通过netlink达成内核模块和应用层通信
发布时间：2021-11-24 20:49:52 所属栏目：教程 来源：互联网
导读：skb常用操作函数，比较容易弄混 skb_put : skb-tail += len, skb-len += len skb_pull: skb-data += len, skb-len -= len skb_push: skb-data -= len, skb-len += len 内核版本Linux2.6.38，编译环境gcc 4.4.4，CentOS6.0 内核模块代码： /* * author: hoi071
|
skb常用操作函数，比较容易弄混 skb_put : skb->tail += len, skb->len += len skb_pull: skb->data += len, skb->len -= len skb_push: skb->data -= len, skb->len += len 内核版本Linux2.6.38，编译环境gcc 4.4.4，CentOS6.0 内核模块代码： /* * author: hoi0714@163.com * date : 2011-10-29 */ #include <linux/module.h> #include <linux/netlink.h> #include <linux/sched.h> #include <net/sock.h> #include <linux/proc_fs.h> #include <linux/netfilter.h> #include <linux/netfilter_ipv4.h> #include <linux/ip.h> #include <linux/tcp.h> #include <linux/icmp.h> #include <linux/udp.h> #define NETLINK_TEST 30 /* 调试信息 */ #define LOGMSG(fmt, arg...) do{ printk("[func:%s,line:%d]: "fmt, __FUNCTION__, __LINE__, ##arg); }while(0) /* 错误信息 */ #define LOGERR(fmt, arg...) do{ printk("[func:%s,line:%d]: "fmt, __FUNCTION__, __LINE__, ##arg); }while(0) /* 断言 */ #define ASSERT(expr) if (unlikely(!(expr))) { printk("Assertion failed! %s,%s,%s,line=%dn", #expr, __FILE__, __func__, __LINE__); } /* 消息最大值 */ #define MAX_MSG_LEN 1024 enum{ NLMSG_TYPE_NONE = 0, NLMSG_TYPE_SETPID, /* 设置PID */ NLMSG_TYPE_KERNEL, /* 消息来自内核 */ NLMSG_TYPE_APP, /* 消息来自应用层 */ }; struct nlmsg{ int type; /* 消息类型 */ int len; /* 消息长度，包括头部 */ char msg[MAX_MSG_LEN]; /* 消息正文 */ }; /* netlink socket */ static struct sock *g_nl_sk = NULL; static int g_nlpid = -1; /* 应用层接收程序PID */ /* * 发送整个从ip头开始的skb数据到应用层 * * param[in]: sk, skb发送目的socket * param[in]: skb, 待发送的skb * return -1, 失败; 0, 成功 * */ int nl_sendskb(struct sock *sk, struct sk_buff *skb) { struct iphdr *iph = NULL; struct nlmsghdr *nlh = NULL; struct sk_buff *nl_skb = NULL; int skb_len = 0; ASSERT(skb != NULL); ASSERT(sk != NULL); if(g_nlpid < 0) return 0; iph = ip_hdr(skb); skb_len = iph->tot_len; /* NLMSG_SPACE: sizeof(struct nlmsghdr) + len按4字节对齐 */ nl_skb = alloc_skb(NLMSG_SPACE(skb_len), GFP_ATOMIC); if(!nl_skb) { LOGERR("nl_skb == NULL, failed!n"); return -1; } /* * static inline struct nlmsghdr *nlmsg_put(struct sk_buff *skb, u32 pid, u32 seq, * int type, int payload, int flags); * 设置skb->tail指针指向skb->data + sizeof(struct nlmsghdr) + payload * skb->len = sizeof(struct nlmsghdr) + payload */ nlh = nlmsg_put(nl_skb, 0, 0, 0, NLMSG_SPACE(skb_len) - sizeof(struct nlmsghdr), 0); NETLINK_CB(nl_skb).pid = 0; /* 0代表数据来自内核 */ memcpy(NLMSG_DATA(nlh), (char *)iph, htons(iph->tot_len)); /* 将数据发送给进程号22345的进程 */ return netlink_unicast(sk, nl_skb, g_nlpid , MSG_DONTWAIT); } /* * 发送字符串到应用层 * * param[in]: sk, 数据发往的socket * param[in]: pmsg, 待发送字符串 * param[in]: msglen, 待发送字符串长度 * * return: -1, 失败; 0, 成功 * */ int nl_sendmsg(struct sock *sk, struct nlmsg *pmsg) { struct nlmsghdr *nlh = NULL; struct sk_buff *nl_skb = NULL; int msglen = pmsg->len; ASSERT(pmsg != NULL); ASSERT(sk != NULL); if(g_nlpid < 0) return 0; nl_skb = alloc_skb(NLMSG_SPACE(msglen), GFP_ATOMIC); if(!nl_skb) { LOGERR("nl_skb == NULL, msglen = %d, failed!n", msglen); return -1; } nlh = nlmsg_put(nl_skb, 0, 0, 0, NLMSG_SPACE(msglen) - NLMSG_HDRLEN, 0); NETLINK_CB(nl_skb).pid = 0; memcpy(NLMSG_DATA(nlh), pmsg, msglen); return netlink_unicast(sk, nl_skb, g_nlpid , MSG_DONTWAIT); } /* * 从应用层接收数据, netlink_kernel_create注册的回调 * param[in]: skb, 包含netlink数据的skb * * skb常用操作函数 * skb_put : skb->tail += len, skb->len += len * skb_pull: skb->data += len, skb->len -= len * skb_push: skb->data -= len, skb->len += len */ static void nl_recvmsg(struct sk_buff *skb) { struct nlmsg *pmsg = NULL; struct nlmsghdr *nlh = NULL; uint32_t rlen = 0; while(skb->len >= NLMSG_SPACE(0)) { nlh = nlmsg_hdr(skb); if(nlh->nlmsg_len < sizeof(*nlh) || skb->len < nlh->nlmsg_len) return; rlen = NLMSG_ALIGN(nlh->nlmsg_len); if(rlen > skb->len) rlen = skb->len; pmsg = (struct nlmsg*)NLMSG_DATA(nlh); switch(pmsg->type) { case NLMSG_TYPE_SETPID: g_nlpid = nlh->nlmsg_pid; LOGMSG("pid: %dn", g_nlpid); LOGMSG("msg: %sn", pmsg->msg); break; case NLMSG_TYPE_KERNEL: break; case NLMSG_TYPE_APP: break; } /* 获取下一条netlink消息 */ skb_pull(skb, rlen); } } /* * netfilter PRE_ROUTING钩子 * */ unsigned int pre_routing_hook(unsigned int hooknum, struct sk_buff *skb, const struct net_device *in, const struct net_device *out, int (*okfn)(struct sk_buff *)) { char *psend = "msg for kernel"; struct nlmsg msg; int ret = 0; msg.type = NLMSG_TYPE_KERNEL; msg.len = strlen(psend) + offsetof(struct nlmsg, msg) + 1; memcpy(msg.msg, psend, msg.len); //ret = nl_sendskb(g_nl_sk, skb); ret = nl_sendmsg(g_nl_sk, &msg); //LOGMSG("okn"); return NF_ACCEPT; } static struct nf_hook_ops local_in_ops __read_mostly = { .hook = pre_routing_hook, .owner = THIS_MODULE, .pf = PF_INET, .hooknum = NF_INET_PRE_ROUTING, .priority = NF_IP_PRI_FIRST }; static int __init nl_init(void) { int ret = 0; /* * struct sock *netlink_kernel_create(struct net *net, int unit, unsigned int groups, * void (*input)(struct sk_buff *skb), * struct mutex *cb_mutex, struct module *module) */ g_nl_sk = netlink_kernel_create(&init_net, NETLINK_TEST, 0, nl_recvmsg, NULL, THIS_MODULE); if (!g_nl_sk) { LOGERR("Fail to create netlink socket.n"); return -1; } ret = nf_register_hook(&local_in_ops); if(ret < 0) { LOGMSG("nf_register_hook failed!n"); goto sock_release; } LOGMSG("ok!n"); return 0; sock_release: if(g_nl_sk) sock_release(g_nl_sk->sk_socket); return -1; } static void __exit nl_exit(void) { synchronize_sched(); if(g_nl_sk) sock_release(g_nl_sk->sk_socket); nf_unregister_hook(&local_in_ops); LOGMSG("ok!n"); } module_init(nl_init); module_exit(nl_exit); MODULE_LICENSE("GPL"); MODULE_AUTHOR("hoi0714@163.com");  （编辑：常州站长网） 【声明】本站内容均来自网络，其相关言论仅代表作者个人观点，不代表本站立场。若无意侵犯到您的权利，请及时与联系站长删除相关内容! |
