关于daemonize()

lock_kernel();
siginitsetinv(¤t->blocked, SHUTDOWN_SIGS);
daemonize();
target_data.thread_id = current;
unlock_kernel();
注：SHUTDOWN_SIGS前面定义为# define SHUTDOWN_SIGS (sigmask(SIGKILL)|sigmask(SIGINT)|sigmask(SIGTERM)) target_data是前面定义的一个全局数据结构，thread_id是它的一个task_struct 结构的一个成员。这里的daemonize函数是什么作用?

   1. linux-2.4.20\kernel\sched.c start from line 1283:  
   2.   
   3. /* 
   4. *        Put all the gunge required to become a kernel thread without 
   5. *        attached user resources in one place where it belongs. 
   6. */  
   7.   
   8. void daemonize(void)  
   9. {  
  10.         struct fs_struct *fs;  
  11.   
  12.   
  13.         /* 
  14.          * If we were started as result of loading a module, close all of the 
  15.          * user space pages.  We don't need them, and if we didn't close them 
  16.          * they would be locked into memory. 
  17.          */  
  18.         exit_mm(current); //因为是内核级daemon，所以不会再需要用户级的内存页面，释放之。这些页面是因为加载模块而申请到的。  
  19.   
  20.         current->session = 1;  
  21.         current->pgrp = 1;  
  22.         current->tty = NULL;//因为是内核级daemon，所以不再需要终端控制台，所属会话为init会话，parent group为1（init），这样就成为init任务的一部分了。下面的代码则是释放文件系统相关资源，并且转为指向init任务的文件资源。  
  23.         /* Become as one with the init task */  
  24.   
  25.         exit_fs(current);        /* current->fs->count--; */  
  26.         fs = init_task.fs;  
  27.         current->fs = fs;  
  28.         atomic_inc(&fs->count);  
  29.         exit_files(current);  
  30.         current->files = init_task.files;  
  31.         atomic_inc(*t->files->count);  
  32. }