nfsd: track last inode only in use_wgather case
[safe/jmp/linux-2.6] / include / net / ip_vs.h
index d328255..bbae1e8 100644 (file)
 
 #include <net/checksum.h>
 #include <linux/netfilter.h>           /* for union nf_inet_addr */
+#include <linux/ip.h>
 #include <linux/ipv6.h>                        /* for struct ipv6hdr */
 #include <net/ipv6.h>                  /* for ipv6_addr_copy */
 
+struct ip_vs_iphdr {
+       int len;
+       __u8 protocol;
+       union nf_inet_addr saddr;
+       union nf_inet_addr daddr;
+};
+
+static inline void
+ip_vs_fill_iphdr(int af, const void *nh, struct ip_vs_iphdr *iphdr)
+{
+#ifdef CONFIG_IP_VS_IPV6
+       if (af == AF_INET6) {
+               const struct ipv6hdr *iph = nh;
+               iphdr->len = sizeof(struct ipv6hdr);
+               iphdr->protocol = iph->nexthdr;
+               ipv6_addr_copy(&iphdr->saddr.in6, &iph->saddr);
+               ipv6_addr_copy(&iphdr->daddr.in6, &iph->daddr);
+       } else
+#endif
+       {
+               const struct iphdr *iph = nh;
+               iphdr->len = iph->ihl * 4;
+               iphdr->protocol = iph->protocol;
+               iphdr->saddr.ip = iph->saddr;
+               iphdr->daddr.ip = iph->daddr;
+       }
+}
+
+static inline void ip_vs_addr_copy(int af, union nf_inet_addr *dst,
+                                  const union nf_inet_addr *src)
+{
+#ifdef CONFIG_IP_VS_IPV6
+       if (af == AF_INET6)
+               ipv6_addr_copy(&dst->in6, &src->in6);
+       else
+#endif
+       dst->ip = src->ip;
+}
+
+static inline int ip_vs_addr_equal(int af, const union nf_inet_addr *a,
+                                  const union nf_inet_addr *b)
+{
+#ifdef CONFIG_IP_VS_IPV6
+       if (af == AF_INET6)
+               return ipv6_addr_equal(&a->in6, &b->in6);
+#endif
+       return a->ip == b->ip;
+}
+
 #ifdef CONFIG_IP_VS_DEBUG
 #include <linux/net.h>
 
 extern int ip_vs_get_debug_level(void);
+
+static inline const char *ip_vs_dbg_addr(int af, char *buf, size_t buf_len,
+                                        const union nf_inet_addr *addr,
+                                        int *idx)
+{
+       int len;
+#ifdef CONFIG_IP_VS_IPV6
+       if (af == AF_INET6)
+               len = snprintf(&buf[*idx], buf_len - *idx, "[%pI6]",
+                              &addr->in6) + 1;
+       else
+#endif
+               len = snprintf(&buf[*idx], buf_len - *idx, "%pI4",
+                              &addr->ip) + 1;
+
+       *idx += len;
+       BUG_ON(*idx > buf_len + 1);
+       return &buf[*idx - len];
+}
+
+#define IP_VS_DBG_BUF(level, msg...)                   \
+    do {                                               \
+           char ip_vs_dbg_buf[160];                    \
+           int ip_vs_dbg_idx = 0;                      \
+           if (level <= ip_vs_get_debug_level())       \
+                   printk(KERN_DEBUG "IPVS: " msg);    \
+    } while (0)
+#define IP_VS_ERR_BUF(msg...)                          \
+    do {                                               \
+           char ip_vs_dbg_buf[160];                    \
+           int ip_vs_dbg_idx = 0;                      \
+           printk(KERN_ERR "IPVS: " msg);              \
+    } while (0)
+
+/* Only use from within IP_VS_DBG_BUF() or IP_VS_ERR_BUF macros */
+#define IP_VS_DBG_ADDR(af, addr)                       \
+    ip_vs_dbg_addr(af, ip_vs_dbg_buf,                  \
+                  sizeof(ip_vs_dbg_buf), addr,         \
+                  &ip_vs_dbg_idx)
+
 #define IP_VS_DBG(level, msg...)                       \
     do {                                               \
            if (level <= ip_vs_get_debug_level())       \
@@ -51,6 +141,8 @@ extern int ip_vs_get_debug_level(void);
                pp->debug_packet(pp, skb, ofs, msg);    \
     } while (0)
 #else  /* NO DEBUGGING at ALL */
+#define IP_VS_DBG_BUF(level, msg...)  do {} while (0)
+#define IP_VS_ERR_BUF(msg...)  do {} while (0)
 #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)
@@ -73,13 +165,13 @@ extern int ip_vs_get_debug_level(void);
     do {                                                               \
            if (level <= ip_vs_get_debug_level())                       \
                    printk(KERN_DEBUG "Enter: %s, %s line %i\n",        \
-                          __FUNCTION__, __FILE__, __LINE__);           \
+                          __func__, __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__);       \
+                              __func__, __FILE__, __LINE__);       \
     } while (0)
 #else
 #define EnterFunction(level)   do {} while (0)
@@ -92,8 +184,8 @@ extern int ip_vs_get_debug_level(void);
 /*
  *      The port number of FTP service (in network order).
  */
-#define FTPPORT  __constant_htons(21)
-#define FTPDATA  __constant_htons(20)
+#define FTPPORT  cpu_to_be16(21)
+#define FTPDATA  cpu_to_be16(20)
 
 /*
  *      TCP State Values
@@ -163,27 +255,10 @@ struct ip_vs_estimator {
 
 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 */
-
-       /*
-        * Don't add anything before the lock, because we use memcpy() to copy
-        * the members before the lock to struct ip_vs_stats_user in
-        * ip_vs_ctl.c.
-        */
+       struct ip_vs_stats_user ustats;         /* statistics */
+       struct ip_vs_estimator  est;            /* estimator */
 
        spinlock_t              lock;           /* spin lock */
-
-       struct ip_vs_estimator  est;            /* estimator */
 };
 
 struct dst_entry;
@@ -205,21 +280,23 @@ struct ip_vs_protocol {
 
        void (*exit)(struct ip_vs_protocol *pp);
 
-       int (*conn_schedule)(struct sk_buff *skb,
+       int (*conn_schedule)(int af, 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,
+       (*conn_in_get)(int af,
+                      const struct sk_buff *skb,
                       struct ip_vs_protocol *pp,
-                      const struct iphdr *iph,
+                      const struct ip_vs_iphdr *iph,
                       unsigned int proto_off,
                       int inverse);
 
        struct ip_vs_conn *
-       (*conn_out_get)(const struct sk_buff *skb,
+       (*conn_out_get)(int af,
+                       const struct sk_buff *skb,
                        struct ip_vs_protocol *pp,
-                       const struct iphdr *iph,
+                       const struct ip_vs_iphdr *iph,
                        unsigned int proto_off,
                        int inverse);
 
@@ -229,7 +306,8 @@ struct ip_vs_protocol {
        int (*dnat_handler)(struct sk_buff *skb,
                            struct ip_vs_protocol *pp, struct ip_vs_conn *cp);
 
-       int (*csum_check)(struct sk_buff *skb, struct ip_vs_protocol *pp);
+       int (*csum_check)(int af, struct sk_buff *skb,
+                         struct ip_vs_protocol *pp);
 
        const char *(*state_name)(int state);
 
@@ -309,6 +387,45 @@ struct ip_vs_conn {
 
 
 /*
+ *     Extended internal versions of struct ip_vs_service_user and
+ *     ip_vs_dest_user for IPv6 support.
+ *
+ *     We need these to conveniently pass around service and destination
+ *     options, but unfortunately, we also need to keep the old definitions to
+ *     maintain userspace backwards compatibility for the setsockopt interface.
+ */
+struct ip_vs_service_user_kern {
+       /* virtual service addresses */
+       u16                     af;
+       u16                     protocol;
+       union nf_inet_addr      addr;           /* virtual ip address */
+       u16                     port;
+       u32                     fwmark;         /* firwall mark of service */
+
+       /* virtual service options */
+       char                    *sched_name;
+       unsigned                flags;          /* virtual service flags */
+       unsigned                timeout;        /* persistent timeout in sec */
+       u32                     netmask;        /* persistent netmask */
+};
+
+
+struct ip_vs_dest_user_kern {
+       /* destination server address */
+       union nf_inet_addr      addr;
+       u16                     port;
+
+       /* real server options */
+       unsigned                conn_flags;     /* connection flags */
+       int                     weight;         /* destination weight */
+
+       /* thresholds for active connections */
+       u32                     u_threshold;    /* upper threshold */
+       u32                     l_threshold;    /* lower threshold */
+};
+
+
+/*
  *     The information about the virtual service offered to the net
  *     and the forwarding entries
  */
@@ -485,16 +602,8 @@ extern void ip_vs_init_hash_table(struct list_head *table, int rows);
 #ifndef CONFIG_IP_VS_TAB_BITS
 #define CONFIG_IP_VS_TAB_BITS   12
 #endif
-/* make sure that IP_VS_CONN_TAB_BITS is located in [8, 20] */
-#if CONFIG_IP_VS_TAB_BITS < 8
-#define IP_VS_CONN_TAB_BITS    8
-#endif
-#if CONFIG_IP_VS_TAB_BITS > 20
-#define IP_VS_CONN_TAB_BITS    20
-#endif
-#if 8 <= CONFIG_IP_VS_TAB_BITS && CONFIG_IP_VS_TAB_BITS <= 20
+
 #define IP_VS_CONN_TAB_BITS    CONFIG_IP_VS_TAB_BITS
-#endif
 #define IP_VS_CONN_TAB_SIZE     (1 << IP_VS_CONN_TAB_BITS)
 #define IP_VS_CONN_TAB_MASK     (IP_VS_CONN_TAB_SIZE - 1)
 
@@ -506,11 +615,16 @@ enum {
 };
 
 extern struct ip_vs_conn *ip_vs_conn_in_get
-(int protocol, __be32 s_addr, __be16 s_port, __be32 d_addr, __be16 d_port);
+(int af, int protocol, const union nf_inet_addr *s_addr, __be16 s_port,
+ const union nf_inet_addr *d_addr, __be16 d_port);
+
 extern struct ip_vs_conn *ip_vs_ct_in_get
-(int protocol, __be32 s_addr, __be16 s_port, __be32 d_addr, __be16 d_port);
+(int af, int protocol, const union nf_inet_addr *s_addr, __be16 s_port,
+ const union nf_inet_addr *d_addr, __be16 d_port);
+
 extern struct ip_vs_conn *ip_vs_conn_out_get
-(int protocol, __be32 s_addr, __be16 s_port, __be32 d_addr, __be16 d_port);
+(int af, int protocol, const union nf_inet_addr *s_addr, __be16 s_port,
+ const union nf_inet_addr *d_addr, __be16 d_port);
 
 /* put back the conn without restarting its timer */
 static inline void __ip_vs_conn_put(struct ip_vs_conn *cp)
@@ -521,8 +635,9 @@ extern void ip_vs_conn_put(struct ip_vs_conn *cp);
 extern void ip_vs_conn_fill_cport(struct ip_vs_conn *cp, __be16 cport);
 
 extern struct ip_vs_conn *
-ip_vs_conn_new(int proto, __be32 caddr, __be16 cport, __be32 vaddr, __be16 vport,
-              __be32 daddr, __be16 dport, unsigned flags,
+ip_vs_conn_new(int af, int proto, const union nf_inet_addr *caddr, __be16 cport,
+              const union nf_inet_addr *vaddr, __be16 vport,
+              const union nf_inet_addr *daddr, __be16 dport, unsigned flags,
               struct ip_vs_dest *dest);
 extern void ip_vs_conn_expire_now(struct ip_vs_conn *cp);
 
@@ -538,24 +653,32 @@ static inline void ip_vs_control_del(struct ip_vs_conn *cp)
 {
        struct ip_vs_conn *ctl_cp = cp->control;
        if (!ctl_cp) {
-               IP_VS_ERR("request control DEL for uncontrolled: "
-                         "%d.%d.%d.%d:%d to %d.%d.%d.%d:%d\n",
-                         NIPQUAD(cp->caddr),ntohs(cp->cport),
-                         NIPQUAD(cp->vaddr),ntohs(cp->vport));
+               IP_VS_ERR_BUF("request control DEL for uncontrolled: "
+                             "%s:%d to %s:%d\n",
+                             IP_VS_DBG_ADDR(cp->af, &cp->caddr),
+                             ntohs(cp->cport),
+                             IP_VS_DBG_ADDR(cp->af, &cp->vaddr),
+                             ntohs(cp->vport));
+
                return;
        }
 
-       IP_VS_DBG(7, "DELeting control for: "
-                 "cp.dst=%d.%d.%d.%d:%d ctl_cp.dst=%d.%d.%d.%d:%d\n",
-                 NIPQUAD(cp->caddr),ntohs(cp->cport),
-                 NIPQUAD(ctl_cp->caddr),ntohs(ctl_cp->cport));
+       IP_VS_DBG_BUF(7, "DELeting control for: "
+                     "cp.dst=%s:%d ctl_cp.dst=%s:%d\n",
+                     IP_VS_DBG_ADDR(cp->af, &cp->caddr),
+                     ntohs(cp->cport),
+                     IP_VS_DBG_ADDR(cp->af, &ctl_cp->caddr),
+                     ntohs(ctl_cp->cport));
 
        cp->control = NULL;
        if (atomic_read(&ctl_cp->n_control) == 0) {
-               IP_VS_ERR("BUG control DEL with n=0 : "
-                         "%d.%d.%d.%d:%d to %d.%d.%d.%d:%d\n",
-                         NIPQUAD(cp->caddr),ntohs(cp->cport),
-                         NIPQUAD(cp->vaddr),ntohs(cp->vport));
+               IP_VS_ERR_BUF("BUG control DEL with n=0 : "
+                             "%s:%d to %s:%d\n",
+                             IP_VS_DBG_ADDR(cp->af, &cp->caddr),
+                             ntohs(cp->cport),
+                             IP_VS_DBG_ADDR(cp->af, &cp->vaddr),
+                             ntohs(cp->vport));
+
                return;
        }
        atomic_dec(&ctl_cp->n_control);
@@ -565,17 +688,22 @@ static inline void
 ip_vs_control_add(struct ip_vs_conn *cp, struct ip_vs_conn *ctl_cp)
 {
        if (cp->control) {
-               IP_VS_ERR("request control ADD for already controlled: "
-                         "%d.%d.%d.%d:%d to %d.%d.%d.%d:%d\n",
-                         NIPQUAD(cp->caddr),ntohs(cp->cport),
-                         NIPQUAD(cp->vaddr),ntohs(cp->vport));
+               IP_VS_ERR_BUF("request control ADD for already controlled: "
+                             "%s:%d to %s:%d\n",
+                             IP_VS_DBG_ADDR(cp->af, &cp->caddr),
+                             ntohs(cp->cport),
+                             IP_VS_DBG_ADDR(cp->af, &cp->vaddr),
+                             ntohs(cp->vport));
+
                ip_vs_control_del(cp);
        }
 
-       IP_VS_DBG(7, "ADDing control for: "
-                 "cp.dst=%d.%d.%d.%d:%d ctl_cp.dst=%d.%d.%d.%d:%d\n",
-                 NIPQUAD(cp->caddr),ntohs(cp->cport),
-                 NIPQUAD(ctl_cp->caddr),ntohs(ctl_cp->cport));
+       IP_VS_DBG_BUF(7, "ADDing control for: "
+                     "cp.dst=%s:%d ctl_cp.dst=%s:%d\n",
+                     IP_VS_DBG_ADDR(cp->af, &cp->caddr),
+                     ntohs(cp->cport),
+                     IP_VS_DBG_ADDR(cp->af, &ctl_cp->caddr),
+                     ntohs(ctl_cp->cport));
 
        cp->control = ctl_cp;
        atomic_inc(&ctl_cp->n_control);
@@ -653,7 +781,8 @@ extern struct ip_vs_stats ip_vs_stats;
 extern const struct ctl_path net_vs_ctl_path[];
 
 extern struct ip_vs_service *
-ip_vs_service_get(__u32 fwmark, __u16 protocol, __be32 vaddr, __be16 vport);
+ip_vs_service_get(int af, __u32 fwmark, __u16 protocol,
+                 const union nf_inet_addr *vaddr, __be16 vport);
 
 static inline void ip_vs_service_put(struct ip_vs_service *svc)
 {
@@ -661,14 +790,16 @@ static inline void ip_vs_service_put(struct ip_vs_service *svc)
 }
 
 extern struct ip_vs_dest *
-ip_vs_lookup_real_service(__u16 protocol, __be32 daddr, __be16 dport);
+ip_vs_lookup_real_service(int af, __u16 protocol,
+                         const union nf_inet_addr *daddr, __be16 dport);
+
 extern int ip_vs_use_count_inc(void);
 extern void ip_vs_use_count_dec(void);
 extern int ip_vs_control_init(void);
 extern void ip_vs_control_cleanup(void);
 extern struct ip_vs_dest *
-ip_vs_find_dest(__be32 daddr, __be16 dport,
-                __be32 vaddr, __be16 vport, __u16 protocol);
+ip_vs_find_dest(int af, const union nf_inet_addr *daddr, __be16 dport,
+               const union nf_inet_addr *vaddr, __be16 vport, __u16 protocol);
 extern struct ip_vs_dest *ip_vs_try_bind_dest(struct ip_vs_conn *cp);
 
 
@@ -712,6 +843,19 @@ extern int ip_vs_icmp_xmit
 (struct sk_buff *skb, struct ip_vs_conn *cp, struct ip_vs_protocol *pp, int offset);
 extern void ip_vs_dst_reset(struct ip_vs_dest *dest);
 
+#ifdef CONFIG_IP_VS_IPV6
+extern int ip_vs_bypass_xmit_v6
+(struct sk_buff *skb, struct ip_vs_conn *cp, struct ip_vs_protocol *pp);
+extern int ip_vs_nat_xmit_v6
+(struct sk_buff *skb, struct ip_vs_conn *cp, struct ip_vs_protocol *pp);
+extern int ip_vs_tunnel_xmit_v6
+(struct sk_buff *skb, struct ip_vs_conn *cp, struct ip_vs_protocol *pp);
+extern int ip_vs_dr_xmit_v6
+(struct sk_buff *skb, struct ip_vs_conn *cp, struct ip_vs_protocol *pp);
+extern int ip_vs_icmp_xmit_v6
+(struct sk_buff *skb, struct ip_vs_conn *cp, struct ip_vs_protocol *pp,
+ int offset);
+#endif
 
 /*
  *     This is a simple mechanism to ignore packets when
@@ -756,7 +900,12 @@ static inline char ip_vs_fwd_tag(struct ip_vs_conn *cp)
 }
 
 extern void ip_vs_nat_icmp(struct sk_buff *skb, struct ip_vs_protocol *pp,
-               struct ip_vs_conn *cp, int dir);
+                          struct ip_vs_conn *cp, int dir);
+
+#ifdef CONFIG_IP_VS_IPV6
+extern void ip_vs_nat_icmp_v6(struct sk_buff *skb, struct ip_vs_protocol *pp,
+                             struct ip_vs_conn *cp, int dir);
+#endif
 
 extern __sum16 ip_vs_checksum_complete(struct sk_buff *skb, int offset);
 
@@ -764,14 +913,25 @@ static inline __wsum ip_vs_check_diff4(__be32 old, __be32 new, __wsum oldsum)
 {
        __be32 diff[2] = { ~old, new };
 
-       return csum_partial((char *) diff, sizeof(diff), oldsum);
+       return csum_partial(diff, sizeof(diff), oldsum);
+}
+
+#ifdef CONFIG_IP_VS_IPV6
+static inline __wsum ip_vs_check_diff16(const __be32 *old, const __be32 *new,
+                                       __wsum oldsum)
+{
+       __be32 diff[8] = { ~old[3], ~old[2], ~old[1], ~old[0],
+                           new[3],  new[2],  new[1],  new[0] };
+
+       return csum_partial(diff, sizeof(diff), oldsum);
 }
+#endif
 
 static inline __wsum ip_vs_check_diff2(__be16 old, __be16 new, __wsum oldsum)
 {
        __be16 diff[2] = { ~old, new };
 
-       return csum_partial((char *) diff, sizeof(diff), oldsum);
+       return csum_partial(diff, sizeof(diff), oldsum);
 }
 
 #endif /* __KERNEL__ */