Asynchronous log processing method based on message driving

Abstract

The invention discloses an asynchronous log processing method based on message driving. The method comprises the following steps: acquiring log information of a plurality of user threads; adding a first mutual exclusion lock to the multi-thread concurrent user thread; selecting a memory block matched with the log information in size from the static buffer area according to the log information of each user thread, and adding a call lock to the selected memory block to obtain an allocated memory block; releasing the call lock, and writing corresponding log information into the unlocked allocated memory block; a calling thread is notified through a calling event, so that the calling thread reads multiple pieces of log information from the static buffer area and sends the multiple pieces of log information to multiple processing threads; each processing thread controls the output format of the log information through a custom callback function. According to the method, a double-cache mechanism is adopted, simultaneous reading and writing of multiple threads can be achieved, mutual blocking of the reading and writing threads is effectively avoided, the asynchronous log processing efficiency is improved, and the real-time performance and reliability of an automatic driving system are guaranteed.

Description

technical field [0001] The invention relates to the technical field of automatic driving, in particular to a message-driven asynchronous log processing method. Background technique [0002] As autonomous driving becomes the mainstream trend of future cars, various types of autonomous vehicles are constantly being born. The core of autonomous vehicles is a central computing platform, which is responsible for computing tasks such as vehicle fusion perception, positioning, and planning. The real-time requirements of the driving system are extremely high; and due to the inevitable need for logs in practical engineering applications to locate and analyze the problems of the automatic driving system, track the real-time operating status of the automatic driving system, and there is a problem in the process of writing the log. Frequent I / O operations will affect system performance. [0003] Existing log systems include a synchronous log system and an asynchronous log system; among...

AI Technical Summary

Problems solved by technology

[0005] However, the asynchronous log system cannot support multiple output sources at the same time. Neither the corresponding log level nor the log output format can be customized. It must rely on a fixed format, resulting in low readability of the log; at the same time, Because the string copy involves dynamic allocation of memory and switching between user mode and kernel mode, the performance consumption is large, resulting in reduced efficiency; and because the process of writing logs will involve multiple threads reading and writing the same buffer, On the one hand, no corresponding measures have been taken to solve the problem of lock granularity, resulting in an increase in lock granularity; on the other hand, effective concurrent reading and writing cannot be guaranteed, resulting in thread blocking and reducing the efficiency of log processing

[0006] The log system is very important in the application of fault location. The general log system design does not have particularly high performance requirements, but the automatic driving system has high concurrency and low latency application scenarios due to real-time and reliability requirements. In addition, the design of the log system itself has a certain complexity, and the general log system is not very feasible when dealing with automatic driving scenarios.

Images