CN118672743A - Deterministic task scheduling method, system and readable storage medium - Google Patents

Abstract

The present invention provides a deterministic task scheduling method, system and readable storage medium, including obtaining processor information and task information, wherein the task information at least includes task cycle, inter-task dependency, inter-task communication limit value and task worst execution time; based on the processor information and the task information, statically scheduling multiple tasks according to the scheduling algorithm to obtain static scheduling results, and obtaining standard running start times corresponding to the multiple tasks based on the static scheduling results; based on the static scheduling results, the standard running start time, the unified time reference point and the task cycle, regularly waking up and scheduling multiple processors to process corresponding tasks. The present invention can solve the problem of uncertainty in traditional task scheduling methods, and has positive significance for the performance and reliability of real-time systems.

Description

Deterministic task scheduling method, system and readable storage medium

Technical Field

The present invention relates to the technical field of task scheduling, and in particular to a deterministic task scheduling method, system and readable storage medium.

Background Art

Task scheduling refers to the process in which the system executes tasks at a specific agreed time in order to automatically complete specific tasks. Task scheduling can free up more manpower and allow the system to automatically execute tasks, saving manpower.

In real-time systems, task scheduling plays a vital role in the reliability and responsiveness of the system. Different task scheduling strategies will affect the performance and reliability of tasks accordingly. Traditional task scheduling methods often have certain uncertainties. For example, the execution time of tasks cannot be guaranteed, which will affect the stability and responsiveness of the system.

Summary of the invention

In response to the above technical problems, the present invention provides a deterministic task scheduling method, system and readable storage medium, which can solve the problem of uncertainty in traditional task scheduling methods.

A first aspect of the present invention provides a deterministic task scheduling method, comprising:

Acquire processor information and task information, wherein the task information at least includes task cycle, inter-task dependency, inter-task communication limit value and task worst execution time;

Based on the processor information and the task information, statically schedule the multiple tasks according to the scheduling algorithm to obtain static scheduling results, and obtain standard running start times corresponding to the multiple tasks based on the static scheduling results, wherein the static scheduling results at least include task running start times, and the standard running start times are task running start times of the multiple tasks based on a unified time reference point;

Based on the static scheduling result, the standard running start time, the unified time reference point and the task cycle, multiple processors are regularly woken up and scheduled to process corresponding tasks.

As an optional implementation, the inter-task dependency is the degree of operational dependency between tasks; and/or the inter-task communication limit is the maximum time required for inter-task data communication; and/or the worst task execution time is the longest time that the task can be executed under an operational condition.

As an optional implementation manner, the statically scheduling the multiple tasks according to the scheduling algorithm based on the processor information and the task information to obtain the static scheduling result includes:

Determine the number of cores of each processor based on the processor information;

Determine the minimum delay of the task link based on the inter-task dependency and the inter-task communication limit value;

Determine the maximum CPU load based on the task cycle and the worst execution time of the task;

Determine the task running start time based on the processor information and the CPU maximum load, wherein the task running start time is at least used to determine processor load balancing;

When the task link delay is minimal and the processor load is balanced, the tasks corresponding to each processor are statically scheduled based on the core number of each processor to obtain a static scheduling table.

As an optional implementation, the CPU number corresponding to each task may also be configured based on the static orchestration result to perform task core binding.

As an optional implementation, the deterministic task scheduling method further includes dynamically lowering the priority order of the task in the static scheduling result and giving priority to scheduling the task with a high priority when there is at least one task whose running time exceeds the corresponding task worst execution time.

A second aspect of the present invention provides a deterministic task scheduling method, which is applied between multiple processors, comprising:

Configure the processors to communicate in master-slave mode, and each processor has a corresponding master scheduling server or slave scheduling server;

The processor is configured to adopt the Server/Client mode and communicate internally, and the processor has a scheduling Server and a scheduling Client internally, and the scheduling Client integrates the static orchestration results;

The static scheduling result is generated by statically scheduling multiple tasks according to the processor information and the task information according to the scheduling algorithm, wherein the task information includes the task cycle, the inter-task dependency, the inter-task communication limit value and the worst task execution time, and based on the static scheduling result, the standard running start time corresponding to the multiple tasks is obtained, and the standard running start time is the task running start time of the multiple tasks based on a unified time reference point;

The master scheduling server calculates a unified time reference point and synchronizes it to all slave scheduling servers;

The scheduling server synchronizes the unified time reference point to all scheduling clients;

The scheduling Client periodically wakes up and schedules multiple processors to process corresponding tasks based on the static scheduling result, the standard running start time, the unified time reference point and the task cycle.

As an optional implementation, the static arrangement result is a static arrangement table;

The static scheduling table can generate a configuration file including the task CPU number, task cycle, task worst execution time and standard running start time;

When multiple processors are woken up and scheduled to process corresponding tasks at a scheduled time, multiple processors are woken up and scheduled to process corresponding tasks at a scheduled time according to a preset scheduling rule.

A third aspect of the present invention provides a deterministic task scheduling system, comprising:

An acquisition unit, at least used to acquire processor information and task information, wherein the task information at least includes a task cycle, inter-task dependency, an inter-task communication limit value, and a worst-case execution time of a task;

a scheduling unit, at least configured to perform static scheduling on the plurality of tasks according to a scheduling algorithm based on the processor information and the task information to obtain a static scheduling result, and to obtain standard running start times corresponding to the plurality of tasks based on the static scheduling result, wherein the static scheduling result at least includes a task running start time, and the standard running start time is the task running start time of the plurality of tasks based on a unified time reference point;

The scheduling unit is at least used to regularly wake up and schedule multiple processors to process corresponding tasks based on the static scheduling result, the standard operation start time, the unified time reference point and the task cycle.

A fourth aspect of the present invention provides a deterministic task scheduling system, which is applied between multiple processors and includes:

The inter-processor configuration unit is at least used to configure the processors to adopt a master-slave mode and communicate with each other, and each processor has a corresponding master scheduling Server or slave scheduling Server;

The processor configuration unit is at least used to configure the processor to adopt the Server/Client mode and communicate, and the processor has a scheduling Server and a scheduling Client, and the scheduling Client integrates the static orchestration result;

A static scheduling unit, at least used to generate a static scheduling result required by the configuration unit in the processor, wherein the static scheduling result is generated by statically scheduling multiple tasks according to the processor information and the task information in accordance with the scheduling algorithm, wherein the task information includes a task cycle, a dependency between tasks, a communication limit value between tasks, and a worst-case execution time of tasks, and based on the static scheduling result, a standard running start time corresponding to the multiple tasks is obtained, wherein the standard running start time is the task running start time of the multiple tasks based on a unified time reference point;

The time synchronization unit is used at least by the master scheduling server to calculate a unified time reference point and synchronize it to all slave scheduling servers;

A time conversion unit, at least used to synchronize the unified time reference point from the scheduling server to all scheduling clients;

The scheduling wake-up unit is at least used for the scheduling Client to regularly wake up and schedule multiple processors to process corresponding tasks based on the static scheduling result, the standard operation start time, the unified time reference point and the task cycle.

A fifth aspect of the present invention provides an electronic device, comprising:

At least one processor; and at least one memory communicatively connected to the processor, wherein: the memory stores program instructions executable by the processor, and the processor calls the program instructions to execute the deterministic task scheduling method as described in the first aspect of the present invention or the second aspect of the present invention.

A sixth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a computer, the deterministic task scheduling method as described in the first aspect of the present invention or the second aspect of the present invention is executed.

A seventh aspect of the present invention provides a computer program product, which includes computer instructions. When the computer instructions are executed by a processor, they implement the deterministic task scheduling method as described in the first aspect of the present invention or the second aspect of the present invention.

The present invention ensures that the tasks have a unified time base for task execution time by obtaining static scheduling results and standard running start time, which can improve the stability and reliability of task scheduling. The present invention can be widely applied to various real-time systems, and can optimize the response time of the overall task link while ensuring scheduling certainty, so that the system can respond more quickly.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a schematic flow chart of a deterministic task scheduling method according to an embodiment of the present invention.

FIG. 2 is a schematic flow chart of another deterministic task scheduling method according to an embodiment of the present invention.

FIG. 3 is a schematic diagram of an interface of a static scheduling table according to an embodiment of the present invention.

FIG. 4 is a schematic flow chart of another deterministic task scheduling method according to an embodiment of the present invention.

FIG5 is a diagram showing the operating principle of a deterministic task scheduling method according to an embodiment of the present invention.

FIG6 is a block diagram of a deterministic task scheduling system according to an embodiment of the present invention.

FIG. 7 is a block diagram of yet another deterministic task scheduling system according to an embodiment of the present invention.

FIG. 8 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.

DETAILED DESCRIPTION

The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

It should be understood that the terms "first", "second", and "third" etc. in the claims, specifications, and drawings of the present disclosure are used to distinguish different objects rather than to describe a specific order. The terms "include" and "comprise" used in the specification and claims of the present disclosure indicate the presence of the described features, wholes, steps, operations, elements, and/or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and/or their collections. It should also be understood that the terms used in this disclosure specification are only for the purpose of describing specific embodiments and are not intended to limit the present disclosure.

Take the automobile autonomous driving technology as an example. With the development of autonomous driving technology and the improvement of automobile performance, while it brings convenience to users, it has higher requirements for real-time performance. Since there are a large number of asynchronous tasks in the system, if there is no appropriate scheduling strategy, the CPU resources will be improperly occupied, which will affect the user's driving safety.

The deterministic task scheduling method provided by the present invention is applied to a multi-processor real-time system, which can ensure that each task is executed on time according to a predetermined strategy and eliminate the uncertainty of traditional scheduling methods.

Referring to FIGS. 1-3 , the present invention provides a deterministic task scheduling method, comprising the following steps:

Step 110: Obtain processor information and task information.

Specifically, the processor information includes the number of processors, that is, the number of cores of each processor. If the number is fixed, the more processors there are, the more scheduling methods can be sufficient. The number of processors needs to be considered and involved in the scheduling during static scheduling. Ideally, the number of processors is greater than or equal to the number of tasks;

Specifically, the task information includes the number of tasks, the task period, the inter-task dependency, the inter-task communication limit, and the worst task execution time.

The inter-task dependency is the degree of operational dependency between tasks, which can be used to quantify the operational dependency between tasks. For example, the dependency between task A and task B is 1, and the dependency between task B and task C is 0, indicating that there is an operational dependency between task A and task B, but there is no operational dependency between task B and task C. In the actual deterministic scheduling link, priority will be given to scheduling two tasks with operational dependencies in a continuous manner in time to achieve the purpose of minimum delay.

In addition, the inter-task communication limit is also called the maximum time required for data communication between tasks (Worst-Case Communication Time, WCCT); the worst-case execution time (Worst-Case Execution Time, WCET) of the task is the longest time that the task can be executed under an operational condition, wherein the maximum CPU load of the task can be roughly calculated based on the ratio of WCET to Period, which serves as a reference for determining processor load balancing during subsequent static scheduling.

Step S120: Based on the processor information and the task information, statically schedule the multiple tasks according to the scheduling algorithm to obtain static scheduling results, and obtain standard running start times of the multiple tasks based on the static scheduling results. The static scheduling results at least include task running start times, and the standard running start time is the task running start time of the multiple tasks based on a unified time reference point.

Specifically, the scheduling algorithm can be implemented by referring to the idea of traversing all combinations, that is, traversing all combinations of all tasks on different CPUs and different start times to obtain the global optimal solution. It can also be implemented by referring to existing static scheduling algorithm rules, such as simulated annealing algorithm or genetic algorithm, etc., which will not be repeated here.

Specifically, the static scheduling result can be presented in the form of a static scheduling table. The static scheduling table for multiple processors schedules multiple tasks on multiple processors together.

Specifically, the standard operation start time is referred to as Eerliest, and each task is scheduled based on a unified time reference point (or the same time reference point) through GPTP. Since two tasks with operation dependencies are deployed on different processors, the wake-up time points of these tasks need to be calculated based on the unified time reference point T0 under GPTP. At this time, the time difference between the start times of the two tasks can reflect this operation dependency. The purpose of using GPTP to synchronize the time of tasks on multiple processors is to ensure that tasks on multiple processors in a system are based on a unified time reference point T0. In addition, if there are other protocols or methods that achieve the purpose of the present invention, GPTP can be replaced, and the technical concept to be protected by the present invention will not be affected.

As shown in FIG. 2 , in one embodiment of the present invention, with respect to step S200: the static scheduling of multiple tasks according to the scheduling algorithm based on the processor information and the task information to obtain a static scheduling result includes:

Step S121: determining the number of cores of each processor based on the processor information;

Step S122: determining a minimum delay of a task link based on the inter-task dependency and the inter-task communication limit value;

Step S123: determining the maximum CPU load based on the task cycle and the worst execution time of the task;

Step S124: determining the task running start time based on the processor information and the CPU maximum load, wherein the task running start time is at least used to determine processor load balancing;

Step S125: When the task link delay is minimized and the processor load is balanced, the tasks corresponding to each processor are statically scheduled based on the core number of each processor to obtain a static scheduling table. The static scheduling table specifies the CPU number and standard running start time of each task. The CPU number is related to the bound core, and the standard running start time is related to the wake-up time point (or scheduling time point).

Specifically, the principle of static scheduling in the present invention is to minimize the overall link delay of the task and balance the CPU load to ensure real-time and reliability. Among them, the overall link delay of the task is minimized, and two tasks with running dependencies need to be scheduled continuously in time, and only the WCCT time gap is reserved between tasks as much as possible. To ensure CPU load balance, the following two aspects need to be considered:

First, from the perspective of CPU cores: the total CPU load of tasks on each CPU core needs to be balanced, that is, tasks need to be evenly distributed to all CPUs;

Secondly, from the dimension of time: at any time, the number of CPU cores in operation is balanced, that is, more CPU resources need to be reserved for non-deterministic tasks at any time to prevent non-deterministic tasks from being unable to continue due to failure to obtain CPU resources in time, thereby causing BUGs. Therefore, the start time of each task needs to be considered in the dimension of time.

In addition, based on the evaluation of complex equilibrium, local optimization algorithms and the like can be implemented, which will not be elaborated here.

The present invention obtains static scheduling results and standard operation start time based on the consideration of how to ensure that the task link has minimum delay and how to ensure CPU load balancing, and presents the static scheduling structure in the form of a visual static scheduling table to ensure the determinism of task scheduling, especially the normal operation of non-deterministic tasks.

Step S130: based on the static scheduling result, the standard operation start time, the unified time reference point and the task cycle, regularly wake up and schedule multiple processors to process corresponding tasks.

Exemplarily, the static scheduling table obtained based on the processor information and the task information includes specific information related to the actual running time of each task, such as each task period. In addition, based on the static scheduling table, the task running start time and the task running start time of each task relative to the unified time reference point T0, i.e., the Eerliest, can also be obtained. Since the scheduling of tasks is periodic, all wake-up time points of the task can be calculated according to the following formula: Ta=T0+N*Period+Earliest, where N≥0, and N is the number of wake-ups.

Here, the deterministic scheduling method provided by the present invention can be widely applied to various real-time systems. While ensuring the scheduling determinism, the response time of the overall task link can also be optimized, so that the system can respond more quickly.

In one embodiment of the present invention, the static scheduling result can also be used to configure the CPU number corresponding to each task to perform task core binding to prevent deterministic tasks from being affected by other non-deterministic tasks.

Generally speaking, deterministic tasks are usually at the user level, and the priority of user-level threads is generally not recommended to be higher than certain special threads in the kernel, such as hard interrupt threads (after the Linux RT patch is applied, interrupts are forced to be threaded). In order to avoid the impact of similar higher-priority threads on confirmatory tasks, on the one hand, the execution time of higher-priority threads can be minimized (low-priority processing can be run in low-priority threads); on the other hand, the core binding strategy for deterministic scheduling tasks can be enabled to monopolize one or more deterministic tasks on one or several CPU cores to avoid the impact of high-priority threads.

Deterministic scheduling requires that the worst-case execution time of deterministic tasks is predictable, but developers cannot fully guarantee this, so special handling is required for abnormal situations, mainly including the following two abnormal situations:

The first abnormal situation: It is believed that there are different priorities among deterministic tasks, so a relatively higher scheduling priority needs to be set for high-priority tasks. When the running time of the high-priority task exceeds the WCET due to an abnormal situation, it may squeeze out other low-priority deterministic tasks. There are two results at this time. The first result is that there are still idle CPU resources at this time, so the low-priority task can still grab the CPU resources to run, and the impact is small at this time; the second result is that there are no idle CPU resources at this time, which will cause a large scheduling delay for the low-priority task.

The second exception is that all deterministic tasks are considered equal, so these deterministic tasks will be set with an initial higher scheduling priority.

In order to solve the above problem, the present invention proposes that when there is at least one task whose running time exceeds the corresponding task worst execution time, the priority order of the task in the static scheduling result is dynamically lowered, and the task with high priority is scheduled first.

Specifically, when the running time of a task exceeds the corresponding task WCET, it means that the task has an abnormal situation. If the task continues to run, it will inevitably occupy the running time of other deterministic tasks. At this time, the task priority strategy can be dynamically adjusted. When the running time of the task exceeds the task WCET, the priority of the task is lowered so that the task can give up CPU resources to other deterministic tasks. When the CPU resources are idle, the task with a lowered priority can still seize the CPU resources to continue to run and complete. If the task that seizes the CPU resources to continue to run at this time also encounters a similar situation, it can be postponed or an alarm can be issued.

The priority strategy proposed in the present invention is proposed based on the actual deterministic task running scenario. Through the priority stratification strategy, the CPU resources of high-priority tasks can be guaranteed first. By dynamically adjusting the task priority strategy, when a single task has an exception, the priority of the task is lowered to give up CPU resources to prevent affecting other deterministic tasks.

In order to better express the technical solution to be protected by the present invention, as shown in FIG. 4 and FIG. 5 , the second aspect of the present invention proposes a deterministic task scheduling method, which is applied between multiple processors and mainly includes the following steps:

Step S210: configure the processors to communicate in master-slave mode, and each processor has a corresponding master scheduling server or slave scheduling server.

Step S220: The processor is configured to adopt the Server/Client mode and communicate internally, and the processor has a scheduling Server and a scheduling Client internally, and the scheduling Client integrates the static orchestration result.

Specifically, the static scheduling result is generated by statically scheduling multiple tasks according to the processor information and the task information according to the scheduling algorithm, the task information includes the task cycle, the dependency between tasks, the communication limit value between tasks and the worst execution time of the task, and based on the static scheduling result, the standard running start time corresponding to the multiple tasks is obtained, and the standard running start time is the task running start time of the multiple tasks based on the unified time reference point.

Step S230: The master scheduling server calculates a unified time reference point and synchronizes it to all slave scheduling servers;

Step S240: The scheduling server converts the unified time reference point into a monotonic time reference point, and synchronizes it to all scheduling clients; specifically, in order to avoid the impact of time rollback or time jump, the actual scheduling clock uses a monotonic clock.

Step S250: the scheduling Client periodically wakes up and schedules multiple processors to process corresponding tasks based on the static scheduling result, the standard operation start time, the unified time reference point and the task cycle.

Specifically, the static scheduling result is expressed in at least a static scheduling table; the input task information Peroid and WCET of the static scheduling algorithm, the output task CPU number and Earliest of the static scheduling algorithm, these four parameters are included in the configuration file.

In step S250, the scheduling Client periodically wakes up and schedules multiple processors to process corresponding tasks based on the static scheduling result, the standard operation start time, the unified time reference point and the task cycle, including:

Step S251: Timely wake up and schedule multiple processors to process corresponding tasks. Specifically, the preset scheduling rule refers to the scheduling time point formula: Ta=T0+N*Period+Earliest. The configuration file is used to store the static scheduling results and the task CPU number and Earliest obtained based on the static scheduling results.

Specifically, please continue to refer to FIG. 3 and FIG. 5 , which show a specific deterministic task scheduling working principle diagram between processor A and processor B. The following steps are now described:

Step ①, when the Client (scheduling Client) and Server (scheduling Server) inside processor A and processor B establish communication accordingly, and the scheduling Client is integrated with a configuration file, it corresponds to step S220;

Step ②: Establishing communication between processor A and processor B, corresponding to step S210;

Step 3: The master scheduling server confirms T0 and broadcasts it, calculates the unified scheduling time reference point T0, and synchronizes it to all slave scheduling servers, corresponding to step S230;

Step ④: The scheduling server converts the unified time reference point T0 from the GPTP time reference point into a monotonic time reference point, and broadcasts it to all scheduling clients, corresponding to step S240;

Step ⑤: After the scheduling Client obtains the monotonic time reference point, it reads the parameters such as Period and Earliest in the configuration file, combines the configuration file and the unified time reference point information, and wakes up the scheduling task regularly according to the scheduling time point formula: Ta=T0+N*Period+Earliest, corresponding to step S250, and executes task scheduling through the scheduling Client.

Among them, Figure 3 shows the task numbers of processor A and processor B respectively (for example, task_a, task_b, task_c, task_d, task_e, task_f, task_g, task_h and task_i). Taking task task_e of processor B as an example, based on the scheduling time point formula Ta=T0+N*Period+Earliest, where N is the number of wake-ups, it is calculated. When N=0, it means that it has not been awakened before, and its first awakening time is Ta=T0+N*Period+Earliest=T0+0*20ms+3ms=3ms; when N=1, it means that its second awakening time is Ta=T0+N*Period+Earliest=T0+2*20ms+3ms=23ms.

Here, in this way, task scheduling between processors and within processors can be achieved to ensure the performance and reliability of the real-time system.

As shown in FIG6 , the third aspect of the present invention provides a deterministic task scheduling system, including:

The acquisition unit 310 is at least used to acquire processor information and task information, wherein the task information at least includes task cycle, inter-task dependency, inter-task communication limit value and task worst execution time;

The scheduling unit 320 is at least used to perform static scheduling on the multiple tasks according to the scheduling algorithm based on the processor information and the task information to obtain a static scheduling result, and obtain standard running start times corresponding to the multiple tasks based on the static scheduling result, wherein the static scheduling result at least includes a task running start time, and the standard running start time is the task running start time of the multiple tasks based on a unified time reference point;

The scheduling unit 330 is at least used to regularly wake up and schedule multiple processors to process corresponding tasks based on the static scheduling result, the standard operation start time, the unified time reference point and the task cycle.

As shown in FIG. 7 , a fourth aspect of the present invention provides a deterministic task scheduling system, which is applied between multiple processors and includes:

The inter-processor configuration unit 410 is at least used to configure the processors to adopt a master-slave mode and communicate with each other, and to configure each processor to have a corresponding master scheduling server or slave scheduling server;

The processor configuration unit 420 is at least used to configure the processor to adopt the Server/Client mode and communicate, and the processor has a scheduling Server and a scheduling Client, and the scheduling Client integrates the static orchestration result;

The static scheduling unit 430 is at least used to generate a static scheduling result required by the configuration unit in the processor, wherein the static scheduling result is generated by statically scheduling multiple tasks according to the processor information and the task information according to the scheduling algorithm, wherein the task information includes a task cycle, a task dependency, a task communication limit value, and a task worst execution time, and based on the static scheduling result, a standard running start time corresponding to the multiple tasks is obtained, wherein the standard running start time is the task running start time of the multiple tasks based on a unified time reference point;

The time synchronization unit 440 is at least used for the master scheduling server to calculate a unified time reference point and synchronize it to all slave scheduling servers;

The time conversion unit 450 is at least used to convert the unified time reference point into a monotonic time reference point from the scheduling server, and synchronize it to all scheduling clients;

The scheduling wake-up unit 460 is at least used for the scheduling Client to regularly wake up and schedule multiple processors to process corresponding tasks based on the static scheduling result, the standard operation start time, the unified time reference point and the task cycle.

As shown in FIG8 , the fifth aspect of the present invention provides an electronic device, including:

At least one processor; and at least one memory communicatively connected to the processor, wherein: the memory stores program instructions executable by the processor, and the processor calls the program instructions to execute the deterministic task scheduling method as described in any of the above embodiments.

A sixth aspect of the present invention provides a computer program product, which includes computer instructions, and when the computer instructions are executed by a processor, the deterministic task scheduling method as described in any of the above embodiments is implemented.

A seventh aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a computer, the deterministic task scheduling method as described in any one of the above embodiments is executed.

It can be understood that the computer-readable storage medium may include: any entity or device capable of carrying a computer program, a recording medium, a USB flash drive, a mobile hard disk, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), and a software distribution medium, etc. The computer program includes a computer program code. The computer program code may be in source code form, object code form, an executable file, or some intermediate form, etc. The computer-readable storage medium may include: any entity or device capable of carrying a computer program code, a recording medium, a USB flash drive, a mobile hard disk, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), and a software distribution medium, etc.

In certain embodiments of the present invention, the electronic device may include a controller or a processor, and the controller is a single-chip microcomputer chip that integrates a processor, a memory, a communication module, etc. The processor may refer to a processor included in the controller. The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc.

Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code that includes one or more executable instructions for implementing the steps of a specific logical function or process, and the scope of the preferred embodiments of the present invention includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present invention belong.

Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims (12)

1. A deterministic task scheduling method, comprising:

acquiring processor information and task information, wherein the task information at least comprises task periods, inter-task dependencies, inter-task communication limit values and worst task execution time;

Based on the processor information and the task information, performing static arrangement on a plurality of tasks according to an arrangement algorithm to obtain a static arrangement result, and acquiring standard operation starting time corresponding to the plurality of tasks based on the static arrangement result, wherein the static arrangement result at least comprises task operation starting time, and the standard operation starting time is task operation starting time of the plurality of tasks based on a unified time reference point;

and scheduling a plurality of processors to process corresponding tasks in a timing wake-up mode based on the static arrangement result, the standard running starting time, the unified time reference point and the task period.

2. The deterministic task scheduling method according to claim 1, wherein the inter-task dependencies are inter-task run dependencies; and/or the inter-task communication limit value is a maximum time required for inter-task data communication; and/or the worst execution time of the task is the longest time that the task can be executed under the executable condition.

3. The deterministic task scheduling method according to claim 2, wherein said statically arranging a plurality of tasks according to an arrangement algorithm based on said processor information and said task information to obtain a static arrangement result comprises:

determining the core number of each processor based on the processor information;

determining a minimum time delay of a task link based on the inter-task dependencies and the inter-task communication limit value;

Determining a CPU maximum load based on the task period and the task worst execution time;

Determining the task running starting time based on the processor information and the CPU maximum load, wherein the task running starting time is at least used for determining processor load balancing;

And when the time delay of the task link is minimum and the loads of the processors are balanced, carrying out static arrangement on the tasks corresponding to the processors based on the core numbers of the processors to obtain a static arrangement table.

4. A deterministic task scheduling method according to any one of claims 1-3, wherein the CPU number corresponding to each task is also configurable based on the static orchestration result to perform task binding.

5. A deterministic task scheduling method according to any one of claims 1-3, further comprising dynamically lowering the priority order of said tasks in the static orchestration result when there is at least one task present running longer than the worst execution time of the corresponding task, prioritizing said prioritized tasks.

6. A deterministic task scheduling method, applied between multiple processors, comprising:

configuring each processor to adopt a master-slave mode and communicate, wherein each processor is provided with a corresponding master scheduling Server or slave scheduling Server respectively;

Configuring a processor to adopt a Server/Client mode and communicate, wherein a scheduling Server and a scheduling Client are arranged in the processor, and the scheduling Client is integrated with a static arrangement result;

The static arrangement result is generated by carrying out static arrangement on a plurality of tasks according to arrangement algorithm according to processor information and task information, the task information comprises task period, inter-task dependency, inter-task communication limit value and task worst execution time, and standard operation starting time corresponding to the plurality of tasks is obtained based on the static arrangement result, and the standard operation starting time is the task operation starting time of the plurality of tasks based on a unified time reference point;

the master scheduling Server calculates a unified time reference point and synchronizes the unified time reference point to all the slave scheduling servers;

Synchronizing the unified time reference point to all scheduling clients from a scheduling Server;

And the scheduling Client wakes up and schedules a plurality of processors to process corresponding tasks at regular time based on the static arrangement result, the standard running starting time, the unified time reference point and the task period.

7. The deterministic task scheduling method according to claim 6, wherein the static orchestration result is a static orchestration table;

The static arrangement table can correspondingly generate a configuration file containing a task CPU number, a task period, a task worst execution time and a standard operation starting time;

And when the processors process the corresponding tasks, the processors process the corresponding tasks according to the preset scheduling rules in a timing wake-up scheduling mode.

8. A deterministic task scheduling system, comprising:

The system comprises an acquisition unit, a processing unit and a processing unit, wherein the acquisition unit is at least used for acquiring processor information and task information, and the task information at least comprises a task period, a task dependence degree, a task communication limit value and a task worst execution time;

The arrangement unit is at least used for carrying out static arrangement on a plurality of tasks according to an arrangement algorithm based on the processor information and the task information to obtain a static arrangement result, and obtaining standard running starting time corresponding to the plurality of tasks based on the static arrangement result, wherein the static arrangement result at least comprises task running starting time, and the standard running starting time is the task running starting time of the plurality of tasks based on a unified time reference point;

And the scheduling unit is at least used for scheduling the plurality of processors to process corresponding tasks in a timing wake-up mode based on the static arrangement result, the standard running starting time, the unified time reference point and the task period.

9. A deterministic task scheduling system for use between multiple processors, comprising:

The inter-processor configuration unit is at least used for configuring each processor to adopt a master-slave mode and communicate, and each processor is respectively provided with a corresponding master scheduling Server or slave scheduling Server;

The processor internal configuration unit is at least used for configuring the internal of the processor to adopt a Server/Client mode and communicate, and the internal of the processor is provided with a scheduling Server and a scheduling Client, and the scheduling Client is integrated with a static arrangement result;

The static arrangement unit is at least used for generating a static arrangement result required by the configuration unit in the processor, the static arrangement result is generated by carrying out static arrangement on a plurality of tasks according to an arrangement algorithm according to processor information and task information, the task information comprises task periods, inter-task dependencies, inter-task communication limit values and task worst execution time, and standard running starting time corresponding to the plurality of tasks is obtained based on the static arrangement result, and the standard running starting time is the task running starting time of the plurality of tasks based on a uniform time reference point;

The time synchronization unit is at least used for calculating a unified time reference point by the master scheduling Server and synchronizing the unified time reference point to all the slave scheduling servers;

the time conversion unit is at least used for synchronizing the unified time reference point to all dispatching clients from the dispatching Server;

And the scheduling awakening unit is at least used for the scheduling Client to wake up and schedule a plurality of processors to process corresponding tasks at regular time based on the static arrangement result, the standard running starting time, the unified time reference point and the task period.

10. An electronic device, comprising:

at least one processor; and at least one memory communicatively coupled to the processor, wherein: the memory stores program instructions executable by the processor, the processor invoking the program instructions capable of performing the deterministic task scheduling method according to any of claims 1-7.

11. A computer readable storage medium, characterized in that a computer program is stored thereon, which, when being run by a computer, performs the deterministic task scheduling method according to any one of claims 1 to 7.

12. A computer program product comprising computer instructions which, when executed by a processor, implement the deterministic task scheduling method according to any one of claims 1-7.