The difference between soft real and hard real-time operating system __ucos-ii

The operating system is divided into the hardware category: Chip-level embedding (processor with program or algorithm), module-level embedding (a core module in the System) system-level embedding.
According to the software category real-time requirements: Non-real-time system, soft real-time system, hard real-time system. For real time systems, his correctness is not only related to the logic correctness of the system, but also to the system response time, if the system response does not meet the time limit, even if it can get the correct output, we can only say that he is a failure response. "soft" means that failure to meet specified time constraints does not lead to catastrophic consequences, but is disastrous for hard-time systems.

In a time-sharing operating system, computer resources are allocated evenly to all work within the system. It doesn't matter how long each task takes to complete in a time-sharing system, but in a real-time operating system, the most important concern is how long each task can be completed. In short, the biggest difference between real time and time-sharing operating systems is the concept of time limit (deadline).

From the above differences can be seen in the real-time operating system, the system must be in a specific time to complete the specified application, with strong rigidity, while the time-sharing operating system is to pay attention to the allocation of resources to each application, not too concerned about the progress of each application, what time can be completed. However, even the real-time operating system, the degree of rigidity and flexibility is also different, as if the hardness of the system is different, so the so-called hard time (hard real-time) and soft real-time (soft real-time). Hard real-time systems have a rigid, immutable time limit that does not allow any errors that exceed the time limit. A timeout error can cause damage or even a system failure or cause the system to fail to achieve its intended goal. The time limit for a soft real time system is flexible and flexible, and it can tolerate accidental timeout errors. The consequences of failure are not serious, such as only a slight reduction in the throughput of the system in the network. The key difference between hard real time and soft real-time is that soft real-time can only provide real-time in statistical sense. For example, some applications require the system to ensure that an action is completed within a specified period of time in 95% cases, not necessarily 100%. Real-time is to complete the operation in a specified period, the time of operation has the operating system factor, also has the user software factor. Non-real-time operating systems are the last time that an operating system cannot guarantee even the highest priority task to begin execution. Soft real-time operating system refers to the operating system can only guarantee the execution of the highest priority in XX time user code, but the user software can be completed in a timely manner, the operating system regardless. Hard real-time, the operating system has been responsible for the operation completed, the world's first hard real-time operating system is still in development, when it can be completed, ask God. Emphasis on a concept, not real-time, soft real-time, hard real-time three concepts, not the speed of the standard, it is a definition of the concept. Linux and RTLinux ratio, high priority task arrival, whether the shortest time or average time, Linux is better than RTLinux, but RTLinux's final time is better than Linux, so rtlinux real-time than Linux, this is the meaning of certainty. Even if you use a unshielded interrupt to complete the operation, it can only be soft real-time, not hard, because the unshielded interrupt does not tell you when to complete the work.