從ip_queue到nfnetlink_queue(上)
1. 前言
在2.4內核中出現了ip_queue，用於將數據包從內核空間傳遞到用戶空間，其不足之處是只能有一個應用程序接收內核數據。到了2.6.14以後，新增了nfnetlink_queue，理論上可最大可支持65536個應用程序接口，而且可以兼容ip_queue。
不過從內核到用戶空間的通道還是隻有一個，實際上netfilter對每個協議族也只有一個隊列，這裏說的65536個子隊列的實現就象 802.1Q實現VLAN一樣是在數據包中設置ID號來區分的，不同ID的包都通過相同的接口傳輸，只是在兩端根據ID號進行了分類處理。
2. 用戶空間
用戶空間的支持庫包括兩個：libnfnetlink和libnetfilter_queue，後者需要前者支持，其源碼可到netfilter網站上下載。
2.1 數據結構
2.1.1
/*linux_nfnetlink.h */
// 基本屬性結構
struct nfattr
{
// 長度
u_int16_t nfa_len;
// leix
u_int16_t nfa_type; /* we use 15 bits for the type, and the highest
* bit to indicate whether the payload is nested */
} __attribute__ ((packed));
// nf基本消息結構
struct nfgenmsg {
u_int8_t nfgen_family; /* AF_xxx */
u_int8_t version; /* nfnetlink version */
u_int16_t res_id; /* resource id */
} __attribute__ ((packed));
/* nfnl是netfilter netlink的縮寫 */
// nfnl回調結構
struct nfnl_callback
{
int (*call)(struct sock *nl, struct sk_buff *skb,
struct nlmsghdr *nlh, struct nfattr *cda[], int *errp);
u_int16_t attr_count; /* number of nfattr's */
};
// netfilter netlink子系統結構
struct nfnetlink_subsystem
{
const char *name;
__u8 subsys_id; /* nfnetlink subsystem ID */
__u8 cb_count; /* number of callbacks */
struct nfnl_callback *cb; /* callback for individual types */
};
2.1.2
/*libnfnetlink.h */
/* nfnl是netfilter netlink的縮寫 */
// netfilter的netlink消息頭結構
struct nfnlhdr {
struct nlmsghdr nlh;
struct nfgenmsg nfmsg;
};
// netfilter的netlink回調函數
struct nfnl_callback {
int (*call)(struct nlmsghdr *nlh, struct nfattr *nfa[], void *data);
void *data;
u_int16_t attr_count;
};
2.1.3
/* libnfnetlink.c */
// netfilter netlink子系統的handle結構
struct nfnl_subsys_handle {
struct nfnl_handle *nfnlh;
u_int32_t subscriptions;
u_int8_t subsys_id;
u_int8_t cb_count;
struct nfnl_callback *cb; /* array of callbacks */
};
#define NFNL_MAX_SUBSYS 16 /* enough for now */
// netfilter netlink的handle結構
struct nfnl_handle {
int fd; // socket
struct sockaddr_nl local; // 本地netlink socket信息
struct sockaddr_nl peer; // 對端的netlink socket信息
u_int32_t subscriptions;
u_int32_t seq; // 序號
u_int32_t dump;
struct nlmsghdr *last_nlhdr; // 上一個消息
struct nfnl_subsys_handle subsys[NFNL_MAX_SUBSYS+1]; // 各子系統的handle
};
/*linux_nfnetlink_queue.h */
/* nfqnl是netfilter queue netlink的縮寫 */
// nfqueue_netlink消息類型
enum nfqnl_msg_types {
NFQNL_MSG_PACKET, /* packet from kernel to userspace */
NFQNL_MSG_VERDICT, /* verdict from userspace to kernel */
NFQNL_MSG_CONFIG, /* connect to a particular queue */
NFQNL_MSG_MAX
};
// nfqueue_netlink消息包頭結構
struct nfqnl_msg_packet_hdr {
u_int32_t packet_id; /* unique ID of packet in queue */
u_int16_t hw_protocol; /* hw protocol (network order) */
u_int8_t hook; /* netfilter hook */
} __attribute__ ((packed));
// nfqueue_netlink消息包頭的硬件地址信息
struct nfqnl_msg_packet_hw {
u_int16_t hw_addrlen;
u_int16_t _pad;
u_int8_t hw_addr[8];
} __attribute__ ((packed));
// nfqueue_netlink消息數據包時間戳，都是64位的
struct nfqnl_msg_packet_timestamp {
aligned_u64 sec;
aligned_u64 usec;
} __attribute__ ((packed));
// nfqueue_netlink屬性類型
enum nfqnl_attr_type {
NFQA_UNSPEC,
NFQA_PACKET_HDR,
NFQA_VERDICT_HDR, /* nfqnl_msg_verdict_hrd */
NFQA_MARK, /* u_int32_t nfmark */
NFQA_TIMESTAMP, /* nfqnl_msg_packet_timestamp */
NFQA_IFINDEX_INDEV, /* u_int32_t ifindex */
NFQA_IFINDEX_OUTDEV, /* u_int32_t ifindex */
NFQA_IFINDEX_PHYSINDEV, /* u_int32_t ifindex */
NFQA_IFINDEX_PHYSOUTDEV, /* u_int32_t ifindex */
NFQA_HWADDR, /* nfqnl_msg_packet_hw */
NFQA_PAYLOAD, /* opaque data payload */
__NFQA_MAX
};
#define NFQA_MAX (__NFQA_MAX - 1)
// nfqueue_netlink數據包的裁定結果頭結構
struct nfqnl_msg_verdict_hdr {
u_int32_t verdict;
u_int32_t id;
} __attribute__ ((packed));
// nfqueue_netlink消息的配置命令類型
enum nfqnl_msg_config_cmds {
NFQNL_CFG_CMD_NONE,
NFQNL_CFG_CMD_BIND, // 和隊列綁定
NFQNL_CFG_CMD_UNBIND, // 取消和隊列的綁定
NFQNL_CFG_CMD_PF_BIND, // 和協議族綁定
NFQNL_CFG_CMD_PF_UNBIND, // 取消和協議族的綁定
};
// nfqueue_netlink消息的配置命令結構
struct nfqnl_msg_config_cmd {
u_int8_t command; /* nfqnl_msg_config_cmds */
u_int8_t _pad;
u_int16_t pf; /* AF_xxx for PF_[UN]BIND */
} __attribute__ ((packed));
// nfqueue_netlink消息的配置模式類型
enum nfqnl_config_mode {
NFQNL_COPY_NONE, // 不拷貝
NFQNL_COPY_META, // 只拷貝頭部信息
NFQNL_COPY_PACKET, // 拷貝整個數據包
};
// nfqueue_netlink消息的配置結構
struct nfqnl_msg_config_params {
u_int32_t copy_range;
u_int8_t copy_mode; /* enum nfqnl_config_mode */
} __attribute__ ((packed));
// nfqueue_netlink屬性類型
enum nfqnl_attr_config {
NFQA_CFG_UNSPEC,
NFQA_CFG_CMD, /* nfqnl_msg_config_cmd */
NFQA_CFG_PARAMS, /* nfqnl_msg_config_params */
__NFQA_CFG_MAX
};
#define NFQA_CFG_MAX (__NFQA_CFG_MAX-1)
2.1.4
/* libnetfilter_queue.c */
// netfilter queue的handle結構
struct nfq_handle
{
struct nfnl_handle *nfnlh; // nf netlink handle
struct nfnl_subsys_handle *nfnlssh; // 子系統的handle
struct nfq_q_handle *qh_list; //
};
// netfilter queue的節點隊列結構
struct nfq_q_handle
{
// 下一個節點
struct nfq_q_handle *next;
// 該隊列的handle
struct nfq_handle *h;
// id號，每個queue有唯一ID，一共可支持65536個queue
u_int16_t id;
// typedef int nfq_callback(struct nfq_q_handle *gh, struct nfgenmsg *nfmsg,
// struct nfq_data *nfad, void *data);
// nf queue的回調函數
nfq_callback *cb;
// nf queue的回調函數輸入數據指針
void *data;
};
struct nfq_data {
struct nfattr **data;
};
2.2 接口函數聲明
2.2.1 libnfnetlink
實際上這些函數和宏都是被netfilter_queue的接口函數所包裝，一般用戶應用程序中不用直接調用這些函數或宏。
/* libnfnetlink.h */
extern int nfnl_fd(struct nfnl_handle *h);
/* get a new library handle */
extern struct nfnl_handle *nfnl_open(void);
extern int nfnl_close(struct nfnl_handle *);
extern struct nfnl_subsys_handle *nfnl_subsys_open(struct nfnl_handle *,
u_int8_t, u_int8_t,
unsigned int);
extern void nfnl_subsys_close(struct nfnl_subsys_handle *);
/* sending of data */
extern int nfnl_send(struct nfnl_handle *, struct nlmsghdr *);
extern int nfnl_sendmsg(const struct nfnl_handle *, const struct msghdr *msg,
unsigned int flags);
extern int nfnl_sendiov(const struct nfnl_handle *nfnlh,
const struct iovec *iov, unsigned int num,
unsigned int flags);
extern void nfnl_fill_hdr(struct nfnl_subsys_handle *, struct nlmsghdr *,
unsigned int, u_int8_t, u_int16_t, u_int16_t,
u_int16_t);
extern int nfnl_talk(struct nfnl_handle *, struct nlmsghdr *, pid_t,
unsigned, struct nlmsghdr *,
int (*)(struct sockaddr_nl *, struct nlmsghdr *, void *),
void *);
/* simple challenge/response */
extern int nfnl_listen(struct nfnl_handle *,
int (*)(struct sockaddr_nl *, struct nlmsghdr *, void *),
void *);
/* receiving */
extern ssize_t nfnl_recv(const struct nfnl_handle *h, unsigned char *buf, size_t len);
extern int nfnl_callback_register(struct nfnl_subsys_handle *,
u_int8_t type, struct nfnl_callback *cb);
extern int nfnl_callback_unregister(struct nfnl_subsys_handle *, u_int8_t type);
extern int nfnl_handle_packet(struct nfnl_handle *, char *buf, int len);
/* parsing */
extern struct nfattr *nfnl_parse_hdr(const struct nfnl_handle *nfnlh,
const struct nlmsghdr *nlh,
struct nfgenmsg **genmsg);
extern int nfnl_check_attributes(const struct nfnl_handle *nfnlh,
const struct nlmsghdr *nlh,
struct nfattr *tb[]);
extern struct nlmsghdr *nfnl_get_msg_first(struct nfnl_handle *h,
const unsigned char *buf,
size_t len);
extern struct nlmsghdr *nfnl_get_msg_next(struct nfnl_handle *h,
const unsigned char *buf,
size_t len);
#define nfnl_attr_present(tb, attr) /
(tb[attr-1])
#define nfnl_get_data(tb, attr, type) /
({ type __ret = 0; /
if (tb[attr-1]) /
__ret = *(type *)NFA_DATA(tb[attr-1]); /
__ret; /
})
#define nfnl_get_pointer_to_data(tb, attr, type) /
({ type *__ret = NULL; /
if (tb[attr-1]) /
__ret = NFA_DATA(tb[attr-1]); /
__ret; /
})
/* nfnl attribute handling functions */
extern int nfnl_addattr_l(struct nlmsghdr *, int, int, void *, int);
extern int nfnl_addattr16(struct nlmsghdr *, int, int, u_int16_t);
extern int nfnl_addattr32(struct nlmsghdr *, int, int, u_int32_t);
extern int nfnl_nfa_addattr_l(struct nfattr *, int, int, void *, int);
extern int nfnl_nfa_addattr16(struct nfattr *, int, int, u_int16_t);
extern int nfnl_nfa_addattr32(struct nfattr *, int, int, u_int32_t);
extern int nfnl_parse_attr(struct nfattr **, int, struct nfattr *, int);
#define nfnl_parse_nested(tb, max, nfa) /
nfnl_parse_attr((tb), (max), NFA_DATA((nfa)), NFA_PAYLOAD((nfa)))
#define nfnl_nest(nlh, bufsize, type) /
({ struct nfattr *__start = NLMSG_TAIL(nlh); /
nfnl_addattr_l(nlh, bufsize, (NFNL_NFA_NEST | type), NULL, 0); /
__start; })
#define nfnl_nest_end(nlh, tail) /
({ (tail)->nfa_len = (void *) NLMSG_TAIL(nlh) - (void *) tail; })
extern void nfnl_build_nfa_iovec(struct iovec *iov, struct nfattr *nfa,
u_int16_t type, u_int32_t len,
unsigned char *val);
extern unsigned int nfnl_rcvbufsiz(struct nfnl_handle *h, unsigned int size);
extern void nfnl_dump_packet(struct nlmsghdr *, int, char *);
2.2.2 libnetfilter_queue
/* libnetfilter_queue.h */
// 打開一個nfqueue的handle,返回NULL表示失敗
extern struct nfq_handle *nfq_open(void);
// 打開nf netlink handle對應的nfqueue
extern struct nfq_handle *nfq_open_nfnl(struct nfnl_handle *nfnlh);
// 關閉nfqueue
extern int nfq_close(struct nfq_handle *h);
// 對nfqueue綁定協議族
extern int nfq_bind_pf(struct nfq_handle *h, u_int16_t pf);
// 對nfqueue取消協議族綁定
extern int nfq_unbind_pf(struct nfq_handle *h, u_int16_t pf);
// 建立具體的queue的handle,由num指定queue的序號, 返回NULL失敗
extern struct nfq_q_handle *nfq_create_queue(struct nfq_handle *h,
u_int16_t num,
nfq_callback *cb,
void *data);
// 釋放隊列
extern int nfq_destroy_queue(struct nfq_q_handle *qh);
// 處理隊列的數據包
extern int nfq_handle_packet(struct nfq_handle *h, char *buf, int len);
// 設置queue handle的數據拷貝模式
extern int nfq_set_mode(struct nfq_q_handle *qh,
u_int8_t mode, unsigned int len);
// 設置數據包的判定結果, id用於指定具體的包, buf和data_len用於傳遞修改後的數據
extern int nfq_set_verdict(struct nfq_q_handle *qh,
u_int32_t id,
u_int32_t verdict,
u_int32_t data_len,
unsigned char *buf);
// 設置數據包的mark值
extern int nfq_set_verdict_mark(struct nfq_q_handle *qh,
u_int32_t id,
u_int32_t verdict,
u_int32_t mark,
u_int32_t datalen,
unsigned char *buf);
/* message parsing function */
// 從緩衝區原始數據中返回消息頭結構
extern struct nfqnl_msg_packet_hdr *
nfq_get_msg_packet_hdr(struct nfq_data *nfad);
// 獲取數據的mark信息
extern u_int32_t nfq_get_nfmark(struct nfq_data *nfad);
extern int nfq_get_timestamp(struct nfq_data *nfad, struct timeval *tv);
/* return 0 if not set */
// 返回數據包進入網卡的索引號
extern u_int32_t nfq_get_indev(struct nfq_data *nfad);
// 返回數據包進入的物理網卡的索引號
extern u_int32_t nfq_get_physindev(struct nfq_data *nfad);
// 返回數據包發出網卡的索引號
extern u_int32_t nfq_get_outdev(struct nfq_data *nfad);
// 返回數據包發出的物理網卡的索引號
extern u_int32_t nfq_get_physoutdev(struct nfq_data *nfad);
// 返回數據包硬件地址
extern struct nfqnl_msg_packet_hw *nfq_get_packet_hw(struct nfq_data *nfad);
/* return -1 if problem, length otherwise */
// 獲取數據包中載荷地址
extern int nfq_get_payload(struct nfq_data *nfad, char **data);
2.3 netfilter queue接口函數的實現
/* libnetfilter_queue.c */
// 刪除一個queue節點
// 各個nfq_q_handle結構都是其nfq_handle中的qh_list鏈表中的一個節點
// 所以刪除節點就是將其從鏈表中移出即可，該函數不進行內存釋放操作
// 結構可表示如下：
// nfq_handle -> qh_list
// ^ |
// | V
// | nfq_q_handle -> nfq_q_handle -> ...
// | | |
// |______________|_______________|_________________
static void del_qh(struct nfq_q_handle *qh)
{
struct nfq_q_handle *cur_qh, *prev_qh = NULL;
for (cur_qh = qh->h->qh_list; cur_qh; cur_qh = cur_qh->next) {
if (cur_qh == qh) {
if (prev_qh)
prev_qh->next = qh->next;
else
qh->h->qh_list = qh->next;
return;
}
prev_qh = cur_qh;
}
}
// 把一個nfq_q_handle結構添加到鏈表中
static void add_qh(struct nfq_q_handle *qh)
{
qh->next = qh->h->qh_list;
qh->h->qh_list = qh;
}
// 根據ID號找nfq_q_handle節點
static struct nfq_q_handle *find_qh(struct nfq_handle *h, u_int16_t id)
{
struct nfq_q_handle *qh;
for (qh = h->qh_list; qh; qh = qh->next) {
if (qh->id == id)
return qh;
}
return NULL;
}
/* build a NFQNL_MSG_CONFIG message */
// 向netlink socket發送配置信息，該函數是static的，外部函數不可見
static int
__build_send_cfg_msg(struct nfq_handle *h, u_int8_t command,
u_int16_t queuenum, u_int16_t pf)
{
char buf[NFNL_HEADER_LEN
+NFA_LENGTH(sizeof(struct nfqnl_msg_config_cmd))];
struct nfqnl_msg_config_cmd cmd;
struct nlmsghdr *nmh = (struct nlmsghdr *) buf;
nfnl_fill_hdr(h->nfnlssh, nmh, 0, AF_UNSPEC, queuenum,
NFQNL_MSG_CONFIG, NLM_F_REQUEST|NLM_F_ACK);
cmd.command = command;
cmd.pf = htons(pf);
nfnl_addattr_l(nmh, sizeof(buf), NFQA_CFG_CMD, &cmd, sizeof(cmd));
return nfnl_talk(h->nfnlh, nmh, 0, 0, NULL, NULL, NULL);
}
// 接收netlink數據包，該函數是也static的，外部函數不可見
static int __nfq_rcv_pkt(struct nlmsghdr *nlh, struct nfattr *nfa[],
void *data)
{
struct nfgenmsg *nfmsg = NLMSG_DATA(nlh);
struct nfq_handle *h = data;
u_int16_t queue_num = ntohs(nfmsg->res_id);
// 根據ID找到nfq_q_handle
struct nfq_q_handle *qh = find_qh(h, queue_num);
struct nfq_data nfqa;
if (!qh)
return -ENODEV;
if (!qh->cb)
return -ENODEV;
nfqa.data = nfa;
// 調用nfq_q_handle的回調函數
return qh->cb(qh, nfmsg, &nfqa, qh->data);
}
// 固定的回調結構
static struct nfnl_callback pkt_cb = {
.call = &__nfq_rcv_pkt,
.attr_count = NFQA_MAX,
};
/* public interface */
// 返回nfq_handle的netlink handle
struct nfnl_handle *nfq_nfnlh(struct nfq_handle *h)
{
return h->nfnlh;
}
// 返回nfq_handle的netlink handle的netlink套接字
int nfq_fd(struct nfq_handle *h)
{
return nfnl_fd(nfq_nfnlh(h));
}
struct nfq_handle *nfq_open(void)
{
// 先打開netlink handle
struct nfnl_handle *nfnlh = nfnl_open();
struct nfq_handle *qh;
if (!nfnlh)
return NULL;
// 再調用nfq_open_nfnl()打開nf queue handle
qh = nfq_open_nfnl(nfnlh);
if (!qh)
nfnl_close(nfnlh);
return qh;
}
// 已存在netlink handle時打開nfq_handle
struct nfq_handle *nfq_open_nfnl(struct nfnl_handle *nfnlh)
{
struct nfq_handle *h;
int err;
// 分配內存
h = malloc(sizeof(*h));
if (!h)
return NULL;
memset(h, 0, sizeof(*h));
// 把nfq_handle和netlink handle連接起來
h->nfnlh = nfnlh;
// 打開NFNL_SUBSYS_QUEUE類型的子系統
h->nfnlssh = nfnl_subsys_open(h->nfnlh, NFNL_SUBSYS_QUEUE,
NFQNL_MSG_MAX, 0);
if (!h->nfnlssh) {
/* FIXME: nfq_errno */
goto out_free;
}
// 登記回調處理函數
pkt_cb.data = h;
err = nfnl_callback_register(h->nfnlssh, NFQNL_MSG_PACKET, &pkt_cb);
if (err < 0) {
nfq_errno = err;
goto out_close;
}
return h;
out_close:
nfnl_subsys_close(h->nfnlssh);
out_free:
free(h);
return NULL;
}
int nfq_close(struct nfq_handle *h)
{
int ret;
// 關閉子系統
nfnl_subsys_close(h->nfnlssh);
// 關閉netlink handle
ret = nfnl_close(h->nfnlh);
if (ret == 0)
free(h);
return ret;
}
/* bind nf_queue from a specific protocol family */
// 以下函數均是調用__build_send_cfg_msg()函數向內核發送消息命令
int nfq_bind_pf(struct nfq_handle *h, u_int16_t pf)
{
return __build_send_cfg_msg(h, NFQNL_CFG_CMD_PF_BIND, 0, pf);
}
/* unbind nf_queue from a specific protocol family */
int nfq_unbind_pf(struct nfq_handle *h, u_int16_t pf)
{
return __build_send_cfg_msg(h, NFQNL_CFG_CMD_PF_UNBIND, 0, pf);
}
/* bind this socket to a specific queue number */
// 生成一個號碼爲num的隊列
struct nfq_q_handle *nfq_create_queue(struct nfq_handle *h,
u_int16_t num,
nfq_callback *cb,
void *data)
{
int ret;
struct nfq_q_handle *qh;
if (find_qh(h, num))
return NULL;
// 分配queue節點空間, 設置相應參數
qh = malloc(sizeof(*qh));
memset(qh, 0, sizeof(*qh));
qh->h = h;
qh->id = num;
qh->cb = cb;
qh->data = data;
ret = __build_send_cfg_msg(h, NFQNL_CFG_CMD_BIND, num, 0);
if (ret < 0) {
nfq_errno = ret;
free(qh);
return NULL;
}
// 添加到隊列中
add_qh(qh);
return qh;
}
/* unbind this socket from a specific queue number */
// 釋放隊列
int nfq_destroy_queue(struct nfq_q_handle *qh)
{
int ret = __build_send_cfg_msg(qh->h, NFQNL_CFG_CMD_UNBIND, qh->id, 0);
if (ret == 0) {
del_qh(qh);
free(qh);
}
return ret;
}
int nfq_handle_packet(struct nfq_handle *h, char *buf, int len)
{
// 實際就是netlink處理包
return nfnl_handle_packet(h->nfnlh, buf, len);
}
int nfq_set_mode(struct nfq_q_handle *qh,
u_int8_t mode, u_int32_t range)
{
char buf[NFNL_HEADER_LEN
+NFA_LENGTH(sizeof(struct nfqnl_msg_config_params))];
struct nfqnl_msg_config_params params;
struct nlmsghdr *nmh = (struct nlmsghdr *) buf;
nfnl_fill_hdr(qh->h->nfnlssh, nmh, 0, AF_UNSPEC, qh->id,
NFQNL_MSG_CONFIG, NLM_F_REQUEST|NLM_F_ACK);
params.copy_range = htonl(range);
params.copy_mode = mode;
nfnl_addattr_l(nmh, sizeof(buf), NFQA_CFG_PARAMS, ¶ms,
sizeof(params));
return nfnl_talk(qh->h->nfnlh, nmh, 0, 0, NULL, NULL, NULL);
}
static int __set_verdict(struct nfq_q_handle *qh, u_int32_t id,
u_int32_t verdict, u_int32_t mark, int set_mark,
u_int32_t data_len, unsigned char *data)
{
struct nfqnl_msg_verdict_hdr vh;
char buf[NFNL_HEADER_LEN
+NFA_LENGTH(sizeof(mark))
+NFA_LENGTH(sizeof(vh))];
struct nlmsghdr *nmh = (struct nlmsghdr *) buf;
struct iovec iov[3];
int nvecs;
/* This must be declared here (and not inside the data
* handling block) because the iovec points to this. */
struct nfattr data_attr;
memset(iov, 0, sizeof(iov));
// 設置裁定結果頭
vh.verdict = htonl(verdict);
vh.id = htonl(id);
nfnl_fill_hdr(qh->h->nfnlssh, nmh, 0, AF_UNSPEC, qh->id,
NFQNL_MSG_VERDICT, NLM_F_REQUEST);
/* add verdict header */
nfnl_addattr_l(nmh, sizeof(buf), NFQA_VERDICT_HDR, &vh, sizeof(vh));
// 設置數據包的mark值
if (set_mark)
nfnl_addattr32(nmh, sizeof(buf), NFQA_MARK, mark);
iov[0].iov_base = nmh;
iov[0].iov_len = NLMSG_TAIL(nmh) - (void *)nmh;
nvecs = 1;
if (data_len) {
// 如果數據進行修改要傳回內核,相應將數據添加到要發送到內核的數據向量中
nfnl_build_nfa_iovec(&iov[1], &data_attr, NFQA_PAYLOAD,
data_len, data);
nvecs += 2;
/* Add the length of the appended data to the message
* header. The size of the attribute is given in the
* nfa_len field and is set in the nfnl_build_nfa_iovec()
* function. */
nmh->nlmsg_len += data_attr.nfa_len;
}
// 向內核發送數據向量
return nfnl_sendiov(qh->h->nfnlh, iov, nvecs, 0);
}
int nfq_set_verdict(struct nfq_q_handle *qh, u_int32_t id,
u_int32_t verdict, u_int32_t data_len,
unsigned char *buf)
{
return __set_verdict(qh, id, verdict, 0, 0, data_len, buf);
}
int nfq_set_verdict_mark(struct nfq_q_handle *qh, u_int32_t id,
u_int32_t verdict, u_int32_t mark,
u_int32_t datalen, unsigned char *buf)
{
return __set_verdict(qh, id, verdict, mark, 1, datalen, buf);
}
/*************************************************************
* Message parsing functions
*************************************************************/
// 以下函數均是調用nfnl_get_pointer_to_data()和nfnl_get_data()函數獲取
// 指定數據
struct nfqnl_msg_packet_hdr *nfq_get_msg_packet_hdr(struct nfq_data *nfad)
{
return nfnl_get_pointer_to_data(nfad->data, NFQA_PACKET_HDR,
struct nfqnl_msg_packet_hdr);
}
uint32_t nfq_get_nfmark(struct nfq_data *nfad)
{
return ntohl(nfnl_get_data(nfad->data, NFQA_MARK, u_int32_t));
}
int nfq_get_timestamp(struct nfq_data *nfad, struct timeval *tv)
{
struct nfqnl_msg_packet_timestamp *qpt;
qpt = nfnl_get_pointer_to_data(nfad->data, NFQA_TIMESTAMP,
struct nfqnl_msg_packet_timestamp);
if (!qpt)
return -1;
tv->tv_sec = __be64_to_cpu(qpt->sec);
tv->tv_usec = __be64_to_cpu(qpt->usec);
return 0;
}
/* all nfq_get_*dev() functions return 0 if not set, since
linux學習 only allows
* ifindex >= 1, see net/core/dev.c:2600 (in 2.6.13.1) */
u_int32_t nfq_get_indev(struct nfq_data *nfad)
{
return ntohl(nfnl_get_data(nfad->data, NFQA_IFINDEX_INDEV, u_int32_t));
}
u_int32_t nfq_get_physindev(struct nfq_data *nfad)
{
return ntohl(nfnl_get_data(nfad->data, NFQA_IFINDEX_PHYSINDEV, u_int32_t));
}
u_int32_t nfq_get_outdev(struct nfq_data *nfad)
{
return ntohl(nfnl_get_data(nfad->data, NFQA_IFINDEX_OUTDEV, u_int32_t));
}
u_int32_t nfq_get_physoutdev(struct nfq_data *nfad)
{
return ntohl(nfnl_get_data(nfad->data, NFQA_IFINDEX_PHYSOUTDEV, u_int32_t));
}
struct nfqnl_msg_packet_hw *nfq_get_packet_hw(struct nfq_data *nfad)
{
return nfnl_get_pointer_to_data(nfad->data, NFQA_HWADDR,
struct nfqnl_msg_packet_hw);
}
int nfq_get_payload(struct nfq_data *nfad, char **data)
{
*data = nfnl_get_pointer_to_data(nfad->data, NFQA_PAYLOAD, char);
if (*data)
return NFA_PAYLOAD(nfad->data[NFQA_PAYLOAD-1]);
return -1;
}
2.4 程序實例
/* nfqnl_test.c */
#include <stdio.h>
#include <stdlib.h>
#include <unistd.h>
#include <netinet/in.h>
#include <linux/netfilter.h> /* for NF_ACCEPT */
#include <libnetfilter_queue/libnetfilter_queue.h>
/* returns packet id */
// 對數據包的處理函數, 本示例僅用於打印數據包的信息
static u_int32_t print_pkt (struct nfq_data *tb)
{
int id = 0;
struct nfqnl_msg_packet_hdr *ph;
u_int32_t mark,ifi;
int ret;
char *data;
// 提取數據包頭信息,包括id,協議和hook點信息
ph = nfq_get_msg_packet_hdr(tb);
if (ph){
id = ntohl(ph->packet_id);
printf("hw_protocol=0x%04x hook=%u id=%u ",
ntohs(ph->hw_protocol), ph->hook, id);
}
// 獲取數據包的mark值, 也就是內核skb的nfmark值
mark = nfq_get_nfmark(tb);
if (mark)
printf("mark=%u ", mark);
// 獲取數據包的進入網卡的索引號
ifi = nfq_get_indev(tb);
if (ifi)
printf("indev=%u ", ifi);
// 獲取數據包的發出網卡的索引號
ifi = nfq_get_outdev(tb);
if (ifi)
printf("outdev=%u ", ifi);
// 獲取數據包載荷,data指針指向載荷,從實際的IP頭開始
ret = nfq_get_payload(tb, &data);
if (ret >= 0)
printf("payload_len=%d ", ret);
fputc('/n', stdout);
return id;
}
// 回調函數定義, 基本結構是先處理包,然後返回裁定
static int cb(struct nfq_q_handle *qh, struct nfgenmsg *nfmsg,
struct nfq_data *nfa, void *data)
{
// 數據包處理
u_int32_t id = print_pkt(nfa);
printf("entering callback/n");
// 設置裁定
return nfq_set_verdict(qh, id, NF_ACCEPT, 0, NULL);
}
int main(int argc, char **argv)
{
struct nfq_handle *h;
struct nfq_q_handle *qh;
struct nfnl_handle *nh;
int fd;
int rv;
char buf[4096];
printf("opening library handle/n");
// 打開nfq_handle
h = nfq_open();
if (!h) {
fprintf(stderr, "error during nfq_open()/n");
exit(1);
}
printf("unbinding existing nf_queue handler for AF_INET (if any)/n");
// 先解開和AF_INET的綁定
if (nfq_unbind_pf(h, AF_INET) < 0) {
fprintf(stderr, "error during nfq_unbind_pf()/n");
exit(1);
}
printf("binding nfnetlink_queue as nf_queue handler for AF_INET/n");
// 綁定到AF_INET
if (nfq_bind_pf(h, AF_INET) < 0) {
fprintf(stderr, "error during nfq_bind_pf()/n");
exit(1);
}
printf("binding this socket to queue '0'/n");
// 建立nfq_q_handle, 號碼是0, 回調函數是cb
// 可建立多個queue,用不同的號碼區分即可
qh = nfq_create_queue(h, 0, &cb, NULL);
if (!qh) {
fprintf(stderr, "error during nfq_create_queue()/n");
exit(1);
}
printf("setting copy_packet mode/n");
// 設置數據拷貝模式, 全包拷貝
if (nfq_set_mode(qh, NFQNL_COPY_PACKET, 0xffff) < 0) {
fprintf(stderr, "can't set packet_copy mode/n");
exit(1);
}
nh = nfq_nfnlh(h);
fd = nfnl_fd(nh);
// 從netlink套接字接收數據
while ((rv = recv(fd, buf, sizeof(buf), 0)) && rv >= 0) {
printf("pkt received/n");
// 處理數據,最終會調用到相應的回調函數
nfq_handle_packet(h, buf, rv);
}
printf("unbinding from queue 0/n");
// 釋放隊列
nfq_destroy_queue(qh);
#ifdef INSANE
/* normally, applications SHOULD NOT issue this command, since
* it detaches other programs/sockets from AF_INET, too ! */
printf("unbinding from AF_INET/n");
nfq_unbind_pf(h, AF_INET);
#endif
printf("closing library handle/n");
// 關閉nfq_handle
nfq_close(h);
exit(0);
}
2.4 包裝libipq
可用netlink_queue包裝libipq，ipq就相當於是號碼爲0的一個nfqueue而已：
/* libipq_compat.c */
struct ipq_handle *ipq_create_handle(u_int32_t flags, u_int32_t protocol)
{
int status;
struct ipq_handle *h;
h = (struct ipq_handle *)malloc(sizeof(struct ipq_handle));
if (h == NULL) {
ipq_errno = IPQ_ERR_HANDLE;
return NULL;
}
memset(h, 0, sizeof(struct ipq_handle));
// 打開ipq的nfqueue handle
h->nfqnlh = nfq_open();
if (!h->nfqnlh) {
ipq_errno = IPQ_ERR_SOCKET;
goto err_free;
}
// 綁定到PF_INET或PF_INET6
if (protocol == PF_INET)
status = nfq_bind_pf(h->nfqnlh, PF_INET);
else if (protocol == PF_INET6)
status = nfq_bind_pf(h->nfqnlh, PF_INET6);
else {
ipq_errno = IPQ_ERR_PROTOCOL;
goto err_close;
}
h->family = protocol;
if (status < 0) {
ipq_errno = IPQ_ERR_BIND;
goto err_close;
}
// 按號碼0建立queue，無回調函數，數據包由ipq直接讀後處理
h->qh = nfq_create_queue(h->nfqnlh, 0, NULL, NULL);
if (!h->qh) {
ipq_errno = IPQ_ERR_BIND;
goto err_close;
}
return h;
err_close:
nfq_close(h->nfqnlh);
err_free:
free(h);
return NULL;
}
/*
* No error condition is checked here at this stage, but it may happen
* if/when reliable messaging is implemented.
*/
int ipq_destroy_handle(struct ipq_handle *h)
{
if (h) {
nfq_close(h->nfqnlh);
free(h);
}
return 0;
}
int ipq_set_mode(const struct ipq_handle *h,
u_int8_t mode, size_t range)
{
return nfq_set_mode(h->qh, mode, range);
}
/*
* timeout is in microseconds (1 second is 1000000 (1 million) microseconds)
*
*/
// ipq_read包裝得有點疑問，實際沒進行接收操作，需要顯式的recv接收數據包
// 現在的ipq_read只是對接收的數據進行解析
ssize_t ipq_read(const struct ipq_handle *h,
unsigned char *buf, size_t len, int timeout)
{
struct nfattr *tb[NFQA_MAX];
struct nlmsghdr *nlh = (struct nlmsghdr *)buf;
struct nfgenmsg *msg = NULL;
struct nfattr *nfa;
//return ipq_netlink_recvfrom(h, buf, len, timeout);
/* This really sucks. We have to copy the whole packet
* in order to build a data structure that is compatible to
* the old ipq interface... */
nfa = nfnl_parse_hdr(nfq_nfnlh(h->nfqnlh), nlh, &msg);
if (!msg || !nfa)
return 0;
if (msg->nfgen_family != h->family)
return 0;
nfnl_parse_attr(tb, NFQA_MAX, nfa, 0xffff);
return 0;
}
int ipq_message_type(const unsigned char *buf)
{
return ((struct nlmsghdr*)buf)->nlmsg_type;
}
int ipq_get_msgerr(const unsigned char *buf)
{
struct nlmsghdr *h = (struct nlmsghdr *)buf;
struct nlmsgerr *err = (struct nlmsgerr*)NLMSG_DATA(h);
return -err->error;
}
ipq_packet_msg_t *ipq_get_packet(const unsigned char *buf)
{
return NLMSG_DATA((struct nlmsghdr *)(buf));
}
int ipq_set_verdict(const struct ipq_handle *h,
ipq_id_t id,
unsigned int verdict,
size_t data_len,
unsigned char *buf)
{
return nfq_set_verdict(h->qh, id, verdict, data_len, buf);
}
/* Not implemented yet */
int ipq_ctl(const struct ipq_handle *h, int request, ...)
{
return 1;
}
char *ipq_errstr(void)
{
return ipq_strerror(ipq_errno);
}
void ipq_perror(const char *s)
{
if (s)
fputs(s, stderr);
else
fputs("ERROR", stderr);
if (ipq_errno)
fprintf(stderr, ": %s", ipq_errstr());
if (errno)
fprintf(stderr, ": %s", strerror(errno));
fputc('/n', stderr);
}
...... 待續 ......
