing end line 234, and the data flow program given as an example ends. In addition, the variable "out" output by line 14
Since the destination is the output of the program, the destination line number 630 is “s in lc”.
(meaning of sink) is displayed.

Debugging is executed as described above, and the execution order and execution information regarding the source program executed by the data flow type information processing device are obtained.

[Effects of the Invention] As described above, according to the present invention, it is possible to obtain execution information of a source program represented by the execution order and the number of programs corresponding to the execution of a data flow program. development can be carried out easily and accurately.

[Brief explanation of the drawing]

FIG. 1 is a schematic block diagram of an embodiment of the present invention. FIG. 2 is a diagram showing an example of a source program applied to an embodiment of the present invention. FIG. 3 is a diagram showing an example in which the source program shown in FIG. 2 is developed into a data flow program. FIG. 4 is a diagram showing the results of debugging according to an embodiment of the present invention. FIG. 5 is a block diagram showing an example of the configuration of a conventional data flow type information processing device. FIG. 6 is a diagram showing the structure of a data packet circulating through each part of the information processing apparatus shown in FIG. 5. FIG. 7 is a diagram showing a portion of the data flow program stored in the program memory. FIG. 8 is a block diagram showing the configuration of the program storage section of the data flow type information processing device. In the figure, 10 is a debug execution device, 20 is an input file, 21 is a program file, 22 is a debug information file, 30 is a command file, 40 is a console, 50 is an internal file, 51 is a PS file, 52 is a data packet file, 53 is an EDS-EDS map file, 60 is a debug display file, 70 is a result file, and 80 is a display.

Claims (1)

[Scope of Claim] A data flow program debugging device for debugging the operation of a data flow program in units of lines written in a source program, comprising: a program storage means for storing the data flow program; The data flow program, which includes a plurality of functional parts that simulate functions corresponding to a plurality of component parts in a processing device to be executed, and which is stored in the program storage means, is sequentially executed for each line of the corresponding source program. a program execution means to execute; a state storage means for storing execution states of the plurality of functional parts of the program execution means corresponding to each line of the source program; execution of the data flow program in each corresponding line of the source program; a data transition storage means for storing the progress of processing and execution of data used in each functional unit of the program execution means; a debugging means showing the correspondence between each line of the source program and each node of the data flow program; debug information storage means for storing information; debug information display means for displaying debug information regarding each line of the source program in connection with execution of each node of the data flow program; and the debug information stored in the debug information storage means. A debugging device for a data flow program, comprising control means for controlling the program execution means, the state storage means, and the data transition storage means according to information.