In this demo we want to show that a thread can yield the utilization of the CPU directly to another thread.
To illustrate the mechanism in it, we will explain:
yield_to(insrc/user/thread.c)
When this function is called, the kernel will check if the thread we specified is available to yield to.
If the thread can be yielded to, we will set next_thread to the address of this thread and call the do_scheduler function.
Since we have decided next_thread, the scheduler will do housekeeping tasks and help us to setup the thread we had already chosen.
If the thread we specified is unavailable to yield, do_scheduler will not be called and we will go back to the point we called yield_to().
At the beginning of this demo, we create 3 named threads first, second and last. After the thread setup ends, the thread first will get the use of the CPU and yield to the thread second specified by its thread ID (tid).
When the thread second gets the use, it yields to first and last respectively and the thread first exits normally. This shows that we can yield to a lower tid as well as a higher one.
After the thread second does yield(), the thread last will reach its end. At this point, there is only thread second in the all_threads array. So the next time it tries to do yield_to() or yield(), it should fail and thus yielding will not occur.
As a result, the thread second will exit normally, and the kernel will also exit since there are no more threads to run.