Coroutine processing and management method based on interrupt reentry mechanism

Abstract

The invention discloses a coroutine processing and management method based on an interrupt reentry mechanism; and in a processing method, a coroutine comprises a main coroutine and a sub-coroutine, the interrupt reentry step comprises the following steps: S1, enabling the main coroutine to throw an interrupt request, and so a bottom layer framework of the main coroutine responds to the interrupt request and asynchronously sends a task request; s2, switching to an external system processing logic, wherein the main co-thread releases thread resources; s3, executing reentry of the task request response into the sub-coroutine through a callback interface, and recovering to a context state during interruption; s4, enabling the sub-coroutines to continue to process subsequent logic until execution is finished, and then thread resources are released; or using the sub-coroutine as a new main coroutine to throw out the interrupt request, and repeating the steps S1 to S4; in the management method, the coroutine combination is an application, and the application program entry OnProcess () comprises a CallBatch () interface in the processing method. According to the invention, the developmentand maintenance complexity and workload of a scheduling program with a large number of parallel branches are greatly reduced.

Description

technical field [0001] The invention belongs to the technical field of computers, and in particular relates to a coroutine processing and management method based on an interrupt reentry mechanism. Background technique [0002] In business application development, it is often necessary to develop some schedulers. These programs need to advance many branch processes at the same time, and many branch processes will take a long time to execute due to external cooperation, such as figure 1 The scheduling process shown (wherein, B, C, and D all represent submitting a batch task and waiting for the completion result of the task). The above-mentioned scheduler is developed in a common programming language, and usually cooperates with the outside world in a synchronous blocking manner. Therefore, only multi-threading can be used to achieve multi-branch parallel advancement. Its working mechanism is as follows: figure 2 shown. Obviously, the problem with such a program is that once...

AI Technical Summary

Problems solved by technology

Obviously, the problem with such a program is that once large-scale parallel advancement is required (for example, more than 1,000 parallel branches), and many branches have long-term synchronous blocking scenarios, the runtime thread resource consumption is huge

Obviously, the consumption of thread resources is minimized, but the problem with such a program is that the programming model is complex, does not conform to people's thinking habits, the maintainability of the code is poor, and it cannot even support component-based development, and the Process method cannot be encapsulated into A method that returns the completion result when the E action completes

Unfortunately, there are few languages ​​that support this coroutine programming model, and the most widely used languages ​​in the industry such as Java and C++ do not support it; and even if the language itself supports this coroutine programming model, most languages ​​generally do not support it. The program automatically has some high-availability features, such as the same scheduler can be switched to run, and the program can continue to execute after the system is restored.

Images