linux创建进程.docVIP

fork,vfork,clone都是linux的系统调用，用来创建子进程的（确切说vfork创造出来的是线程）。 先介绍下进程必须的4要点： a.要有一段程序供该进程运行，就像一场戏剧要有一个剧本一样。该程序是可以被多个进程共享的，多场戏剧用一个剧本一样。 b.有起码的私有财产，就是进程专用的系统堆栈空间。 c.有“户口”，既操作系统所说的进程控制块，在linux中具体实现是task_struct d.有独立的存储空间。 当一个进程缺少d条件时候，我们称其为线程。 1.fork 创造的子进程复制了父亲进程的资源，包括内存的内容task_struct内容（2个进程的pid不同）。这里是资源的复制不是指针的复制。下面的例子可以看出 [root@liumengli program]# cat testFork.c#includestdio.h int main() {??????? int count = 1;??????? int child; ??????? if(!(child = fork())) { //开始创建子进程??????????????? printf(This is son, his count is: %d. and his pid is: %d\n, ++count, getpid());//子进程的内容??????? } else {??????????????? printf(This is father, his count is: %d, his pid is: %d\n, count, getpid());??????? }}[root@liumengli program]# gcc testFork.c -o testFork[root@liumengli program]# ./testForkThis is son, his count is: 2. and his pid is: 3019This is father, his count is: 1, his pid is: 3018[root@liumengli program]#?从代码里面可以看出2者的pid不同，内存资源count是值得复制，子进程改变了count的值，而父进程中的count没有被改变。有人认为这样大批量的复制会导致执行效率过低。其实在复制过程中，子进程复制了父进程的task_struct，系统堆栈空间和页面表，这意味着上面的程序，我们没有执行count++前，其实子进程和父进程的count指向的是同一块内存。而当子进程改变了父进程的变量时候，会通过copy_on_write的手段为所涉及的页面建立一个新的副本。所以当我们执行++count后，这时候子进程才新建了一个页面复制原来页面的内容，基本资源的复制是必须的，而且是高效的。整体看上去就像是父进程的独立存储空间也复制了一遍。 ? 其次，我们看到子进程和父进程直接没有互相干扰，明显2者资源都独立了。我们看下面程序 [root@liumengli program]# cat testFork.c#includestdio.h int main() {??????? int count = 1;??????? int child; ??????? if(!(child = fork())) {??????????????? int i;??????????????? for(i = 0; i 200; i++) {??????????????????????? printf(This is son, his count is: %d. and his pid is: %d\n, i, getpid());??????????????? }??????? } else {??????????????? printf(This is father, his count is: %d, his pid is: %d\n, count, getpid());??????? }}[root@liumengli program]# gcc testFork.c -o testFork[root@liumengli program]# ./testFork... This is son, his count is: 46. and his pid is: 4092This is son, his count is: 47. and his pid is: 4092This is son, his count is: 48. and his pid is: 4092This is son, his count is: 4