/*
* The three way handshake has completed - we got a valid ACK or DATAACK -
* now create the new socket.
*
* This is the equivalent of TCP's tcp_v4_syn_recv_sock
*/
struct sock *dccp_v4_request_recv_sock(const struct sock *sk,
struct sk_buff *skb,
struct request_sock *req,
struct dst_entry *dst,
struct request_sock *req_unhash,
bool *own_req)
{
struct inet_request_sock *ireq;
struct inet_sock *newinet;
struct sock *newsk;
if (sk_acceptq_is_full(sk))
goto exit_overflow;
newsk = dccp_create_openreq_child(sk, req, skb);
if (newsk == NULL)
goto exit_nonewsk;
newinet = inet_sk(newsk);
ireq = inet_rsk(req);
sk_daddr_set(newsk, ireq->ir_rmt_addr);
sk_rcv_saddr_set(newsk, ireq->ir_loc_addr);
newinet->inet_saddr = ireq->ir_loc_addr;
RCU_INIT_POINTER(newinet->inet_opt, rcu_dereference(ireq->ireq_opt));
newinet->mc_index = inet_iif(skb);
newinet->mc_ttl = ip_hdr(skb)->ttl;
newinet->inet_id = jiffies;
if (dst == NULL && (dst = inet_csk_route_child_sock(sk, newsk, req)) == NULL)
goto put_and_exit;
sk_setup_caps(newsk, dst);
dccp_sync_mss(newsk, dst_mtu(dst));
if (__inet_inherit_port(sk, newsk) < 0)
goto put_and_exit;
*own_req = inet_ehash_nolisten(newsk, req_to_sk(req_unhash));
if (*own_req)
ireq->ireq_opt = NULL;
else
newinet->inet_opt = NULL;
return newsk;
exit_overflow:
__NET_INC_STATS(sock_net(sk), LINUX_MIB_LISTENOVERFLOWS);
exit_nonewsk:
dst_release(dst);
exit:
__NET_INC_STATS(sock_net(sk), LINUX_MIB_LISTENDROPS);
return NULL;
put_and_exit:
newinet->inet_opt = NULL;
inet_csk_prepare_forced_close(newsk);
dccp_done(newsk);
goto exit;
}
/*
* The three way handshake has completed - we got a valid ACK or DATAACK -
* now create the new socket.
*
* This is the equivalent of TCP's tcp_v4_syn_recv_sock
*/
struct sock *dccp_v4_request_recv_sock(struct sock *sk, struct sk_buff *skb,
struct request_sock *req,
struct dst_entry *dst)
{
struct inet_request_sock *ireq;
struct inet_sock *newinet;
struct sock *newsk;
if (sk_acceptq_is_full(sk))
goto exit_overflow;
newsk = dccp_create_openreq_child(sk, req, skb);
if (newsk == NULL)
goto exit_nonewsk;
newinet = inet_sk(newsk);
ireq = inet_rsk(req);
newinet->inet_daddr = ireq->rmt_addr;
newinet->inet_rcv_saddr = ireq->loc_addr;
newinet->inet_saddr = ireq->loc_addr;
newinet->inet_opt = ireq->opt;
ireq->opt = NULL;
newinet->mc_index = inet_iif(skb);
newinet->mc_ttl = ip_hdr(skb)->ttl;
newinet->inet_id = jiffies;
if (dst == NULL && (dst = inet_csk_route_child_sock(sk, newsk, req)) == NULL)
goto put_and_exit;
sk_setup_caps(newsk, dst);
dccp_sync_mss(newsk, dst_mtu(dst));
if (__inet_inherit_port(sk, newsk) < 0)
goto put_and_exit;
__inet_hash_nolisten(newsk, NULL);
return newsk;
exit_overflow:
NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_LISTENOVERFLOWS);
exit_nonewsk:
dst_release(dst);
exit:
NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_LISTENDROPS);
return NULL;
put_and_exit:
bh_unlock_sock(newsk);
sock_put(newsk);
goto exit;
}
/*
* The three way handshake has completed - we got a valid ACK or DATAACK -
* now create the new socket.
*
* This is the equivalent of TCP's tcp_v4_syn_recv_sock
*/
struct sock *dccp_v4_request_recv_sock(struct sock *sk, struct sk_buff *skb,
struct request_sock *req,
struct dst_entry *dst)
{
struct inet_request_sock *ireq;
struct inet_sock *newinet;
struct dccp_sock *newdp;
struct sock *newsk;
if (sk_acceptq_is_full(sk))
goto exit_overflow;
if (dst == NULL && (dst = inet_csk_route_req(sk, req)) == NULL)
goto exit;
newsk = dccp_create_openreq_child(sk, req, skb);
if (newsk == NULL)
goto exit;
sk_setup_caps(newsk, dst);
newdp = dccp_sk(newsk);
newinet = inet_sk(newsk);
ireq = inet_rsk(req);
newinet->daddr = ireq->rmt_addr;
newinet->rcv_saddr = ireq->loc_addr;
newinet->saddr = ireq->loc_addr;
newinet->opt = ireq->opt;
ireq->opt = NULL;
newinet->mc_index = inet_iif(skb);
newinet->mc_ttl = skb->nh.iph->ttl;
newinet->id = jiffies;
dccp_sync_mss(newsk, dst_mtu(dst));
__inet_hash_nolisten(newsk);
__inet_inherit_port(sk, newsk);
return newsk;
exit_overflow:
NET_INC_STATS_BH(LINUX_MIB_LISTENOVERFLOWS);
exit:
NET_INC_STATS_BH(LINUX_MIB_LISTENDROPS);
dst_release(dst);
return NULL;
}
int dccp_v4_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len)
{
const struct sockaddr_in *usin = (struct sockaddr_in *)uaddr;
struct inet_sock *inet = inet_sk(sk);
struct dccp_sock *dp = dccp_sk(sk);
__be16 orig_sport, orig_dport;
__be32 daddr, nexthop;
struct flowi4 *fl4;
struct rtable *rt;
int err;
struct ip_options_rcu *inet_opt;
dp->dccps_role = DCCP_ROLE_CLIENT;
if (addr_len < sizeof(struct sockaddr_in))
return -EINVAL;
if (usin->sin_family != AF_INET)
return -EAFNOSUPPORT;
nexthop = daddr = usin->sin_addr.s_addr;
inet_opt = rcu_dereference_protected(inet->inet_opt,
sock_owned_by_user(sk));
if (inet_opt != NULL && inet_opt->opt.srr) {
if (daddr == 0)
return -EINVAL;
nexthop = inet_opt->opt.faddr;
}
orig_sport = inet->inet_sport;
orig_dport = usin->sin_port;
fl4 = &inet->cork.fl.u.ip4;
rt = ip_route_connect(fl4, nexthop, inet->inet_saddr,
RT_CONN_FLAGS(sk), sk->sk_bound_dev_if,
IPPROTO_DCCP,
orig_sport, orig_dport, sk);
if (IS_ERR(rt))
return PTR_ERR(rt);
if (rt->rt_flags & (RTCF_MULTICAST | RTCF_BROADCAST)) {
ip_rt_put(rt);
return -ENETUNREACH;
}
if (inet_opt == NULL || !inet_opt->opt.srr)
daddr = fl4->daddr;
if (inet->inet_saddr == 0)
inet->inet_saddr = fl4->saddr;
sk_rcv_saddr_set(sk, inet->inet_saddr);
inet->inet_dport = usin->sin_port;
sk_daddr_set(sk, daddr);
inet_csk(sk)->icsk_ext_hdr_len = 0;
if (inet_opt)
inet_csk(sk)->icsk_ext_hdr_len = inet_opt->opt.optlen;
/*
* Socket identity is still unknown (sport may be zero).
* However we set state to DCCP_REQUESTING and not releasing socket
* lock select source port, enter ourselves into the hash tables and
* complete initialization after this.
*/
dccp_set_state(sk, DCCP_REQUESTING);
err = inet_hash_connect(&dccp_death_row, sk);
if (err != 0)
goto failure;
rt = ip_route_newports(fl4, rt, orig_sport, orig_dport,
inet->inet_sport, inet->inet_dport, sk);
if (IS_ERR(rt)) {
err = PTR_ERR(rt);
rt = NULL;
goto failure;
}
/* OK, now commit destination to socket. */
sk_setup_caps(sk, &rt->dst);
dp->dccps_iss = secure_dccp_sequence_number(inet->inet_saddr,
inet->inet_daddr,
inet->inet_sport,
inet->inet_dport);
inet->inet_id = dp->dccps_iss ^ jiffies;
err = dccp_connect(sk);
rt = NULL;
if (err != 0)
goto failure;
out:
return err;
failure:
/*
* This unhashes the socket and releases the local port, if necessary.
*/
dccp_set_state(sk, DCCP_CLOSED);
ip_rt_put(rt);
sk->sk_route_caps = 0;
inet->inet_dport = 0;
goto out;
}
int dccp_v4_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len)
{
struct inet_sock *inet = inet_sk(sk);
struct dccp_sock *dp = dccp_sk(sk);
const struct sockaddr_in *usin = (struct sockaddr_in *)uaddr;
struct rtable *rt;
__be32 daddr, nexthop;
int tmp;
int err;
dp->dccps_role = DCCP_ROLE_CLIENT;
if (addr_len < sizeof(struct sockaddr_in))
return -EINVAL;
if (usin->sin_family != AF_INET)
return -EAFNOSUPPORT;
nexthop = daddr = usin->sin_addr.s_addr;
if (inet->opt != NULL && inet->opt->srr) {
if (daddr == 0)
return -EINVAL;
nexthop = inet->opt->faddr;
}
tmp = ip_route_connect(&rt, nexthop, inet->inet_saddr,
RT_CONN_FLAGS(sk), sk->sk_bound_dev_if,
IPPROTO_DCCP,
inet->inet_sport, usin->sin_port, sk, 1);
if (tmp < 0)
return tmp;
if (rt->rt_flags & (RTCF_MULTICAST | RTCF_BROADCAST)) {
ip_rt_put(rt);
return -ENETUNREACH;
}
if (inet->opt == NULL || !inet->opt->srr)
daddr = rt->rt_dst;
if (inet->inet_saddr == 0)
inet->inet_saddr = rt->rt_src;
inet->inet_rcv_saddr = inet->inet_saddr;
inet->inet_dport = usin->sin_port;
inet->inet_daddr = daddr;
inet_csk(sk)->icsk_ext_hdr_len = 0;
if (inet->opt != NULL)
inet_csk(sk)->icsk_ext_hdr_len = inet->opt->optlen;
/*
* Socket identity is still unknown (sport may be zero).
* However we set state to DCCP_REQUESTING and not releasing socket
* lock select source port, enter ourselves into the hash tables and
* complete initialization after this.
*/
dccp_set_state(sk, DCCP_REQUESTING);
err = inet_hash_connect(&dccp_death_row, sk);
if (err != 0)
goto failure;
err = ip_route_newports(&rt, IPPROTO_DCCP, inet->inet_sport,
inet->inet_dport, sk);
if (err != 0)
goto failure;
/* OK, now commit destination to socket. */
sk_setup_caps(sk, &rt->u.dst);
dp->dccps_iss = secure_dccp_sequence_number(inet->inet_saddr,
inet->inet_daddr,
inet->inet_sport,
inet->inet_dport);
inet->inet_id = dp->dccps_iss ^ jiffies;
err = dccp_connect(sk);
rt = NULL;
if (err != 0)
goto failure;
out:
return err;
failure:
/*
* This unhashes the socket and releases the local port, if necessary.
*/
dccp_set_state(sk, DCCP_CLOSED);
ip_rt_put(rt);
sk->sk_route_caps = 0;
inet->inet_dport = 0;
goto out;
}
/* This will initiate an outgoing connection. */
int tcp_v4_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len)
{
struct inet_sock *inet = inet_sk(sk);
struct tcp_sock *tp = tcp_sk(sk);
struct sockaddr_in *usin = (struct sockaddr_in *)uaddr;
struct rtable *rt;
u32 daddr, nexthop;
int tmp;
int err;
if (addr_len < sizeof(struct sockaddr_in))
return -EINVAL;
if (usin->sin_family != AF_INET)
return -EAFNOSUPPORT;
nexthop = daddr = usin->sin_addr.s_addr;
if (inet->opt && inet->opt->srr) {
if (!daddr)
return -EINVAL;
nexthop = inet->opt->faddr;
}
tmp = ip_route_connect(&rt, nexthop, inet->saddr,
RT_CONN_FLAGS(sk), sk->sk_bound_dev_if,
IPPROTO_TCP,
inet->sport, usin->sin_port, sk);
if (tmp < 0)
return tmp;
if (rt->rt_flags & (RTCF_MULTICAST | RTCF_BROADCAST)) {
ip_rt_put(rt);
return -ENETUNREACH;
}
if (!inet->opt || !inet->opt->srr)
daddr = rt->rt_dst;
if (!inet->saddr)
inet->saddr = rt->rt_src;
inet->rcv_saddr = inet->saddr;
if (tp->rx_opt.ts_recent_stamp && inet->daddr != daddr) {
/* Reset inherited state */
tp->rx_opt.ts_recent = 0;
tp->rx_opt.ts_recent_stamp = 0;
tp->write_seq = 0;
}
if (tcp_death_row.sysctl_tw_recycle &&
!tp->rx_opt.ts_recent_stamp && rt->rt_dst == daddr) {
struct inet_peer *peer = rt_get_peer(rt);
/* VJ's idea. We save last timestamp seen from
* the destination in peer table, when entering state TIME-WAIT
* and initialize rx_opt.ts_recent from it, when trying new connection.
*/
if (peer && peer->tcp_ts_stamp + TCP_PAWS_MSL >= xtime.tv_sec) {
tp->rx_opt.ts_recent_stamp = peer->tcp_ts_stamp;
tp->rx_opt.ts_recent = peer->tcp_ts;
}
}
inet->dport = usin->sin_port;
inet->daddr = daddr;
inet_csk(sk)->icsk_ext_hdr_len = 0;
if (inet->opt)
inet_csk(sk)->icsk_ext_hdr_len = inet->opt->optlen;
tp->rx_opt.mss_clamp = 536;
/* Socket identity is still unknown (sport may be zero).
* However we set state to SYN-SENT and not releasing socket
* lock select source port, enter ourselves into the hash tables and
* complete initialization after this.
*/
tcp_set_state(sk, TCP_SYN_SENT);
err = inet_hash_connect(&tcp_death_row, sk);
if (err)
goto failure;
err = ip_route_newports(&rt, IPPROTO_TCP, inet->sport, inet->dport, sk);
if (err)
goto failure;
/* OK, now commit destination to socket. */
sk->sk_gso_type = SKB_GSO_TCPV4;
sk_setup_caps(sk, &rt->u.dst);
if (!tp->write_seq)
tp->write_seq = secure_tcp_sequence_number(inet->saddr,
inet->daddr,
inet->sport,
usin->sin_port);
inet->id = tp->write_seq ^ jiffies;
err = tcp_connect(sk);
rt = NULL;
if (err)
goto failure;
return 0;
failure:
/* This unhashes the socket and releases the local port, if necessary. */
tcp_set_state(sk, TCP_CLOSE);
ip_rt_put(rt);
sk->sk_route_caps = 0;
inet->dport = 0;
return err;
}
int dccp_v4_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len)
{
struct inet_sock *inet = inet_sk(sk);
struct dccp_sock *dp = dccp_sk(sk);
const struct sockaddr_in *usin = (struct sockaddr_in *)uaddr;
struct rtable *rt;
__be32 daddr, nexthop;
int tmp;
int err;
dp->dccps_role = DCCP_ROLE_CLIENT;
if (addr_len < sizeof(struct sockaddr_in))
return -EINVAL;
if (usin->sin_family != AF_INET)
return -EAFNOSUPPORT;
nexthop = daddr = usin->sin_addr.s_addr;
if (inet->opt != NULL && inet->opt->srr) {
if (daddr == 0)
return -EINVAL;
nexthop = inet->opt->faddr;
}
tmp = ip_route_connect(&rt, nexthop, inet->saddr,
RT_CONN_FLAGS(sk), sk->sk_bound_dev_if,
IPPROTO_DCCP,
inet->sport, usin->sin_port, sk, 1);
if (tmp < 0)
return tmp;
if (rt->rt_flags & (RTCF_MULTICAST | RTCF_BROADCAST)) {
ip_rt_put(rt);
return -ENETUNREACH;
}
if (inet->opt == NULL || !inet->opt->srr)
daddr = rt->rt_dst;
if (inet->saddr == 0)
inet->saddr = rt->rt_src;
inet->rcv_saddr = inet->saddr;
inet->dport = usin->sin_port;
inet->daddr = daddr;
inet_csk(sk)->icsk_ext_hdr_len = 0;
if (inet->opt != NULL)
inet_csk(sk)->icsk_ext_hdr_len = inet->opt->optlen;
dccp_set_state(sk, DCCP_REQUESTING);
err = inet_hash_connect(&dccp_death_row, sk);
if (err != 0)
goto failure;
err = ip_route_newports(&rt, IPPROTO_DCCP, inet->sport, inet->dport,
sk);
if (err != 0)
goto failure;
sk_setup_caps(sk, &rt->u.dst);
dp->dccps_iss = secure_dccp_sequence_number(inet->saddr, inet->daddr,
inet->sport, inet->dport);
inet->id = dp->dccps_iss ^ jiffies;
err = dccp_connect(sk);
rt = NULL;
if (err != 0)
goto failure;
out:
return err;
failure:
dccp_set_state(sk, DCCP_CLOSED);
ip_rt_put(rt);
sk->sk_route_caps = 0;
inet->dport = 0;
goto out;
}
int serval_ipv4_xmit(struct sk_buff *skb)
{
struct sock *sk = skb->sk;
int err = 0;
#if defined(OS_LINUX_KERNEL)
/*
This is pretty much a copy paste from ip_queue_xmit
(ip_output.c), but which modifications that take into
account Serval specific stuff.
It will route the packet according to the IP stack's routing
table and output for standard IP output processing.
*/
struct iphdr *iph;
struct rtable *rt;
struct inet_sock *inet = inet_sk(sk);
struct ip_options *opt = NULL; /*inet->inet_opt; */
int ifindex;
/*
The SAL has dirtied the control block that IP expects to be
zeroed out.
We need to make sure it is initialized again. Otherwise, there
might be stack corruptions when IP functions try to read the
IPCB. (This happens in, e.g., icmp_send when reading ip options.)
*/
memset(IPCB(skb), 0, sizeof(struct inet_skb_parm));
/*
* Skip all of this if the packet is already routed,
*/
rcu_read_lock();
rt = skb_rtable(skb);
if (rt != NULL) {
LOG_PKT("Packet already routed\n");
goto packet_routed;
}
/* Make sure we can route this packet. */
rt = (struct rtable *)__sk_dst_check(sk, 0);
if (skb->dev) {
ifindex = skb->dev->ifindex;
} else {
ifindex = sk->sk_bound_dev_if;
}
if (rt == NULL) {
struct flowi fl;
serval_flow_init_output(&fl, ifindex,
sk->sk_mark,
RT_CONN_FLAGS(sk), 0,
skb->protocol,
#if (LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,28))
inet_sk_flowi_flags(sk),
#else
0,
#endif
inet->inet_daddr,
inet->inet_saddr,
0, 0);
serval_security_sk_classify_flow(sk, &fl);
rt = serval_ip_route_output_flow(sock_net(sk), &fl, sk, 0);
if (!rt) {
LOG_DBG("No route!\n");
err = -EHOSTUNREACH;
rcu_read_unlock();
goto drop;
}
/* Setup the socket to use this route in the future */
sk_setup_caps(sk, route_dst(rt));
} else {
LOG_PKT("Using route already associated with socket\n");
}
#if defined(ENABLE_DEBUG)
{
char src[18], dst[18];
LOG_PKT("Route found %s -> %s %s\n",
inet_ntop(AF_INET, &rt->rt_src,
src, sizeof(src)),
inet_ntop(AF_INET, &rt->rt_dst,
dst, sizeof(dst)),
route_dst(rt)->dev ?
route_dst(rt)->dev->name : "(null)");
}
#endif
#if (LINUX_VERSION_CODE < KERNEL_VERSION(2, 6, 35))
skb_dst_set(skb, dst_clone(route_dst(rt)));
#else
skb_dst_set_noref(skb, route_dst(rt));
#endif
packet_routed:
if (opt && opt->is_strictroute && rt->rt_dst != rt->rt_gateway) {
err = -EHOSTUNREACH;
rcu_read_unlock();
LOG_DBG("dest is not gateway!\n");
goto drop;
}
/* OK, we know where to send it, allocate and build IP header. */
skb_push(skb, sizeof(struct iphdr) + (opt ? opt->optlen : 0));
skb_reset_network_header(skb);
iph = ip_hdr(skb);
*((__be16 *)iph) = htons((4 << 12) | (5 << 8) | (inet->tos & 0xff));
if (ip_dont_fragment(sk, route_dst(rt)) && !skb->local_df)
iph->frag_off = htons(IP_DF);
else
iph->frag_off = 0;
iph->ttl = ip_select_ttl(inet, route_dst(rt));
iph->protocol = skb->protocol;
iph->saddr = rt->rt_src;
iph->daddr = rt->rt_dst;
if (opt && opt->optlen) {
LOG_WARN("IP options not implemented\n");
/* For some reason, enabling the code below gives the
* error: "Unknown symbol ip_options_build (err 0)"
* when loading the serval.ko module. Seems the
* ip_options_build function is not exported.
*/
/*
iph->ihl += opt->optlen >> 2;
ip_options_build(skb, opt, inet->inet_daddr, rt, 0);
*/
}
ip_select_ident_more(iph, route_dst(rt), sk,
(skb_shinfo(skb)->gso_segs ?: 1) - 1);
skb->priority = sk->sk_priority;
#if (LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,25))
skb->mark = sk->sk_mark;
#endif
err = serval_ip_local_out(skb);
rcu_read_unlock();
#else
/*
FIXME: We should not rely on an outgoing interface here.
Instead, we should route the packet like we do in the
kernel. But, we currently do not have an IP routing table
for userlevel.
*/
if (!skb->dev)
skb->dev = __dev_get_by_index(sock_net(sk),
sk->sk_bound_dev_if);
if (!skb->dev) {
LOG_ERR("no output device set in skb!\n");
err = -ENODEV;
goto drop;
}
err = serval_ipv4_fill_in_hdr(sk, skb, inet_sk(sk)->inet_saddr,
inet_sk(sk)->inet_daddr);
if (err < 0) {
LOG_ERR("hdr failed\n");
goto drop;
}
/* Transmit */
err = serval_ip_local_out(skb);
#endif /* OS_LINUX_KERNEL */
out:
if (err < 0) {
LOG_ERR("xmit failed: %d\n", err);
}
return err;
drop:
LOG_DBG("Dropping skb!\n");
FREE_SKB(skb);
goto out;
}
