/*
* Outer subroutine:
* Connect from a socket to a specified address.
* Both address and port must be specified in argument sin.
* If don't have a local address for this socket yet,
* then pick one.
*/
int
in_pcbconnect(struct inpcb *inp, struct mbuf *nam)
{
struct sockaddr_in *ifaddr;
register struct sockaddr_in *sin = mtod(nam, struct sockaddr_in *);
int error;
/*
* Call inner routine, to assign local interface address.
*/
if ((error = in_pcbladdr(inp, nam, &ifaddr)))
return(error);
if (in_pcblookuphash(inp->inp_pcbinfo, sin->sin_addr, sin->sin_port,
inp->inp_laddr.s_addr ? inp->inp_laddr : ifaddr->sin_addr,
inp->inp_lport, 0) != NULL)
return (EADDRINUSE);
if (inp->inp_laddr.s_addr == INADDR_ANY) {
if (inp->inp_lport == 0)
(void)in_pcbbind(inp, (struct mbuf *)0);
inp->inp_laddr = ifaddr->sin_addr;
}
inp->inp_faddr = sin->sin_addr;
inp->inp_fport = sin->sin_port;
in_pcbrehash(inp);
return (0);
}
/*
* Outer subroutine:
* Connect from a socket to a specified address.
* Both address and port must be specified in argument sin.
* If don't have a local address for this socket yet,
* then pick one.
*/
int
in6_pcbconnect(
struct inpcb *inp,
struct sockaddr *nam,
struct proc *p)
{
struct in6_addr addr6;
struct sockaddr_in6 *sin6 = (struct sockaddr_in6 *)nam;
struct inpcb *pcb;
int error;
unsigned int outif = 0;
/*
* Call inner routine, to assign local interface address.
* in6_pcbladdr() may automatically fill in sin6_scope_id.
*/
if ((error = in6_pcbladdr(inp, nam, &addr6, &outif)) != 0)
return(error);
socket_unlock(inp->inp_socket, 0);
pcb = in6_pcblookup_hash(inp->inp_pcbinfo, &sin6->sin6_addr,
sin6->sin6_port,
IN6_IS_ADDR_UNSPECIFIED(&inp->in6p_laddr)
? &addr6 : &inp->in6p_laddr,
inp->inp_lport, 0, NULL);
socket_lock(inp->inp_socket, 0);
if (pcb != NULL) {
in_pcb_checkstate(pcb, WNT_RELEASE, pcb == inp ? 1 : 0);
return (EADDRINUSE);
}
if (IN6_IS_ADDR_UNSPECIFIED(&inp->in6p_laddr)) {
if (inp->inp_lport == 0) {
error = in6_pcbbind(inp, (struct sockaddr *)0, p);
if (error)
return (error);
}
inp->in6p_laddr = addr6;
inp->in6p_last_outif = outif;
}
if (!lck_rw_try_lock_exclusive(inp->inp_pcbinfo->mtx)) {
/*lock inversion issue, mostly with udp multicast packets */
socket_unlock(inp->inp_socket, 0);
lck_rw_lock_exclusive(inp->inp_pcbinfo->mtx);
socket_lock(inp->inp_socket, 0);
}
inp->in6p_faddr = sin6->sin6_addr;
inp->inp_fport = sin6->sin6_port;
/* update flowinfo - draft-itojun-ipv6-flowlabel-api-00 */
inp->in6p_flowinfo &= ~IPV6_FLOWLABEL_MASK;
if (inp->in6p_flags & IN6P_AUTOFLOWLABEL)
inp->in6p_flowinfo |=
(htonl(ip6_flow_seq++) & IPV6_FLOWLABEL_MASK);
in_pcbrehash(inp);
lck_rw_done(inp->inp_pcbinfo->mtx);
return (0);
}
void
in_pcbdisconnect(struct inpcb *inp)
{
inp->inp_faddr.s_addr = INADDR_ANY;
inp->inp_fport = 0;
in_pcbrehash(inp);
if (inp->inp_socket->so_state & SS_NOFDREF)
in_pcbdetach(inp);
}
/*
* Common subroutine to open a TCP connection to remote host specified
* by struct sockaddr_in in mbuf *nam. Call in_pcbbind to assign a local
* port number if needed. Call in_pcbconnect_setup to do the routing and
* to choose a local host address (interface). If there is an existing
* incarnation of the same connection in TIME-WAIT state and if the remote
* host was sending CC options and if the connection duration was < MSL, then
* truncate the previous TIME-WAIT state and proceed.
* Initialize connection parameters and enter SYN-SENT state.
*/
static int
tcp_connect(struct tcpcb *tp, struct sockaddr *nam, struct thread *td)
{
struct inpcb *inp = tp->t_inpcb, *oinp;
struct socket *so = inp->inp_socket;
struct in_addr laddr;
u_short lport;
int error;
INP_INFO_WLOCK_ASSERT(&V_tcbinfo);
INP_WLOCK_ASSERT(inp);
if (inp->inp_lport == 0) {
error = in_pcbbind(inp, (struct sockaddr *)0, td->td_ucred);
if (error)
return error;
}
/*
* Cannot simply call in_pcbconnect, because there might be an
* earlier incarnation of this same connection still in
* TIME_WAIT state, creating an ADDRINUSE error.
*/
laddr = inp->inp_laddr;
lport = inp->inp_lport;
error = in_pcbconnect_setup(inp, nam, &laddr.s_addr, &lport,
&inp->inp_faddr.s_addr, &inp->inp_fport, &oinp, td->td_ucred);
if (error && oinp == NULL)
return error;
if (oinp)
return EADDRINUSE;
inp->inp_laddr = laddr;
in_pcbrehash(inp);
/*
* Compute window scaling to request:
* Scale to fit into sweet spot. See tcp_syncache.c.
* XXX: This should move to tcp_output().
*/
while (tp->request_r_scale < TCP_MAX_WINSHIFT &&
(TCP_MAXWIN << tp->request_r_scale) < sb_max)
tp->request_r_scale++;
soisconnecting(so);
TCPSTAT_INC(tcps_connattempt);
tp->t_state = TCPS_SYN_SENT;
tcp_timer_activate(tp, TT_KEEP, tcp_keepinit);
tp->iss = tcp_new_isn(tp);
tp->t_bw_rtseq = tp->iss;
tcp_sendseqinit(tp);
return 0;
}
void
in6_pcbdisconnect(
struct inpcb *inp)
{
if (!lck_rw_try_lock_exclusive(inp->inp_pcbinfo->mtx)) {
/*lock inversion issue, mostly with udp multicast packets */
socket_unlock(inp->inp_socket, 0);
lck_rw_lock_exclusive(inp->inp_pcbinfo->mtx);
socket_lock(inp->inp_socket, 0);
}
bzero((caddr_t)&inp->in6p_faddr, sizeof(inp->in6p_faddr));
inp->inp_fport = 0;
/* clear flowinfo - draft-itojun-ipv6-flowlabel-api-00 */
inp->in6p_flowinfo &= ~IPV6_FLOWLABEL_MASK;
in_pcbrehash(inp);
lck_rw_done(inp->inp_pcbinfo->mtx);
if (inp->inp_socket->so_state & SS_NOFDREF)
in6_pcbdetach(inp);
}
static int
tcp6_connect(struct tcpcb *tp, struct sockaddr *nam, struct thread *td)
{
struct inpcb *inp = tp->t_inpcb, *oinp;
struct socket *so = inp->inp_socket;
struct sockaddr_in6 *sin6 = (struct sockaddr_in6 *)nam;
struct in6_addr addr6;
int error;
INP_INFO_WLOCK_ASSERT(&V_tcbinfo);
INP_WLOCK_ASSERT(inp);
if (inp->inp_lport == 0) {
error = in6_pcbbind(inp, (struct sockaddr *)0, td->td_ucred);
if (error)
return error;
}
/*
* Cannot simply call in_pcbconnect, because there might be an
* earlier incarnation of this same connection still in
* TIME_WAIT state, creating an ADDRINUSE error.
* in6_pcbladdr() also handles scope zone IDs.
*/
error = in6_pcbladdr(inp, nam, &addr6);
if (error)
return error;
oinp = in6_pcblookup_hash(inp->inp_pcbinfo,
&sin6->sin6_addr, sin6->sin6_port,
IN6_IS_ADDR_UNSPECIFIED(&inp->in6p_laddr)
? &addr6
: &inp->in6p_laddr,
inp->inp_lport, 0, NULL);
if (oinp)
return EADDRINUSE;
if (IN6_IS_ADDR_UNSPECIFIED(&inp->in6p_laddr))
inp->in6p_laddr = addr6;
inp->in6p_faddr = sin6->sin6_addr;
inp->inp_fport = sin6->sin6_port;
/* update flowinfo - draft-itojun-ipv6-flowlabel-api-00 */
inp->inp_flow &= ~IPV6_FLOWLABEL_MASK;
if (inp->inp_flags & IN6P_AUTOFLOWLABEL)
inp->inp_flow |=
(htonl(ip6_randomflowlabel()) & IPV6_FLOWLABEL_MASK);
in_pcbrehash(inp);
/* Compute window scaling to request. */
while (tp->request_r_scale < TCP_MAX_WINSHIFT &&
(TCP_MAXWIN << tp->request_r_scale) < sb_max)
tp->request_r_scale++;
soisconnecting(so);
TCPSTAT_INC(tcps_connattempt);
tp->t_state = TCPS_SYN_SENT;
tcp_timer_activate(tp, TT_KEEP, tcp_keepinit);
tp->iss = tcp_new_isn(tp);
tp->t_bw_rtseq = tp->iss;
tcp_sendseqinit(tp);
return 0;
}
int
in_pcbbind(struct inpcb *inp, struct mbuf *nam)
{
register struct socket *so = inp->inp_socket;
unsigned short *lastport;
struct sockaddr_in *sin;
u_short lport = 0;
int wild = 0, reuseport = (so->so_options & SO_REUSEPORT);
int error;
if (in_ifaddr == 0)
return (EADDRNOTAVAIL);
if (inp->inp_lport || inp->inp_laddr.s_addr != INADDR_ANY)
return (EINVAL);
if ((so->so_options & (SO_REUSEADDR|SO_REUSEPORT)) == 0 &&
((so->so_proto->pr_flags & PR_CONNREQUIRED) == 0 ||
(so->so_options & SO_ACCEPTCONN) == 0))
wild = 1;
if (nam) {
sin = mtod(nam, struct sockaddr_in *);
if (nam->m_len != sizeof (*sin))
return (EINVAL);
#ifdef notdef
/*
* We should check the family, but old programs
* incorrectly fail to initialize it.
*/
if (sin->sin_family != AF_INET)
return (EAFNOSUPPORT);
#endif
lport = sin->sin_port;
if (IN_MULTICAST(ntohl(sin->sin_addr.s_addr))) {
/*
* Treat SO_REUSEADDR as SO_REUSEPORT for multicast;
* allow complete duplication of binding if
* SO_REUSEPORT is set, or if SO_REUSEADDR is set
* and a multicast address is bound on both
* new and duplicated sockets.
*/
if (so->so_options & SO_REUSEADDR)
reuseport = SO_REUSEADDR|SO_REUSEPORT;
} else if (sin->sin_addr.s_addr != INADDR_ANY) {
sin->sin_port = 0; /* yech... */
if (ifa_ifwithaddr((struct sockaddr *)sin) == 0)
return (EADDRNOTAVAIL);
}
if (lport) {
struct inpcb *t;
/* GROSS */
if (ntohs(lport) < IPPORT_RESERVED &&
(error = suser(p->p_ucred, &p->p_acflag)))
return (EACCES);
if (so->so_uid) {
t = in_pcblookup(inp->inp_pcbinfo, zeroin_addr,
0, sin->sin_addr, lport,
INPLOOKUP_WILDCARD);
if (t && (so->so_uid != t->inp_socket->so_uid))
return (EADDRINUSE);
}
t = in_pcblookup(inp->inp_pcbinfo, zeroin_addr, 0,
sin->sin_addr, lport, wild);
if (t && (reuseport & t->inp_socket->so_options) == 0)
return (EADDRINUSE);
}
inp->inp_laddr = sin->sin_addr;
}
if (lport == 0) {
unsigned short first, last;
int count;
inp->inp_flags |= INP_ANONPORT;
if (inp->inp_flags & INP_HIGHPORT) {
first = ipport_hifirstauto; /* sysctl */
last = ipport_hilastauto;
lastport = &inp->inp_pcbinfo->lasthi;
} else if (inp->inp_flags & INP_LOWPORT) {
if ((error = suser(p->p_ucred, &p->p_acflag)))
return (EACCES);
first = ipport_lowfirstauto; /* 1023 */
last = ipport_lowlastauto; /* 600 */
lastport = &inp->inp_pcbinfo->lastlow;
} else {
first = ipport_firstauto; /* sysctl */
last = ipport_lastauto;
lastport = &inp->inp_pcbinfo->lastport;
}
/*
* Simple check to ensure all ports are not used up causing
* a deadlock here.
*
* We split the two cases (up and down) so that the direction
* is not being tested on each round of the loop.
*/
if (first > last) {
/*
* counting down
*/
count = first - last;
do {
if (count-- <= 0) /* completely used? */
return (EADDRNOTAVAIL);
--*lastport;
if (*lastport > first || *lastport < last)
*lastport = first;
lport = htons(*lastport);
} while (in_pcblookup(inp->inp_pcbinfo,
zeroin_addr, 0, inp->inp_laddr, lport, wild));
} else {
/*
* counting up
*/
count = last - first;
do {
if (count-- <= 0) /* completely used? */
return (EADDRNOTAVAIL);
++*lastport;
if (*lastport < first || *lastport > last)
*lastport = first;
lport = htons(*lastport);
} while (in_pcblookup(inp->inp_pcbinfo,
zeroin_addr, 0, inp->inp_laddr, lport, wild));
}
}
inp->inp_lport = lport;
in_pcbrehash(inp);
return (0);
}
/*
* Connect from a socket to a specified address.
* Both address and port must be specified in argument sin6.
* Eventually, flow labels will have to be dealt with here, as well.
*
* If don't have a local address for this socket yet,
* then pick one.
*
* I believe this has to be called at splnet().
*/
int
in6_pcbconnect(struct inpcb *inp, struct mbuf *nam)
{
struct in6_addr *in6a = NULL;
struct sockaddr_in6 *sin6 = mtod(nam, struct sockaddr_in6 *);
struct ifnet *ifp = NULL; /* outgoing interface */
int error = 0;
struct sockaddr_in6 tmp;
(void)&in6a; /* XXX fool gcc */
if (nam->m_len != sizeof(*sin6))
return (EINVAL);
if (sin6->sin6_family != AF_INET6)
return (EAFNOSUPPORT);
if (sin6->sin6_port == 0)
return (EADDRNOTAVAIL);
/* reject IPv4 mapped address, we have no support for it */
if (IN6_IS_ADDR_V4MAPPED(&sin6->sin6_addr))
return EADDRNOTAVAIL;
/* sanity check for mapped address case */
if (IN6_IS_ADDR_V4MAPPED(&inp->inp_laddr6))
return EINVAL;
/* protect *sin6 from overwrites */
tmp = *sin6;
sin6 = &tmp;
/* KAME hack: embed scopeid */
if (in6_embedscope(&sin6->sin6_addr, sin6, inp, &ifp) != 0)
return EINVAL;
/* this must be cleared for ifa_ifwithaddr() */
sin6->sin6_scope_id = 0;
/* Source address selection. */
/*
* XXX: in6_selectsrc might replace the bound local address
* with the address specified by setsockopt(IPV6_PKTINFO).
* Is it the intended behavior?
*/
in6a = in6_selectsrc(sin6, inp->inp_outputopts6,
inp->inp_moptions6, &inp->inp_route6, &inp->inp_laddr6,
&error);
if (in6a == 0) {
if (error == 0)
error = EADDRNOTAVAIL;
return (error);
}
if (inp->inp_route6.ro_rt)
ifp = inp->inp_route6.ro_rt->rt_ifp;
inp->inp_ipv6.ip6_hlim = (u_int8_t)in6_selecthlim(inp, ifp);
if (in_pcblookup(inp->inp_table, &sin6->sin6_addr, sin6->sin6_port,
IN6_IS_ADDR_UNSPECIFIED(&inp->inp_laddr6) ? in6a : &inp->inp_laddr6,
inp->inp_lport, INPLOOKUP_IPV6)) {
return (EADDRINUSE);
}
if (IN6_IS_ADDR_UNSPECIFIED(&inp->inp_laddr6)) {
if (inp->inp_lport == 0)
(void)in6_pcbbind(inp, NULL, curproc);
inp->inp_laddr6 = *in6a;
}
inp->inp_faddr6 = sin6->sin6_addr;
inp->inp_fport = sin6->sin6_port;
inp->inp_flowinfo &= ~IPV6_FLOWLABEL_MASK;
if (ip6_auto_flowlabel)
inp->inp_flowinfo |=
(htonl(ip6_randomflowlabel()) & IPV6_FLOWLABEL_MASK);
in_pcbrehash(inp);
return (0);
}
int
in6_pcbsetport(struct in6_addr *laddr, struct inpcb *inp, struct proc *p)
{
struct socket *so = inp->inp_socket;
struct inpcbtable *table = inp->inp_table;
u_int16_t first, last;
u_int16_t *lastport = &inp->inp_table->inpt_lastport;
u_int16_t lport = 0;
int count;
int wild = INPLOOKUP_IPV6;
int error;
/* XXX we no longer support IPv4 mapped address, so no tweaks here */
if ((so->so_options & (SO_REUSEADDR|SO_REUSEPORT)) == 0 &&
((so->so_proto->pr_flags & PR_CONNREQUIRED) == 0 ||
(so->so_options & SO_ACCEPTCONN) == 0))
wild |= INPLOOKUP_WILDCARD;
if (inp->inp_flags & INP_HIGHPORT) {
first = ipport_hifirstauto; /* sysctl */
last = ipport_hilastauto;
} else if (inp->inp_flags & INP_LOWPORT) {
if ((error = suser(p, 0)))
return (EACCES);
first = IPPORT_RESERVED-1; /* 1023 */
last = 600; /* not IPPORT_RESERVED/2 */
} else {
first = ipport_firstauto; /* sysctl */
last = ipport_lastauto;
}
/*
* Simple check to ensure all ports are not used up causing
* a deadlock here.
*
* We split the two cases (up and down) so that the direction
* is not being tested on each round of the loop.
*/
if (first > last) {
/*
* counting down
*/
count = first - last;
if (count)
*lastport = first - arc4random_uniform(count);
do {
if (count-- < 0) /* completely used? */
return (EADDRNOTAVAIL);
--*lastport;
if (*lastport > first || *lastport < last)
*lastport = first;
lport = htons(*lastport);
} while (in_baddynamic(*lastport, so->so_proto->pr_protocol) ||
in_pcblookup(table, &zeroin6_addr, 0,
&inp->inp_laddr6, lport, wild));
} else {
/*
* counting up
*/
count = last - first;
if (count)
*lastport = first + arc4random_uniform(count);
do {
if (count-- < 0) /* completely used? */
return (EADDRNOTAVAIL);
++*lastport;
if (*lastport < first || *lastport > last)
*lastport = first;
lport = htons(*lastport);
} while (in_baddynamic(*lastport, so->so_proto->pr_protocol) ||
in_pcblookup(table, &zeroin6_addr, 0,
&inp->inp_laddr6, lport, wild));
}
inp->inp_lport = lport;
in_pcbrehash(inp);
#if 0
inp->inp_flowinfo = 0; /* XXX */
#endif
return 0;
}
/*
* Bind an address (or at least a port) to an PF_INET6 socket.
*/
int
in6_pcbbind(struct inpcb *inp, struct mbuf *nam, struct proc *p)
{
struct socket *so = inp->inp_socket;
struct inpcbtable *head = inp->inp_table;
struct sockaddr_in6 *sin6;
u_short lport = 0;
int wild = INPLOOKUP_IPV6, reuseport = (so->so_options & SO_REUSEPORT);
int error;
/*
* REMINDER: Once up to speed, flow label processing should go here,
* too. (Same with in6_pcbconnect.)
*/
if (in6_ifaddr == 0)
return EADDRNOTAVAIL;
if (inp->inp_lport != 0 || !IN6_IS_ADDR_UNSPECIFIED(&inp->inp_laddr6))
return EINVAL; /* If already bound, EINVAL! */
if ((so->so_options & (SO_REUSEADDR | SO_REUSEPORT)) == 0 &&
((so->so_proto->pr_flags & PR_CONNREQUIRED) == 0 ||
(so->so_options & SO_ACCEPTCONN) == 0))
wild |= INPLOOKUP_WILDCARD;
/*
* If I did get a sockaddr passed in...
*/
if (nam) {
sin6 = mtod(nam, struct sockaddr_in6 *);
if (nam->m_len != sizeof (*sin6))
return EINVAL;
/*
* Unlike v4, I have no qualms about EAFNOSUPPORT if the
* wretched family is not filled in!
*/
if (sin6->sin6_family != AF_INET6)
return EAFNOSUPPORT;
/* KAME hack: embed scopeid */
if (in6_embedscope(&sin6->sin6_addr, sin6, inp, NULL) != 0)
return EINVAL;
/* this must be cleared for ifa_ifwithaddr() */
sin6->sin6_scope_id = 0;
lport = sin6->sin6_port;
/* reject IPv4 mapped address, we have no support for it */
if (IN6_IS_ADDR_V4MAPPED(&sin6->sin6_addr))
return EADDRNOTAVAIL;
if (IN6_IS_ADDR_MULTICAST(&sin6->sin6_addr)) {
/*
* Treat SO_REUSEADDR as SO_REUSEPORT for multicast;
* allow complete duplication of binding if
* SO_REUSEPORT is set, or if SO_REUSEADDR is set
* and a multicast address is bound on both
* new and duplicated sockets.
*/
if (so->so_options & SO_REUSEADDR)
reuseport = SO_REUSEADDR | SO_REUSEPORT;
} else if (!IN6_IS_ADDR_UNSPECIFIED(&sin6->sin6_addr)) {
struct ifaddr *ia = NULL;
sin6->sin6_port = 0; /*
* Yechhhh, because of upcoming
* call to ifa_ifwithaddr(), which
* does bcmp's over the PORTS as
* well. (What about flow?)
*/
sin6->sin6_flowinfo = 0;
if (!(so->so_options & SO_BINDANY) &&
((ia = ifa_ifwithaddr((struct sockaddr *)sin6))
== NULL))
return EADDRNOTAVAIL;
/*
* bind to an anycast address might accidentally
* cause sending a packet with an anycast source
* address, so we forbid it.
*
* We should allow to bind to a deprecated address,
* since the application dare to use it.
* But, can we assume that they are careful enough
* to check if the address is deprecated or not?
* Maybe, as a safeguard, we should have a setsockopt
* flag to control the bind(2) behavior against
* deprecated addresses (default: forbid bind(2)).
*/
if (ia &&
((struct in6_ifaddr *)ia)->ia6_flags &
(IN6_IFF_ANYCAST|IN6_IFF_NOTREADY|IN6_IFF_DETACHED))
return (EADDRNOTAVAIL);
}
if (lport) {
struct inpcb *t;
/*
* Question: Do we wish to continue the Berkeley
* tradition of ports < IPPORT_RESERVED be only for
* root?
* Answer: For now yes, but IMHO, it should be REMOVED!
* OUCH: One other thing, is there no better way of
* finding a process for a socket instead of using
* curproc? (Marked with BSD's {in,}famous XXX ?
*/
if (ntohs(lport) < IPPORT_RESERVED &&
(error = suser(p, 0)))
return error;
t = in_pcblookup(head,
(struct in_addr *)&zeroin6_addr, 0,
(struct in_addr *)&sin6->sin6_addr, lport,
wild);
if (t && (reuseport & t->inp_socket->so_options) == 0)
return EADDRINUSE;
}
inp->inp_laddr6 = sin6->sin6_addr;
}
if (lport == 0) {
error = in6_pcbsetport(&inp->inp_laddr6, inp, p);
if (error != 0)
return error;
} else {
inp->inp_lport = lport;
in_pcbrehash(inp);
}
return 0;
}
