微视linux waitX的意义

深入理解linux 47页里面介绍父进程通过wait4/waitpid等方法监控子进程的退出状态，子进程进入僵死状态时，父进程通过waitX就会促使父进程释放子进程的资源，促使子进程从僵死状态进入终止状态。

我们通过2.6.23版本内核代码分析该场景。

long do_wait(pid_t pid, int options, struct siginfo __user *infop, int __user *stat_addr, struct rusage __user *ru)

	DECLARE_WAITQUEUE(wait, current);
	struct task_struct *tsk;
	
	add_wait_queue(&current->signal->wait_chldexit,&wait);//子进程exit可以唤醒父进程
repeat:
	/*如果父进程没有收到子进程的结束信号，则将在schedule函数执行后睡眠。直到子进程发出结束信号，唤醒父进程。*/
	current->state = TASK_INTERRUPTIBLE;
	
	tsk = current;//父进程被唤醒
	
	do
		struct task_struct *p;
		struct list_head *_p;
		
		list_for_each(_p,&tsk->children)//遍历子进程，根据子进程当前状态决定下一步动作
			p = list_entry(_p, struct task_struct, sibling);
			
			switch (p->state) 
			case TASK_TRACED:
				... ...
				
			case TASK_STOPPED:
				... ...
				
			default:
				if (p->exit_state == EXIT_DEAD)
					continue;
				if (p->exit_state == EXIT_ZOMBIE)
					retval = wait_task_zombie(p, (options & WNOWAIT), infop, stat_addr, ru);
					=>int wait_task_zombie(struct task_struct *p, int noreap, struct siginfo __user *infop, int __user *stat_addr, struct rusage __user *ru)
						state = xchg(&p->exit_state, EXIT_DEAD);
						if (p != NULL)
							release_task(p);
							/*释放子进程描述符*/
							=>void release_task(struct task_struct * p)
								call_rcu(&p->rcu, delayed_put_task_struct);

						return retval;
					break;
			
			tsk = next_thread(tsk);
			
	while (tsk != current);
	
end:
	current->state = TASK_RUNNING;//父进程恢复就绪态
	remove_wait_queue(&current->signal->wait_chldexit,&wait);//子进程信号已经没用了
	
	return retval;