#include <net/tcp.h>
#include <net/ndisc.h>
+#include <net/inet6_hashtables.h>
#include <net/ipv6.h>
#include <net/transp_v6.h>
#include <net/addrconf.h>
static struct tcp_func ipv6_mapped;
static struct tcp_func ipv6_specific;
-/* I have no idea if this is a good hash for v6 or not. -DaveM */
-static __inline__ int tcp_v6_hashfn(struct in6_addr *laddr, u16 lport,
- struct in6_addr *faddr, u16 fport)
-{
- int hashent = (lport ^ fport);
-
- hashent ^= (laddr->s6_addr32[3] ^ faddr->s6_addr32[3]);
- hashent ^= hashent>>16;
- hashent ^= hashent>>8;
- return (hashent & (tcp_hashinfo.ehash_size - 1));
-}
-
-static __inline__ int tcp_v6_sk_hashfn(struct sock *sk)
-{
- struct inet_sock *inet = inet_sk(sk);
- struct ipv6_pinfo *np = inet6_sk(sk);
- struct in6_addr *laddr = &np->rcv_saddr;
- struct in6_addr *faddr = &np->daddr;
- __u16 lport = inet->num;
- __u16 fport = inet->dport;
- return tcp_v6_hashfn(laddr, lport, faddr, fport);
-}
-
static inline int tcp_v6_bind_conflict(const struct sock *sk,
const struct inet_bind_bucket *tb)
{
lock = &tcp_hashinfo.lhash_lock;
inet_listen_wlock(&tcp_hashinfo);
} else {
- sk->sk_hashent = tcp_v6_sk_hashfn(sk);
- list = &tcp_hashinfo.ehash[sk->sk_hashent].chain;
- lock = &tcp_hashinfo.ehash[sk->sk_hashent].lock;
+ unsigned int hash;
+ sk->sk_hash = hash = inet6_sk_ehashfn(sk);
+ hash &= (tcp_hashinfo.ehash_size - 1);
+ list = &tcp_hashinfo.ehash[hash].chain;
+ lock = &tcp_hashinfo.ehash[hash].lock;
write_lock(lock);
}
}
}
-static struct sock *tcp_v6_lookup_listener(struct in6_addr *daddr, unsigned short hnum, int dif)
-{
- struct sock *sk;
- struct hlist_node *node;
- struct sock *result = NULL;
- int score, hiscore;
-
- hiscore=0;
- read_lock(&tcp_hashinfo.lhash_lock);
- sk_for_each(sk, node, &tcp_hashinfo.listening_hash[inet_lhashfn(hnum)]) {
- if (inet_sk(sk)->num == hnum && sk->sk_family == PF_INET6) {
- struct ipv6_pinfo *np = inet6_sk(sk);
-
- score = 1;
- if (!ipv6_addr_any(&np->rcv_saddr)) {
- if (!ipv6_addr_equal(&np->rcv_saddr, daddr))
- continue;
- score++;
- }
- if (sk->sk_bound_dev_if) {
- if (sk->sk_bound_dev_if != dif)
- continue;
- score++;
- }
- if (score == 3) {
- result = sk;
- break;
- }
- if (score > hiscore) {
- hiscore = score;
- result = sk;
- }
- }
- }
- if (result)
- sock_hold(result);
- read_unlock(&tcp_hashinfo.lhash_lock);
- return result;
-}
-
-/* Sockets in TCP_CLOSE state are _always_ taken out of the hash, so
- * we need not check it for TCP lookups anymore, thanks Alexey. -DaveM
- *
- * The sockhash lock must be held as a reader here.
- */
-
-static inline struct sock *__tcp_v6_lookup_established(struct in6_addr *saddr, u16 sport,
- struct in6_addr *daddr, u16 hnum,
- int dif)
-{
- struct sock *sk;
- const struct hlist_node *node;
- const __u32 ports = INET_COMBINED_PORTS(sport, hnum);
- /* Optimize here for direct hit, only listening connections can
- * have wildcards anyways.
- */
- const int hash = tcp_v6_hashfn(daddr, hnum, saddr, sport);
- struct inet_ehash_bucket *head = &tcp_hashinfo.ehash[hash];
-
- read_lock(&head->lock);
- sk_for_each(sk, node, &head->chain) {
- /* For IPV6 do the cheaper port and family tests first. */
- if (INET6_MATCH(sk, saddr, daddr, ports, dif))
- goto hit; /* You sunk my battleship! */
- }
- /* Must check for a TIME_WAIT'er before going to listener hash. */
- sk_for_each(sk, node, &(head + tcp_hashinfo.ehash_size)->chain) {
- const struct inet_timewait_sock *tw = inet_twsk(sk);
-
- if(*((__u32 *)&(tw->tw_dport)) == ports &&
- sk->sk_family == PF_INET6) {
- const struct tcp6_timewait_sock *tcp6tw = tcp6_twsk(sk);
-
- if (ipv6_addr_equal(&tcp6tw->tw_v6_daddr, saddr) &&
- ipv6_addr_equal(&tcp6tw->tw_v6_rcv_saddr, daddr) &&
- (!sk->sk_bound_dev_if || sk->sk_bound_dev_if == dif))
- goto hit;
- }
- }
- read_unlock(&head->lock);
- return NULL;
-
-hit:
- sock_hold(sk);
- read_unlock(&head->lock);
- return sk;
-}
-
-
-static inline struct sock *__tcp_v6_lookup(struct in6_addr *saddr, u16 sport,
- struct in6_addr *daddr, u16 hnum,
- int dif)
-{
- struct sock *sk;
-
- sk = __tcp_v6_lookup_established(saddr, sport, daddr, hnum, dif);
-
- if (sk)
- return sk;
-
- return tcp_v6_lookup_listener(daddr, hnum, dif);
-}
-
-inline struct sock *tcp_v6_lookup(struct in6_addr *saddr, u16 sport,
- struct in6_addr *daddr, u16 dport,
- int dif)
-{
- struct sock *sk;
-
- local_bh_disable();
- sk = __tcp_v6_lookup(saddr, sport, daddr, ntohs(dport), dif);
- local_bh_enable();
-
- return sk;
-}
-
-EXPORT_SYMBOL_GPL(tcp_v6_lookup);
-
-
/*
* Open request hash tables.
*/
}
}
-static int __tcp_v6_check_established(struct sock *sk, __u16 lport,
+static int __tcp_v6_check_established(struct sock *sk, const __u16 lport,
struct inet_timewait_sock **twp)
{
struct inet_sock *inet = inet_sk(sk);
- struct ipv6_pinfo *np = inet6_sk(sk);
- struct in6_addr *daddr = &np->rcv_saddr;
- struct in6_addr *saddr = &np->daddr;
- int dif = sk->sk_bound_dev_if;
+ const struct ipv6_pinfo *np = inet6_sk(sk);
+ const struct in6_addr *daddr = &np->rcv_saddr;
+ const struct in6_addr *saddr = &np->daddr;
+ const int dif = sk->sk_bound_dev_if;
const u32 ports = INET_COMBINED_PORTS(inet->dport, lport);
- const int hash = tcp_v6_hashfn(daddr, inet->num, saddr, inet->dport);
- struct inet_ehash_bucket *head = &tcp_hashinfo.ehash[hash];
+ unsigned int hash = inet6_ehashfn(daddr, inet->num, saddr, inet->dport);
+ struct inet_ehash_bucket *head = inet_ehash_bucket(&tcp_hashinfo, hash);
struct sock *sk2;
const struct hlist_node *node;
struct inet_timewait_sock *tw;
+ prefetch(head->chain.first);
write_lock(&head->lock);
/* Check TIME-WAIT sockets first. */
/* And established part... */
sk_for_each(sk2, node, &head->chain) {
- if (INET6_MATCH(sk2, saddr, daddr, ports, dif))
+ if (INET6_MATCH(sk2, hash, saddr, daddr, ports, dif))
goto not_unique;
}
unique:
BUG_TRAP(sk_unhashed(sk));
__sk_add_node(sk, &head->chain);
- sk->sk_hashent = hash;
+ sk->sk_hash = hash;
sock_prot_inc_use(sk->sk_prot);
write_unlock(&head->lock);
}
}
-static __inline__ int tcp_v6_iif(struct sk_buff *skb)
-{
- return IP6CB(skb)->iif;
-}
-
static int tcp_v6_connect(struct sock *sk, struct sockaddr *uaddr,
int addr_len)
{
if (final_p)
ipv6_addr_copy(&fl.fl6_dst, final_p);
- if ((err = xfrm_lookup(&dst, &fl, sk, 0)) < 0) {
- dst_release(dst);
+ if ((err = xfrm_lookup(&dst, &fl, sk, 0)) < 0)
goto failure;
- }
if (saddr == NULL) {
saddr = &fl.fl6_src;
int type, int code, int offset, __u32 info)
{
struct ipv6hdr *hdr = (struct ipv6hdr*)skb->data;
- struct tcphdr *th = (struct tcphdr *)(skb->data+offset);
+ const struct tcphdr *th = (struct tcphdr *)(skb->data+offset);
struct ipv6_pinfo *np;
struct sock *sk;
int err;
struct tcp_sock *tp;
__u32 seq;
- sk = tcp_v6_lookup(&hdr->daddr, th->dest, &hdr->saddr, th->source, skb->dev->ifindex);
+ sk = inet6_lookup(&tcp_hashinfo, &hdr->daddr, th->dest, &hdr->saddr,
+ th->source, skb->dev->ifindex);
if (sk == NULL) {
ICMP6_INC_STATS_BH(__in6_dev_get(skb->dev), ICMP6_MIB_INERRORS);
goto out;
req = tcp_v6_search_req(sk, &prev, th->dest, &hdr->daddr,
- &hdr->saddr, tcp_v6_iif(skb));
+ &hdr->saddr, inet6_iif(skb));
if (!req)
goto out;
if (dst == NULL) {
opt = np->opt;
if (opt == NULL &&
- np->rxopt.bits.srcrt == 2 &&
+ np->rxopt.bits.osrcrt == 2 &&
treq->pktopts) {
struct sk_buff *pktopts = treq->pktopts;
struct inet6_skb_parm *rxopt = IP6CB(pktopts);
}
done:
- dst_release(dst);
if (opt && opt != np->opt)
sock_kfree_s(sk, opt, opt->tot_len);
return err;
struct inet6_skb_parm *opt = IP6CB(skb);
if (np->rxopt.all) {
- if ((opt->hop && np->rxopt.bits.hopopts) ||
- ((IPV6_FLOWINFO_MASK&*(u32*)skb->nh.raw) &&
- np->rxopt.bits.rxflow) ||
- (opt->srcrt && np->rxopt.bits.srcrt) ||
- ((opt->dst1 || opt->dst0) && np->rxopt.bits.dstopts))
+ if ((opt->hop && (np->rxopt.bits.hopopts || np->rxopt.bits.ohopopts)) ||
+ ((IPV6_FLOWINFO_MASK & *(u32*)skb->nh.raw) && np->rxopt.bits.rxflow) ||
+ (opt->srcrt && (np->rxopt.bits.srcrt || np->rxopt.bits.osrcrt)) ||
+ ((opt->dst1 || opt->dst0) && (np->rxopt.bits.dstopts || np->rxopt.bits.odstopts)))
return 1;
}
return 0;
buff->csum);
fl.proto = IPPROTO_TCP;
- fl.oif = tcp_v6_iif(skb);
+ fl.oif = inet6_iif(skb);
fl.fl_ip_dport = t1->dest;
fl.fl_ip_sport = t1->source;
/* sk = NULL, but it is safe for now. RST socket required. */
if (!ip6_dst_lookup(NULL, &buff->dst, &fl)) {
- if ((xfrm_lookup(&buff->dst, &fl, NULL, 0)) < 0) {
- dst_release(buff->dst);
+ if ((xfrm_lookup(&buff->dst, &fl, NULL, 0)) < 0)
return;
- }
ip6_xmit(NULL, buff, &fl, NULL, 0);
TCP_INC_STATS_BH(TCP_MIB_OUTSEGS);
buff->csum);
fl.proto = IPPROTO_TCP;
- fl.oif = tcp_v6_iif(skb);
+ fl.oif = inet6_iif(skb);
fl.fl_ip_dport = t1->dest;
fl.fl_ip_sport = t1->source;
if (!ip6_dst_lookup(NULL, &buff->dst, &fl)) {
- if ((xfrm_lookup(&buff->dst, &fl, NULL, 0)) < 0) {
- dst_release(buff->dst);
+ if ((xfrm_lookup(&buff->dst, &fl, NULL, 0)) < 0)
return;
- }
ip6_xmit(NULL, buff, &fl, NULL, 0);
TCP_INC_STATS_BH(TCP_MIB_OUTSEGS);
return;
static struct sock *tcp_v6_hnd_req(struct sock *sk,struct sk_buff *skb)
{
struct request_sock *req, **prev;
- struct tcphdr *th = skb->h.th;
+ const struct tcphdr *th = skb->h.th;
struct sock *nsk;
/* Find possible connection requests. */
req = tcp_v6_search_req(sk, &prev, th->source, &skb->nh.ipv6h->saddr,
- &skb->nh.ipv6h->daddr, tcp_v6_iif(skb));
+ &skb->nh.ipv6h->daddr, inet6_iif(skb));
if (req)
return tcp_check_req(sk, skb, req, prev);
- nsk = __tcp_v6_lookup_established(&skb->nh.ipv6h->saddr,
- th->source,
- &skb->nh.ipv6h->daddr,
- ntohs(th->dest),
- tcp_v6_iif(skb));
+ nsk = __inet6_lookup_established(&tcp_hashinfo, &skb->nh.ipv6h->saddr,
+ th->source, &skb->nh.ipv6h->daddr,
+ ntohs(th->dest), inet6_iif(skb));
if (nsk) {
if (nsk->sk_state != TCP_TIME_WAIT) {
TCP_ECN_create_request(req, skb->h.th);
treq->pktopts = NULL;
if (ipv6_opt_accepted(sk, skb) ||
- np->rxopt.bits.rxinfo ||
- np->rxopt.bits.rxhlim) {
+ np->rxopt.bits.rxinfo || np->rxopt.bits.rxoinfo ||
+ np->rxopt.bits.rxhlim || np->rxopt.bits.rxohlim) {
atomic_inc(&skb->users);
treq->pktopts = skb;
}
/* So that link locals have meaning */
if (!sk->sk_bound_dev_if &&
ipv6_addr_type(&treq->rmt_addr) & IPV6_ADDR_LINKLOCAL)
- treq->iif = tcp_v6_iif(skb);
+ treq->iif = inet6_iif(skb);
if (isn == 0)
isn = tcp_v6_init_sequence(sk,skb);
newsk->sk_backlog_rcv = tcp_v4_do_rcv;
newnp->pktoptions = NULL;
newnp->opt = NULL;
- newnp->mcast_oif = tcp_v6_iif(skb);
+ newnp->mcast_oif = inet6_iif(skb);
newnp->mcast_hops = skb->nh.ipv6h->hop_limit;
/*
if (sk_acceptq_is_full(sk))
goto out_overflow;
- if (np->rxopt.bits.srcrt == 2 &&
+ if (np->rxopt.bits.osrcrt == 2 &&
opt == NULL && treq->pktopts) {
struct inet6_skb_parm *rxopt = IP6CB(treq->pktopts);
if (rxopt->srcrt)
skb_set_owner_r(newnp->pktoptions, newsk);
}
newnp->opt = NULL;
- newnp->mcast_oif = tcp_v6_iif(skb);
+ newnp->mcast_oif = inet6_iif(skb);
newnp->mcast_hops = skb->nh.ipv6h->hop_limit;
/* Clone native IPv6 options from listening socket (if any)
tp = tcp_sk(sk);
if (TCP_SKB_CB(opt_skb)->end_seq == tp->rcv_nxt &&
!((1 << sk->sk_state) & (TCPF_CLOSE | TCPF_LISTEN))) {
- if (np->rxopt.bits.rxinfo)
- np->mcast_oif = tcp_v6_iif(opt_skb);
- if (np->rxopt.bits.rxhlim)
+ if (np->rxopt.bits.rxinfo || np->rxopt.bits.rxoinfo)
+ np->mcast_oif = inet6_iif(opt_skb);
+ if (np->rxopt.bits.rxhlim || np->rxopt.bits.rxohlim)
np->mcast_hops = opt_skb->nh.ipv6h->hop_limit;
if (ipv6_opt_accepted(sk, opt_skb)) {
skb_set_owner_r(opt_skb, sk);
TCP_SKB_CB(skb)->flags = ipv6_get_dsfield(skb->nh.ipv6h);
TCP_SKB_CB(skb)->sacked = 0;
- sk = __tcp_v6_lookup(&skb->nh.ipv6h->saddr, th->source,
- &skb->nh.ipv6h->daddr, ntohs(th->dest), tcp_v6_iif(skb));
+ sk = __inet6_lookup(&tcp_hashinfo, &skb->nh.ipv6h->saddr, th->source,
+ &skb->nh.ipv6h->daddr, ntohs(th->dest),
+ inet6_iif(skb));
if (!sk)
goto no_tcp_socket;
{
struct sock *sk2;
- sk2 = tcp_v6_lookup_listener(&skb->nh.ipv6h->daddr, ntohs(th->dest), tcp_v6_iif(skb));
+ sk2 = inet6_lookup_listener(&tcp_hashinfo,
+ &skb->nh.ipv6h->daddr,
+ ntohs(th->dest), inet6_iif(skb));
if (sk2 != NULL) {
struct inet_timewait_sock *tw = inet_twsk(sk);
inet_twsk_deschedule(tw, &tcp_death_row);
if ((err = xfrm_lookup(&dst, &fl, sk, 0)) < 0) {
sk->sk_err_soft = -err;
- dst_release(dst);
return err;
}
if ((err = xfrm_lookup(&dst, &fl, sk, 0)) < 0) {
sk->sk_route_caps = 0;
- dst_release(dst);
return err;
}
*/
static int tcp_v6_init_sock(struct sock *sk)
{
+ struct inet_connection_sock *icsk = inet_csk(sk);
struct tcp_sock *tp = tcp_sk(sk);
skb_queue_head_init(&tp->out_of_order_queue);
tcp_init_xmit_timers(sk);
tcp_prequeue_init(tp);
- inet_csk(sk)->icsk_rto = TCP_TIMEOUT_INIT;
+ icsk->icsk_rto = TCP_TIMEOUT_INIT;
tp->mdev = TCP_TIMEOUT_INIT;
/* So many TCP implementations out there (incorrectly) count the
sk->sk_state = TCP_CLOSE;
tp->af_specific = &ipv6_specific;
- tp->ca_ops = &tcp_init_congestion_ops;
+ icsk->icsk_ca_ops = &tcp_init_congestion_ops;
sk->sk_write_space = sk_stream_write_space;
sock_set_flag(sk, SOCK_USE_WRITE_QUEUE);
static int tcp_v6_destroy_sock(struct sock *sk)
{
- extern int tcp_v4_destroy_sock(struct sock *sk);
-
tcp_v4_destroy_sock(sk);
return inet6_destroy_sock(sk);
}
jiffies_to_clock_t(timer_expires - jiffies),
icsk->icsk_retransmits,
sock_i_uid(sp),
- tp->probes_out,
+ icsk->icsk_probes_out,
sock_i_ino(sp),
atomic_read(&sp->sk_refcnt), sp,
icsk->icsk_rto,
.flags = INET6_PROTO_NOPOLICY|INET6_PROTO_FINAL,
};
-extern struct proto_ops inet6_stream_ops;
-
static struct inet_protosw tcpv6_protosw = {
.type = SOCK_STREAM,
.protocol = IPPROTO_TCP,