JP6212823B2 - Parallelization support device, execution device, control system, parallelization support method, and program - Google Patents

Description

  The present invention relates to a parallelization support device, an execution device, a control system, a parallelization support method, and a program.

  In a control system used in a plant facility such as a gas turbine, for example, various state quantities are acquired from a measuring instrument provided in the plant, and the plant can be quickly controlled so that the plant can be appropriately controlled according to changes indicated by the state quantities. Is required to respond. Therefore, a real-time OS that can complete the processing within a desired time is generally adopted as an OS (Operating System) of the control system. Further, in a computer in such a control system, a large amount of time is required for a constant cycle calculation that repeats a series of cycles of inputting a signal necessary for control to control a controlled object and performing a calculation in accordance with the signal in a short cycle. Often spent.

  By the way, in recent years, a general-purpose PC server or the like is also equipped with a multi-core CPU (Central Processing Unit), and it is possible to increase the speed by performing parallel processing using the multi-core even in the above-described periodic operation. However, it is generally known that even if an application that does not support multi-core is executed on a computer equipped with a multi-core CPU, the effect of parallelization is hardly obtained. In order to take advantage of multi-core capabilities, programmers need to rewrite their application programs to support multi-cores using specialized knowledge. For example, Patent Document 1 describes a development method in which a definition is assigned to a core for each function in a program, and a programmer approaches the expected performance while executing the program that is read and built.

JP 2011-134055 A

  In general, application engineers are familiar with application specifications but are not familiar with programming knowledge. Conversely, programmers are often not familiar with application specifications even though they can program. Therefore, the application engineer can easily determine which of the processes by the program that constitutes the application can be executed in parallel. However, since there was no technology to parallelize the application, the knowledge of the multi-core can be obtained using that knowledge. Could not be pulled out.

  Accordingly, an object of the present invention is to provide a parallelization support device, an execution device, a control system, a parallelization support method, and a program that can solve the above-described problems.

The first aspect of the present invention displays the execution order and dependency among a plurality of logic sheets for a logic sheet that is a predetermined small unit program corresponding to a group of processes that need to be sequentially executed. Logic sheet-related output unit and a logic sheet aggregation receiving unit that receives an operation of assigning at least a part of the plurality of logic sheets to a program of an execution unit that is required to be completed within a requested time after the execution is started A core selection receiving unit that receives a selection of a plurality of cores to execute the execution unit program configured to include the assigned logic sheet, and assigns the execution unit program to the cores. A core selection support information display unit for displaying an index for Which is a parallel support device.

In the second aspect of the present invention, the index is a load factor of the core when the execution unit program is executed so as to be completed within the requested time in the core that has received the selection. It is characterized by that.

The index according to the third aspect of the present invention is a time required for executing the program when only the program of the execution unit is executed in the core that has received the selection.

The logic sheet aggregation receiving unit according to the fourth aspect of the present invention receives an operation of moving one or more of the logic sheets included in one execution unit program to another execution unit program. It is characterized by that.

According to a fifth aspect of the present invention, for the program to be executed, the identifier of the core received by the core selection receiving unit in the parallelization support apparatus is acquired, and the program of the execution unit is acquired in the core indicated by the acquired identifier. An application execution apparatus characterized by executing the application.

According to a sixth aspect of the present invention, there is provided a plant control system characterized in that the apparatus in the plant is controlled by the execution device according to the fifth aspect.

According to a seventh aspect of the present invention, for a logic sheet that is a predetermined small unit program corresponding to a group of processes that need to be executed sequentially, the execution order and dependency relationships between the plurality of logic sheets are displayed. And accepting an operation of assigning at least a part of the plurality of logic sheets to a program of an execution unit that is requested to be completed within a requested time from the start of execution, and including the assigned logic sheet Supporting parallelization of an application, wherein selection of which of the plurality of cores executes the configured execution unit program is received, and an index for assigning the execution unit program to the core is displayed. Is the method.

According to an eighth aspect of the present invention, a computer of an application parallelization support apparatus performs a plurality of the logics on a logic sheet that is a predetermined small unit program corresponding to a group of processes that need to be sequentially executed. Means for displaying the order of execution and dependency between sheets, and an operation of assigning at least a part of the plurality of logic sheets to a program of an execution unit that is required to be completed within a requested time from the start of execution. Means for accepting, means for accepting selection of which of the plurality of cores execute the execution unit program configured to include the assigned logic sheet, and an index for assigning the execution unit program to the core It is a program characterized by functioning as a means for displaying .

  According to the present invention, it becomes possible for an application engineer to draw out multi-core capabilities using his / her own knowledge.

It is the schematic of the control system containing the parallelization assistance apparatus and execution apparatus by 1st embodiment of this invention. It is explanatory drawing of the control logic arithmetic function in 1st embodiment of this invention. It is a functional block diagram of the parallelization assistance apparatus by 1st embodiment of this invention. It is a 1st figure which shows an example of the operation screen of the parallelization assistance apparatus by 1st embodiment of this invention. It is a 2nd figure which shows an example of the operation screen of the parallelization assistance apparatus by 1st embodiment of this invention. It is a figure which shows the processing flow of the parallelization assistance apparatus by 1st embodiment of this invention. It is a 3rd figure which shows an example of the operation screen of the parallelization assistance apparatus by 1st embodiment of this invention. It is a 4th figure which shows an example of the operation screen of the parallelization assistance apparatus by 1st embodiment of this invention.

<First embodiment>
An application parallelization support apparatus and execution apparatus according to a first embodiment of the present invention will be described below with reference to FIGS.
FIG. 1 is a schematic diagram of a control system including an application parallelization support apparatus and an execution apparatus according to the first embodiment.
As shown in FIG. 1, the control system of the present embodiment includes a parallelization support device 10, an execution device 20, and a control target device 30.
The parallelization support apparatus 10 includes a computer, and an application program for a design tool used by a user is installed and operated.
The execution device 20 includes a computer, and a real-time OS (hereinafter referred to as RTOS) and a control application program are installed and operating. The computer of the execution device 20 includes a multi-core CPU. The execution device 20 is a controller in the control system and controls the control target device 30. Hereinafter, an application program is abbreviated as an application.
The control target device 30 includes a control target 31 such as a gas turbine to be controlled and a control device 32 such as a sensor or an operation end.

  In this control system, with which design tool of the parallelization support device 10 the parallelization support device 10 executes various programs constituting the control application on which of a plurality of CPU cores provided in the computer of the execution device 20 Is specified by the user. Reference numeral 100 indicates an example of an operation screen of a design tool operated by a user. Further, the execution device 20 executes various programs constituting the control application by the designated core, and controls the behavior of the control target device 30.

  Here, the terms used below will be described with reference to FIG. First, the control application executed by the execution device 20 is referred to as a control logic operation function application. The control logic operation function application is composed of programs called a plurality of function operation groups. The program indicated by the functional operation group is a program module of an execution unit, and is executed as a task such as an RTOS, that is, a thread or a process. A program indicated by each functional calculation group is prioritized based on a calculation cycle or the like imposed on the processing of the functional calculation group, is scheduled by the scheduling function of the RTOS, and is executed by the CPU of the execution device 20.

Each functional calculation group is composed of a smaller unit program called a plurality of logic sheets. The logic sheet is a program corresponding to a group of processes that need to be sequentially executed, and includes a plurality of functional operation blocks and data flow information indicating their processing order and input / output relationships. In FIG. 2, the data flow is represented by arrows. Here, the functional operation block is a unit of operations for performing predetermined real number operations, integer operations, and logical operations. The programs indicated by the logic sheets in each functional calculation group are executed in the order in which they are registered or set, and the functional calculation blocks in each logic sheet are executed automatically in the order assigned or set. Note that the processing order indicated by each logic sheet is determined in the order of execution, and the processing within the functional calculation group is executed sequentially.
In the following, the user is used in the sense of an application engineer who is familiar with the specifications and requirements of the control system. The programmer is used to mean a person who can perform coding but does not have deep knowledge of a control method (control logic) of a control target. The programmer creates the program indicated by the above functional calculation group and logic sheet.

FIG. 3 is a functional block diagram of the parallelization support apparatus according to the first embodiment of the present invention.
The parallelization support apparatus 10 in this embodiment is demonstrated using FIG.
As illustrated in FIG. 3, the parallelization support apparatus 10 includes a core selection support information display unit 101, a core selection reception unit 102, a selected core recording unit 103, a functional calculation group output unit 104, a logic sheet related output unit 105, and logic sheet aggregation. A reception unit 106, a core information output unit 107, an operation detection unit 108, and a storage unit 200 are provided.

  The core selection support information display unit 101 outputs information that serves as an index for helping the user to determine which functional calculation group program is executed on which core. The information serving as an index is, for example, the time required to execute the program of the functional calculation group, the CPU load factor for each CPU core, and the like. In a control system such as a plant, it is often required to complete each process within a predetermined time because a control delay may cause a control error due to a slight process delay, resulting in an unexpected accident. Therefore, it is desirable to utilize the capability of a multi-core CPU and distribute the load by assigning and executing the program of each functional calculation group to each core so that each process can be executed without delay. By referring to the information output by the core selection support information display unit 101, the user can complete the processing of each functional calculation group within a desired time, whether the load is smoothed and allocated to each core, etc. Can be confirmed.

The core selection accepting unit 102 accepts input of information as a result of the user selecting which functional operation group the program of which core is to be executed on.
Based on the information received by the core selection receiving unit 102, the selected core recording unit 103 records the identifier of each functional operation group and the core identifier in the storage unit 200 in association with each other.
The functional calculation group output unit 104 outputs a list of functional calculation groups and logic sheets to be assigned to the core. The user selects a functional calculation group from the list output by the functional calculation group output unit 104 and assigns a core for executing the program of the functional calculation group.

  The logic sheet relation output unit 105 outputs a relation between logic sheets. The relationship between logic sheets refers to the execution order and dependency between the processes of “logic sheet 1” and “logic sheet 2”. For example, the programs of “logic sheet 1” and “logic sheet 2” must be sequentially executed in the order of “logic sheet 1” and “logic sheet 2”, or “logic sheet 1” and “logic sheet 2”. "And the process of" logic sheet 3 "can be executed in parallel.

  The logic sheet aggregation receiving unit 106 receives an operation of assigning a logic sheet by a user to a functional calculation group that is an execution program. Specifically, when the user adds a new logic sheet to the functional calculation group or moves a certain logic sheet to another functional calculation group, the change input is accepted. For example, suppose that the program of “functional calculation group 1” is composed of “logic sheet 1” to “logic sheet 10”, and the program of “functional calculation group 1” is processed by occupying one core. Consider a case where the processing cannot be completed within the processing time required by the control system. In such a case, it is assumed that the program of “functional operation group 1” can be divided into “logic sheet 1” to “logic sheet 5” and “logic sheet 6” to “logic sheet 10”. Then, the user can divide “functional calculation group 1” into, for example, “functional calculation group 1A” and “functional calculation group 1B” using the function of the logic sheet aggregation receiving unit 106. Further, the user assigns “logic sheet 1” to “logic sheet 5” to the program of “functional computation group 1A” and assigns “logic sheet 6” to “logic sheet 10” to the program of “functional computation group 1B”. Can do. Note that an operation in which a user assigns a logic sheet to one functional calculation group is referred to as “aggregation”. The logic sheet aggregation receiving unit 106 records the aggregation result in the storage unit 200.

The core information output unit 107 outputs the correspondence relationship between the CPU cores assigned by the user and the functional calculation groups.
The operation detection unit 108 detects an input operation or the like by a user using a mouse or a keyboard, and acquires the operation instruction information.
The storage unit 200 stores the CPU load factor and the time required for processing for each function calculation group for each execution device 20. The CPU load factor is a value indicating how much of the CPU's own processing capacity is spent executing the program when the program of the functional calculation group is executed so that the execution is completed within the time required by the control system. It is. This value may be an average value of actually measured values of CPU load factors when a function calculation group program is executed while actually occupying one core, or may be a value calculated based on the complexity of processing. The time required for processing is the time required from the start of execution to the completion when the program of the functional calculation group occupies one core and is executed.

FIG. 4 is a first diagram illustrating an example of an operation screen of the parallelization support apparatus according to the present embodiment.
An operation screen used when a user assigns a CPU core to a function calculation group will be described with reference to FIG.
It is assumed that the CPU of the execution device 20 includes two cores. Further, it is assumed that the user selects “control logic operation function 1” as an application including a program to be assigned a core by a predetermined operation.
In the area 41 of FIG. 4A, a list of function calculation groups and logic sheets of “control logic calculation function 1” output by the function calculation group output unit 104 is displayed.
In the area 42 of FIG. 4A, a functional calculation group assigned to “Core 1” which is one of the two CPU cores is displayed. Reference numeral 43 in FIG. 4A indicates a functional calculation group selected by the user by a predetermined operation. The predetermined operation may be an operation in which, for example, when a certain functional calculation group is assigned to “core 1”, the functional calculation group is selected in the area 41 and dragged and dropped to the area 42. Reference numeral 44 in FIG. 4A indicates the CPU load factor when the program of “functional operation group 1” is assigned to “core 1” and executed. Reference numeral 45 in FIG. 4A indicates the CPU load factor of all the functional calculation groups assigned to “Core 1”. These CPU load factors are information output by the core selection support information display unit 101. Similarly, in the area 46 of FIG. 4A, the function calculation group assigned to “Core 2” and the CPU load factor when each program is executed are displayed.

In the case of the example of FIG. 4A, “functional calculation group 1” is assigned to “core 1”, and “functional calculation group 2” is assigned to “core 2” which is the remaining one of the two cores. “Functional calculation group 3” and “functional calculation group 4” are assigned. Since the CPU load factor of “function calculation group 1” is 50% (reference numeral 44), the total CPU load factor of “core 1” is 50%. On the other hand, the CPU load factor of “Function calculation group 2” is 40% (reference numeral 47), the CPU load factor of “function calculation group 3” is 50% (reference numeral 48), and the CPU load factor of “function calculation group 4” is Since it is 20% (reference numeral 49), the total CPU load factor of “core 2” is 110% (reference numeral 50).
The user sees this screen and the CPU load factor of “Core 2” exceeds 100%. Therefore, “Function calculation group 2”, “Function calculation group 3”, “Function calculation group” assigned to “Core 2”. The program “4” can grasp that the execution cannot be completed within the requested processing time. Further, the user can reassign, for example, “functional calculation group 4” from “core 2” to “core 1” by a predetermined operation. The predetermined operation may be an operation of selecting “Functional calculation group 4” in the area 46 and dragging and dropping the area 42 in the case of FIG.

FIG. 4B is a screen as a result of the user changing the assignment of the functional calculation group to the core in this way. In the screen shown in FIG. 4B, “functional calculation group 1” and “functional calculation group 4” are assigned to “core 1”, and “functional calculation group 2” and “functional calculation group 3” are assigned to “core 2”. Assigned. By assigning in this way, the total CPU load factor of “Core 1” is 70% (reference numeral 45), and the total load factor of the CPU of “Core 2” is 90% (reference numeral 50), which is kept within 100%. Can do.
Note that the load factor of 100% is an example of the upper limit of program allocation, and the criterion may be, for example, the upper limit of the CPU load factor of each core being within 60%.
Reference numeral 40 indicates an OK button. When the operation detection unit 108 detects that the user presses the OK button 40, the selected core recording unit 103 determines whether or not the total load factor of each core exceeds a predetermined threshold (such as 100%). If not, the selected core recording unit 103 records the identifier of each functional calculation group and the assigned core identifier in the storage unit 200 in association with each other. If the total load factor of each core exceeds a predetermined threshold, the core information output unit 107 may output an error message.

FIG. 5 is a second diagram illustrating an example of an operation screen of the parallelization support apparatus according to the present embodiment.
Another example of information serving as an index output by the core selection support information display unit 101 will be described with reference to FIG.
FIG. 5 is a diagram showing only regions 42 and 46 in FIG. FIG. 5A shows an example in which the core selection support information display unit 101 outputs the required time and the requested processing time for each function calculation group, not the CPU load factor. The required time is the time required to complete the execution when the functional operation group is assigned to the CPU core designated by the user and executed. This value is, for example, a value calculated from the “time required for processing” stored in the storage unit 200 and the total CPU load factor. For example, it may be calculated that the required time when the “time required for processing” of the “functional calculation group 1” is 50 msec and the total CPU load factor is 200% is 100 mec. The requested processing time is an upper limit value of the processing time required for executing the program indicated by the functional calculation group. The core selection support information display unit 101 may output a control cycle indicating a cycle for executing the program of the functional calculation group instead of the request processing time.
In the case of this figure, the required time of “functional calculation group 1” and “functional calculation group 4” assigned to “core 1” is both within the required processing time (reference numerals 51 and 52). For example, the request processing time of “Functional calculation group 1” is 15 msec, while the required time is 10 msec. However, the required time of “functional calculation group 2” and “functional calculation group 3” assigned to “core 2” does not fall within the required processing time (reference numerals 53 and 54).
Looking at this display, the user should understand that the processing of “functional calculation group 2” and “functional calculation group 3” cannot be completed within the requested processing time, and consider how to allocate the functional calculation group to the CPU core. Can do.

  FIG. 5B shows another example in which the core selection support information display unit 101 outputs information serving as an index. In this example, the state of the load on each core is represented by “high”, “medium”, and “low” for each functional calculation group (reference numerals 55, 56, 58, 59). The total load as a result of assigning the functional operation group to each core is also displayed as “high”, “medium”, and “low” (reference numerals 57 and 60). The total load display is replaced with a numerical value such as “20%” if the load in one function calculation group is “low”, “40%” if “medium”, “60%” if high, etc. If the total load value is less than “40%”, “low” may be displayed, and if “40%” or more and less than “60%”, “medium” may be displayed. If the total load exceeds a predetermined maximum load (for example, 100%), “excessive” may be displayed. Further, instead of “high”, “medium”, and “low”, display may be performed according to colors such as “red”, “yellow”, and “blue” according to the load state.

FIG. 6 is a diagram showing a processing flow of the parallelization support apparatus according to the first embodiment.
Processing for setting a core for executing a functional operation group in the parallelization support apparatus will be described using the processing flow of FIG.
As a premise, it is assumed that the control logic operation function application and the execution device 20 that executes the application are designated by a predetermined operation instructing the user to start the setting.
First, the function calculation group output unit 104 reads information on the function calculation group and the logic sheet included in the control logic calculation function application from the storage unit 200 using the identifier of the control logic calculation function application designated in advance. Then, the function calculation group output unit 104 outputs the list information of the function calculation groups as indicated by the area 41 in FIG. 4 (step S1).
In addition, the core information output unit 107 acquires the number of CPU cores included in the execution device 20 from the storage unit 200 using the specified identifier of the execution device 20. Then, the core information output unit 107 displays, as shown by the areas 42 and 46 in FIG. Further, the core information output unit 107 reads the relationship between the functional calculation group designated by the user and the CPU core that executes the program from the storage unit 200, and outputs the correspondence relationship between the CPU core and the functional calculation group (step S2). For example, as shown in FIGS. 4 to 5, a list of functional operation groups assigned to “Core 1” is output to the area 42.

Next, the operation detection unit 108 detects the user's operation for assigning the functional calculation group to the core by the user (step S3). Then, the operation detection unit 108 outputs a signal indicating that an allocation operation has been performed to the core selection support information display unit 101 and the core selection reception unit 102. Then, the core selection receiving unit 102 stores the identifier of the execution device 20, the identifier of the CPU core assigned by the user through a drag-and-drop operation, and the identifier of the functional operation group in a memory or the like.
Next, the core selection support information display unit 101 reads the information stored in the core selection reception unit 102 from the memory, and uses the identifier of the function calculation group and the identifier of the execution device 20 to store the function calculation group from the storage unit 200. Information serving as an index such as CPU load and required time when the program is executed by the execution device 20 is read. Then, the core selection support information display unit 101 outputs the CPU load and the like as indicated by reference numeral 44 in FIG. Further, the core selection support information display unit 101 calculates the total CPU load factor for each core and outputs it as shown by reference numeral 45 in FIG. 4 (step S4).

Next, the operation detecting unit 108 detects whether or not the assignment of the function calculation group to the core by the user is completed (step S5). Specifically, when the operation detection unit 108 detects that the user presses the OK button 40, it may be determined that the assignment has been completed. When the operation detection unit 108 does not detect the completion of assignment (step S5 = No), the processing from step S3 is repeated.
When the operation detection unit 108 detects the completion of assignment (step S5 = Yes), the selected core recording unit 103 reads the information stored in the memory by the core selection receiving unit 102, and the identifier of the execution device 20 and the CPU specified by the user The identifier of the core and the identifier of the functional calculation group are associated and recorded in the storage unit 200 (step S6).
This processing flow is completed.

  And when the program of the functional calculation group to which the user has assigned the core in this way is executed in the execution device 20, the task control program of the execution device 20 designates the program indicated by each functional calculation group by the user. Allocate CPU cores and execute. The execution device 20 acquires the information on the correspondence relationship between the functional calculation group and the CPU core stored in the storage unit 200 via a network, for example.

  According to the present embodiment, the user can select which of the multi-cores to execute the program constituting the application. Further, the user can grasp the influence (CPU load and processing time) when the program is assigned to each core. As a result, even a user who does not have a programming technique for dealing with multi-cores can draw out the capabilities of the multi-core CPU by using his / her knowledge of application specifications.

FIG. 7 is a third diagram illustrating an example of an operation screen of the parallelization support apparatus according to the present embodiment.
Another example of the functional calculation group list output by the functional calculation group output unit 104 will be described with reference to FIG.
Reference numeral 61 in FIG. 7A is a relationship display button. When the operation detection unit 108 detects that the user has pressed the relationship display button 61, the logic sheet relationship output unit 105 outputs the information shown in FIG. 7B. In the case of the example in FIG. 7A, since “functional calculation group 1” is assigned to “core 1” and the CPU load factor when the program of “functional calculation group 1” is executed is 110%, The total CPU load factor of “1” is also 110%, exceeding 100%. Therefore, the user examines whether or not “functional operation group 1” can be divided with reference to the diagram showing the relationship between the logic sheets shown in FIG.

FIG. 7B is a diagram illustrating a relationship between logic sheets output by the logic sheet relationship output unit 105. This diagram is an example of a diagram that displays the dependency relationships of “logic sheet 1-1”, “logic sheet 1-2”, and “logic sheet 1-3” that make up the functional operation group 1. FIG. In this figure, the processing of “logic sheet 1-1” and the processing of “logic sheet 1-2” need to be executed sequentially, and the processing of “logic sheet 1-3” is independent of these processing. Indicates that it can be executed.
For example, the user divides “functional calculation group 1” into two programs “functional calculation group 1A” and “functional calculation group 1B” from the information shown in FIG. 2 ”and“ Logic Sheet 1-3 ”are consolidated into separate functional operation groups, and further consideration is given to assigning each program of“ functional operation group 1A ”and“ functional operation group 1B ”to separate CPU cores. You can also
Information indicating the relationship between the logic sheets is stored in the storage unit 200 in advance, and the logic sheet relationship output unit 105 reads the information from the storage unit 200 and indicates the relationship between the logic sheets by a predetermined description method. May be output. Information indicating the relationship between logic sheets output by the logic sheet relationship output unit 105 may be displayed in a pop-up, for example. Further, in addition to the relationship between the logic sheets, the logic sheet relationship output unit 105 may output information on the processing weight such as a CPU load factor for each logic sheet.

FIG. 8 is a fourth diagram illustrating an example of an operation screen of the parallelization support apparatus according to the present embodiment.
FIG. 8 shows that “functional calculation group 1” is divided into “functional calculation group 1A” and “functional calculation group 1B” by a predetermined operation from the state of FIG. 1 ”and“ logic sheet 1-2 ”are examples of an operation screen as a result of integrating“ logic sheet 1-3 ”into“ functional operation group 1B ”. FIG. 8 shows that, as a result of collecting the logic sheets, the CPU load factor when executing the program of “Functional calculation group 1A” is 95%, and the CPU load factor when executing the program of “Functional calculation group 1B” is 15%. Furthermore, by assigning the program of “Functional calculation group 1B” to “Core 2”, the total CPU load factor of “Core 1” is 95% (reference numeral 45), and the total CPU load factor of “Core 2” is It shows that it became 75% (symbol 50).

  According to the present embodiment, the user can adjust the parallelization by changing the number of aggregation of logic sheets in each functional calculation group. As a result, program assignment to each core can be smoothed more flexibly, the utilization efficiency of the core can be improved, and the processing speed can be increased.

By the way, since the programs of “functional calculation group 1A” and “functional calculation group 1B” assigned as shown in FIG. 8 are executed by different cores, the execution timings are different unless a mechanism for achieving synchronization is used. For example, if the program of “functional calculation group 1A” and the program of “functional calculation group 1B” are assigned to different cores and executed, the process of “logic sheet 1-2” is executed twice at a constant cycle. During this time, the process of “logic sheet 1-3” may be executed only once. Furthermore, only the execution result of the subsequent “logic sheet 1-4” using the execution results of the “logic sheet 1-2” program and the “logic sheet 1-3” program and the execution result of the “logic sheet 1-2” program are used. It is assumed that the logic sheet 1-5 ”exists. In this case, the process of the “logic sheet 1-4” is started after the execution of the program of the “logic sheet 1-3” is completed, whereas the program of the “logic sheet 1-5” As soon as the program of “logic sheet 1-2” is completed, it is executed. At this time, it may be difficult for a programmer to determine whether the influence of the timing difference between the execution of the processes of “logic sheet 1-4” and “logic sheet 1-5” is acceptable in controlling the plant. Many. If the user determines that the deviation is unacceptable for controlling the plant, the user may consolidate the “logic sheet 1-2” and the “logic sheet 1-3” into, for example, the “functional operation group 1A”. it can. Here, it is assumed that the processing of “logic sheet 1-2” and “logic sheet 1-3” is executed once each time the program of “functional operation group 1A” is executed once. Then, if the user aggregates into “functional calculation group 1A” as described above and executes the program of “functional calculation group 1A”, the processes of “logic sheet 1-2” and “logic sheet 1-3” are Since the process is executed the same number of times, the above deviation does not occur. When the sum of the CPU load factors exceeds 100% as a result of the aggregation of the logic sheets as described above, the user executes the program of the “functional operation group 1A” on the execution device 20 on which another CPU with good performance is mounted. You can also consider doing it.
According to the present embodiment, even in such a case, the user appropriately aggregates the logic sheets based on information output from the logic sheet related output unit 105 and knowledge about the control system that the user has, and forms one functional operation group. CPU resources can be used more effectively by assigning them to different cores.

The division of the functional calculation group and the aggregation of the logic sheets from FIG. 7 to FIG. 8 are realized by the following processing, for example.
When the operation detection unit 108 detects an operation of a function calculation group creation instruction by the user, the operation detection unit 108 outputs the instruction signal to the logic sheet aggregation reception unit 106. The logic sheet aggregation receiving unit 106 generates a correspondence relationship between each functional calculation group to which the added functional calculation group is added and the logic sheet (in this case, a logic sheet constituting the newly generated functional calculation group does not yet exist). And stored in the storage unit 200. On the other hand, the functional calculation group output unit 104 reads out the list information indicating the correspondence relationship between the newly generated functional calculation group and the logic sheet from the storage unit 200 and outputs the list information, for example, as indicated by an area 41 in FIG.
Next, when the operation detection unit 108 detects the logic sheet aggregation operation by the user, the operation detection unit 108 outputs the instruction signal to the logic sheet aggregation reception unit 106. Then, the logic sheet aggregation receiving unit 106 associates the identifier of “functional operation group 1A” with the identifiers of “logic sheet 1-1” and “logic sheet 1-2” and stores them in the storage unit 200. Further, the logic sheet aggregation receiving unit 106 stores the identifier of “functional calculation group 1B” and the identifier of “logic sheet 1-3” in the storage unit 200 in association with each other. Then, the functional calculation group output unit 104 outputs list information indicating the correspondence between the newly generated functional calculation group and the logic sheet from the storage unit 200 to the reading area 41. Further, the core information output unit 107 reads the correspondence between the CPU core and the function calculation group from the storage unit 200 and outputs the same as shown in, for example, the areas 42 and 46 in FIG.

  The parallelization support apparatus described above has a computer inside. Each process of the parallelization support apparatus described above is stored in a computer-readable recording medium in the form of a program, and the above process is performed by the computer reading and executing this program. Here, the computer-readable recording medium means a magnetic disk, a magneto-optical disk, a CD-ROM, a DVD-ROM, a semiconductor memory, or the like. Alternatively, the computer program may be distributed to the computer via a communication line, and the computer that has received the distribution may execute the program.

  The program may be for realizing a part of the functions described above. Furthermore, what can implement | achieve the function mentioned above in combination with the program already recorded on the computer system, what is called a difference file (difference program) may be sufficient.

  In addition, it is possible to appropriately replace the components in the above-described embodiments with known components without departing from the spirit of the present invention. The technical scope of the present invention is not limited to the above-described embodiment, and various modifications can be made without departing from the spirit of the present invention.

DESCRIPTION OF SYMBOLS 10 ... Parallelization assistance apparatus 20 ... Execution apparatus 30 ... Control object apparatus 31 ... Control object 32 ... Control apparatus 101 ... Core selection assistance information display part 102 ... Core selection reception Unit 103 ... selected core recording unit 104 ... function calculation group output unit 105 ... logic sheet related output unit 106 ... logic sheet aggregation receiving unit 107 ... core information output unit 108 ... operation detection Part 200 ... Storage part

Claims (8)

For a logic sheet that is a predetermined small unit program corresponding to a group of processes that need to be sequentially executed, a logic sheet relation output unit that displays an execution order and dependency among a plurality of the logic sheets;
A logic sheet aggregation receiving unit that receives an operation of assigning at least a part of the plurality of logic sheets to a program of an execution unit that is requested to be completed within a requested time after starting execution;
A core selection receiving unit that receives selection of which of the plurality of cores to execute the program of the execution unit configured to include the assigned logic sheet;
A core selection support information display unit for displaying an index for allocating the program of the execution unit to the core;
An apparatus for supporting parallelization of an application, comprising: The index is a load factor of the core when the execution unit program is executed to be completed within the requested time in the core that has received the selection. The application parallelization support apparatus described. 3. The parallel application according to claim 1, wherein the index is a time required for executing the program when only the program of the execution unit is executed in the core that has received the selection. 4. Support device. The logic sheet aggregation receiving unit receives an operation of moving one or a plurality of logic sheets included in one execution unit program to another execution unit program,
The application parallelization support apparatus according to any one of claims 1 to 3 . The program to be executed acquires an identifier of a core received by the core selection receiving unit in the application parallelization support device according to any one of claims 1 to 4 , and the core indicated by the acquired identifier An application execution apparatus that executes an execution unit program. 6. A plant control system, wherein an apparatus in the plant is controlled by the application execution apparatus according to claim 5 . For logic sheets that are predetermined small unit programs corresponding to a group of processes that need to be executed sequentially, display the execution order and dependency among the plurality of logic sheets,
Accepting an operation of assigning at least a part of the plurality of logic sheets to a program of an execution unit that is required to be completed within a requested time from the start of execution,
Accepting selection of which of the plurality of cores to execute the program of the execution unit configured to include the assigned logic sheet;
An index for allocating the execution unit program to the core is displayed. Computer for application parallelization support device,
Means for displaying execution order and dependency among a plurality of logic sheets for a logic sheet that is a predetermined small unit program corresponding to a group of processes that need to be sequentially executed ;
Means for accepting an operation of assigning at least a part of the plurality of logic sheets to a program of an execution unit that is required to be completed within a requested time from the start of execution;
Means for accepting selection of which of a plurality of cores to execute the program of the execution unit configured to include the assigned logic sheet;
Means for displaying an index for allocating the execution unit program to the core;
A program characterized by functioning as

Images