ip_vs.h

Linux Kernel 2.6.9 for OMAP1710

/*
 *      IP Virtual Server
 *      data structure and functionality definitions
 */

#ifndef _IP_VS_H
#define _IP_VS_H

#include <asm/types.h>		/* For __uXX types */

#define IP_VS_VERSION_CODE	0x010200
#define NVERSION(version)			\
	(version >> 16) & 0xFF,			\
	(version >> 8) & 0xFF,			\
	version & 0xFF

/*
 *      Virtual Service Flags
 */
#define IP_VS_SVC_F_PERSISTENT	0x0001		/* persistent port */
#define IP_VS_SVC_F_HASHED	0x0002		/* hashed entry */

/*
 *      Destination Server Flags
 */
#define IP_VS_DEST_F_AVAILABLE	0x0001		/* server is available */
#define IP_VS_DEST_F_OVERLOAD	0x0002		/* server is overloaded */

/*
 *      IPVS sync daemon states
 */
#define IP_VS_STATE_NONE	0x0000		/* daemon is stopped */
#define IP_VS_STATE_MASTER	0x0001		/* started as master */
#define IP_VS_STATE_BACKUP	0x0002		/* started as backup */

/*
 *      IPVS socket options
 */
#define IP_VS_BASE_CTL		(64+1024+64)		/* base */

#define IP_VS_SO_SET_NONE	IP_VS_BASE_CTL		/* just peek */
#define IP_VS_SO_SET_INSERT	(IP_VS_BASE_CTL+1)
#define IP_VS_SO_SET_ADD	(IP_VS_BASE_CTL+2)
#define IP_VS_SO_SET_EDIT	(IP_VS_BASE_CTL+3)
#define IP_VS_SO_SET_DEL	(IP_VS_BASE_CTL+4)
#define IP_VS_SO_SET_FLUSH	(IP_VS_BASE_CTL+5)
#define IP_VS_SO_SET_LIST	(IP_VS_BASE_CTL+6)
#define IP_VS_SO_SET_ADDDEST	(IP_VS_BASE_CTL+7)
#define IP_VS_SO_SET_DELDEST	(IP_VS_BASE_CTL+8)
#define IP_VS_SO_SET_EDITDEST	(IP_VS_BASE_CTL+9)
#define IP_VS_SO_SET_TIMEOUT	(IP_VS_BASE_CTL+10)
#define IP_VS_SO_SET_STARTDAEMON (IP_VS_BASE_CTL+11)
#define IP_VS_SO_SET_STOPDAEMON (IP_VS_BASE_CTL+12)
#define IP_VS_SO_SET_RESTORE    (IP_VS_BASE_CTL+13)
#define IP_VS_SO_SET_SAVE       (IP_VS_BASE_CTL+14)
#define IP_VS_SO_SET_ZERO	(IP_VS_BASE_CTL+15)
#define IP_VS_SO_SET_MAX	IP_VS_SO_SET_ZERO

#define IP_VS_SO_GET_VERSION	IP_VS_BASE_CTL
#define IP_VS_SO_GET_INFO	(IP_VS_BASE_CTL+1)
#define IP_VS_SO_GET_SERVICES	(IP_VS_BASE_CTL+2)
#define IP_VS_SO_GET_SERVICE	(IP_VS_BASE_CTL+3)
#define IP_VS_SO_GET_DESTS	(IP_VS_BASE_CTL+4)
#define IP_VS_SO_GET_DEST	(IP_VS_BASE_CTL+5)	/* not used now */
#define IP_VS_SO_GET_TIMEOUT	(IP_VS_BASE_CTL+6)
#define IP_VS_SO_GET_DAEMON	(IP_VS_BASE_CTL+7)
#define IP_VS_SO_GET_MAX	IP_VS_SO_GET_DAEMON


/*
 *      IPVS Connection Flags
 */
#define IP_VS_CONN_F_FWD_MASK	0x0007		/* mask for the fwd methods */
#define IP_VS_CONN_F_MASQ	0x0000		/* masquerading/NAT */
#define IP_VS_CONN_F_LOCALNODE	0x0001		/* local node */
#define IP_VS_CONN_F_TUNNEL	0x0002		/* tunneling */
#define IP_VS_CONN_F_DROUTE	0x0003		/* direct routing */
#define IP_VS_CONN_F_BYPASS	0x0004		/* cache bypass */
#define IP_VS_CONN_F_SYNC	0x0020		/* entry created by sync */
#define IP_VS_CONN_F_HASHED	0x0040		/* hashed entry */
#define IP_VS_CONN_F_NOOUTPUT	0x0080		/* no output packets */
#define IP_VS_CONN_F_INACTIVE	0x0100		/* not established */
#define IP_VS_CONN_F_OUT_SEQ	0x0200		/* must do output seq adjust */
#define IP_VS_CONN_F_IN_SEQ	0x0400		/* must do input seq adjust */
#define IP_VS_CONN_F_SEQ_MASK	0x0600		/* in/out sequence mask */
#define IP_VS_CONN_F_NO_CPORT	0x0800		/* no client port set yet */

/* Move it to better place one day, for now keep it unique */
#define NFC_IPVS_PROPERTY	0x10000

#define IP_VS_SCHEDNAME_MAXLEN	16
#define IP_VS_IFNAME_MAXLEN	16


/*
 *	The struct ip_vs_service_user and struct ip_vs_dest_user are
 *	used to set IPVS rules through setsockopt.
 */
struct ip_vs_service_user {
	/* virtual service addresses */
	u_int16_t		protocol;
	u_int32_t		addr;		/* virtual ip address */
	u_int16_t		port;
	u_int32_t		fwmark;		/* firwall mark of service */

	/* virtual service options */
	char			sched_name[IP_VS_SCHEDNAME_MAXLEN];
	unsigned		flags;		/* virtual service flags */
	unsigned		timeout;	/* persistent timeout in sec */
	u_int32_t		netmask;	/* persistent netmask */
};


struct ip_vs_dest_user {
	/* destination server address */
	u_int32_t		addr;
	u_int16_t		port;

	/* real server options */
	unsigned		conn_flags;	/* connection flags */
	int			weight;		/* destination weight */

	/* thresholds for active connections */
	u_int32_t		u_threshold;	/* upper threshold */
	u_int32_t		l_threshold;	/* lower threshold */
};


/*
 *	IPVS statistics object (for user space)
 */
struct ip_vs_stats_user
{
	__u32                   conns;          /* connections scheduled */
	__u32                   inpkts;         /* incoming packets */
	__u32                   outpkts;        /* outgoing packets */
	__u64                   inbytes;        /* incoming bytes */
	__u64                   outbytes;       /* outgoing bytes */

	__u32			cps;		/* current connection rate */
	__u32			inpps;		/* current in packet rate */
	__u32			outpps;		/* current out packet rate */
	__u32			inbps;		/* current in byte rate */
	__u32			outbps;		/* current out byte rate */
};


/* The argument to IP_VS_SO_GET_INFO */
struct ip_vs_getinfo {
	/* version number */
	unsigned int		version;

	/* size of connection hash table */
	unsigned int		size;

	/* number of virtual services */
	unsigned int		num_services;
};


/* The argument to IP_VS_SO_GET_SERVICE */
struct ip_vs_service_entry {
	/* which service: user fills in these */
	u_int16_t		protocol;
	u_int32_t		addr;		/* virtual address */
	u_int16_t		port;
	u_int32_t		fwmark;		/* firwall mark of service */

	/* service options */
	char			sched_name[IP_VS_SCHEDNAME_MAXLEN];
	unsigned		flags;          /* virtual service flags */
	unsigned		timeout;	/* persistent timeout */
	u_int32_t		netmask;	/* persistent netmask */

	/* number of real servers */
	unsigned int		num_dests;

	/* statistics */
	struct ip_vs_stats_user stats;
};


struct ip_vs_dest_entry {
	u_int32_t		addr;		/* destination address */
	u_int16_t		port;
	unsigned		conn_flags;	/* connection flags */
	int			weight;		/* destination weight */

	u_int32_t		u_threshold;	/* upper threshold */
	u_int32_t		l_threshold;	/* lower threshold */

	u_int32_t		activeconns;	/* active connections */
	u_int32_t		inactconns;	/* inactive connections */
	u_int32_t		persistconns;	/* persistent connections */

	/* statistics */
	struct ip_vs_stats_user stats;
};


/* The argument to IP_VS_SO_GET_DESTS */
struct ip_vs_get_dests {
	/* which service: user fills in these */
	u_int16_t		protocol;
	u_int32_t		addr;		/* virtual address */
	u_int16_t		port;
	u_int32_t		fwmark;		/* firwall mark of service */

	/* number of real servers */
	unsigned int		num_dests;

	/* the real servers */
	struct ip_vs_dest_entry	entrytable[0];
};


/* The argument to IP_VS_SO_GET_SERVICES */
struct ip_vs_get_services {
	/* number of virtual services */
	unsigned int		num_services;

	/* service table */
	struct ip_vs_service_entry entrytable[0];
};


/* The argument to IP_VS_SO_GET_TIMEOUT */
struct ip_vs_timeout_user {
	int			tcp_timeout;
	int			tcp_fin_timeout;
	int			udp_timeout;
};


/* The argument to IP_VS_SO_GET_DAEMON */
struct ip_vs_daemon_user {
	/* sync daemon state (master/backup) */
	int			state;

	/* multicast interface name */
	char			mcast_ifn[IP_VS_IFNAME_MAXLEN];

	/* SyncID we belong to */
	int			syncid;
};


#ifdef __KERNEL__

#include <linux/config.h>
#include <linux/list.h>                 /* for struct list_head */
#include <linux/spinlock.h>             /* for struct rwlock_t */
#include <linux/skbuff.h>               /* for struct sk_buff */
#include <linux/ip.h>                   /* for struct iphdr */
#include <asm/atomic.h>                 /* for struct atomic_t */
#include <linux/netdevice.h>		/* for struct neighbour */
#include <net/dst.h>			/* for struct dst_entry */
#include <net/tcp.h>
#include <net/udp.h>
#include <linux/compiler.h>


#ifdef CONFIG_IP_VS_DEBUG
extern int ip_vs_get_debug_level(void);
#define IP_VS_DBG(level, msg...)			\
    do {						\
	    if (level <= ip_vs_get_debug_level())	\
		    printk(KERN_DEBUG "IPVS: " msg);	\
    } while (0)
#define IP_VS_DBG_RL(msg...)				\
    do {						\
	    if (net_ratelimit())			\
		    printk(KERN_DEBUG "IPVS: " msg);	\
    } while (0)
#define IP_VS_DBG_PKT(level, pp, skb, ofs, msg)		\
    do {						\
	    if (level <= ip_vs_get_debug_level())	\
		pp->debug_packet(pp, skb, ofs, msg);	\
    } while (0)
#define IP_VS_DBG_RL_PKT(level, pp, skb, ofs, msg)	\
    do {						\
	    if (level <= ip_vs_get_debug_level() &&	\
		net_ratelimit())			\
		pp->debug_packet(pp, skb, ofs, msg);	\
    } while (0)
#else	/* NO DEBUGGING at ALL */
#define IP_VS_DBG(level, msg...)  do {} while (0)
#define IP_VS_DBG_RL(msg...)  do {} while (0)
#define IP_VS_DBG_PKT(level, pp, skb, ofs, msg)		do {} while (0)
#define IP_VS_DBG_RL_PKT(level, pp, skb, ofs, msg)	do {} while (0)
#endif

#define IP_VS_BUG() BUG()
#define IP_VS_ERR(msg...) printk(KERN_ERR "IPVS: " msg)
#define IP_VS_INFO(msg...) printk(KERN_INFO "IPVS: " msg)
#define IP_VS_WARNING(msg...) \
	printk(KERN_WARNING "IPVS: " msg)
#define IP_VS_ERR_RL(msg...)				\
    do {						\
	    if (net_ratelimit())			\
		    printk(KERN_ERR "IPVS: " msg);	\
    } while (0)

#ifdef CONFIG_IP_VS_DEBUG
#define EnterFunction(level)						\
    do {								\
	    if (level <= ip_vs_get_debug_level())			\
		    printk(KERN_DEBUG "Enter: %s, %s line %i\n",	\
			   __FUNCTION__, __FILE__, __LINE__);		\
    } while (0)
#define LeaveFunction(level)                                            \
    do {                                                                \
	    if (level <= ip_vs_get_debug_level())                       \
			printk(KERN_DEBUG "Leave: %s, %s line %i\n",    \
			       __FUNCTION__, __FILE__, __LINE__);       \
    } while (0)
#else
#define EnterFunction(level)   do {} while (0)
#define LeaveFunction(level)   do {} while (0)
#endif

#define	IP_VS_WAIT_WHILE(expr)	while (expr) { cpu_relax(); }


/*
 *      The port number of FTP service (in network order).
 */
#define FTPPORT  __constant_htons(21)
#define FTPDATA  __constant_htons(20)

/*
 *      IPVS sysctl variables under the /proc/sys/net/ipv4/vs/
 */
#define NET_IPV4_VS              21

enum {
	NET_IPV4_VS_DEBUG_LEVEL=1,
	NET_IPV4_VS_AMEMTHRESH=2,
	NET_IPV4_VS_AMDROPRATE=3,
	NET_IPV4_VS_DROP_ENTRY=4,
	NET_IPV4_VS_DROP_PACKET=5,
	NET_IPV4_VS_SECURE_TCP=6,
	NET_IPV4_VS_TO_ES=7,
	NET_IPV4_VS_TO_SS=8,
	NET_IPV4_VS_TO_SR=9,
	NET_IPV4_VS_TO_FW=10,
	NET_IPV4_VS_TO_TW=11,
	NET_IPV4_VS_TO_CL=12,
	NET_IPV4_VS_TO_CW=13,
	NET_IPV4_VS_TO_LA=14,
	NET_IPV4_VS_TO_LI=15,
	NET_IPV4_VS_TO_SA=16,
	NET_IPV4_VS_TO_UDP=17,
	NET_IPV4_VS_TO_ICMP=18,
	NET_IPV4_VS_LBLC_EXPIRE=19,
	NET_IPV4_VS_LBLCR_EXPIRE=20,
	NET_IPV4_VS_CACHE_BYPASS=22,
	NET_IPV4_VS_EXPIRE_NODEST_CONN=23,
	NET_IPV4_VS_SYNC_THRESHOLD=24,
	NET_IPV4_VS_NAT_ICMP_SEND=25,
	NET_IPV4_VS_LAST
};

/*
 *      TCP State Values
 */
enum {
	IP_VS_TCP_S_NONE = 0,
	IP_VS_TCP_S_ESTABLISHED,
	IP_VS_TCP_S_SYN_SENT,
	IP_VS_TCP_S_SYN_RECV,
	IP_VS_TCP_S_FIN_WAIT,
	IP_VS_TCP_S_TIME_WAIT,
	IP_VS_TCP_S_CLOSE,
	IP_VS_TCP_S_CLOSE_WAIT,
	IP_VS_TCP_S_LAST_ACK,
	IP_VS_TCP_S_LISTEN,
	IP_VS_TCP_S_SYNACK,
	IP_VS_TCP_S_LAST
};

/*
 *	UDP State Values
 */
enum {
	IP_VS_UDP_S_NORMAL,
	IP_VS_UDP_S_LAST,
};

/*
 *	ICMP State Values
 */
enum {
	IP_VS_ICMP_S_NORMAL,
	IP_VS_ICMP_S_LAST,
};

/*
 *	Delta sequence info structure
 *	Each ip_vs_conn has 2 (output AND input seq. changes).
 *      Only used in the VS/NAT.
 */
struct ip_vs_seq {
	__u32			init_seq;	/* Add delta from this seq */
	__u32			delta;		/* Delta in sequence numbers */
	__u32			previous_delta;	/* Delta in sequence numbers
						   before last resized pkt */
};


/*
 *	IPVS statistics object
 */
struct ip_vs_stats
{
	__u32                   conns;          /* connections scheduled */
	__u32                   inpkts;         /* incoming packets */
	__u32                   outpkts;        /* outgoing packets */
	__u64                   inbytes;        /* incoming bytes */
	__u64                   outbytes;       /* outgoing bytes */

	__u32			cps;		/* current connection rate */
	__u32			inpps;		/* current in packet rate */
	__u32			outpps;		/* current out packet rate */
	__u32			inbps;		/* current in byte rate */
	__u32			outbps;		/* current out byte rate */

	spinlock_t              lock;           /* spin lock */
};

struct ip_vs_conn;
struct ip_vs_app;

struct ip_vs_protocol {
	struct ip_vs_protocol	*next;
	char			*name;
	__u16			protocol;
	int			dont_defrag;
	atomic_t		appcnt;		/* counter of proto app incs */
	int			*timeout_table;	/* protocol timeout table */

	void (*init)(struct ip_vs_protocol *pp);

	void (*exit)(struct ip_vs_protocol *pp);

	int (*conn_schedule)(struct sk_buff *skb,
			     struct ip_vs_protocol *pp,
			     int *verdict, struct ip_vs_conn **cpp);

	struct ip_vs_conn *
	(*conn_in_get)(const struct sk_buff *skb,
		       struct ip_vs_protocol *pp,
		       const struct iphdr *iph,
		       unsigned int proto_off,
		       int inverse);

	struct ip_vs_conn *
	(*conn_out_get)(const struct sk_buff *skb,
			struct ip_vs_protocol *pp,
			const struct iphdr *iph,
			unsigned int proto_off,
			int inverse);

	int (*snat_handler)(struct sk_buff **pskb,
			    struct ip_vs_protocol *pp, struct ip_vs_conn *cp);

	int (*dnat_handler)(struct sk_buff **pskb,
			    struct ip_vs_protocol *pp, struct ip_vs_conn *cp);

	int (*csum_check)(struct sk_buff *skb, struct ip_vs_protocol *pp);

	const char *(*state_name)(int state);

	int (*state_transition)(struct ip_vs_conn *cp, int direction,
				const struct sk_buff *skb,
				struct ip_vs_protocol *pp);

	int (*register_app)(struct ip_vs_app *inc);

	void (*unregister_app)(struct ip_vs_app *inc);

	int (*app_conn_bind)(struct ip_vs_conn *cp);

	void (*debug_packet)(struct ip_vs_protocol *pp,
			     const struct sk_buff *skb,
			     int offset,
			     const char *msg);

	void (*timeout_change)(struct ip_vs_protocol *pp, int flags);

	int (*set_state_timeout)(struct ip_vs_protocol *pp, char *sname, int to);
};

extern struct ip_vs_protocol * ip_vs_proto_get(unsigned short proto);

/*
 *	IP_VS structure allocated for each dynamically scheduled connection
 */
struct ip_vs_conn {
	struct list_head        c_list;         /* hashed list heads */