/*      $NetBSD: in6_src.c,v 1.92 2023/08/03 04:24:55 ozaki-r Exp $     */
/*      $KAME: in6_src.c,v 1.159 2005/10/19 01:40:32 t-momose Exp $     */

/*
* Copyright (C) 1995, 1996, 1997, and 1998 WIDE Project.
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
*    notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
*    notice, this list of conditions and the following disclaimer in the
*    documentation and/or other materials provided with the distribution.
* 3. Neither the name of the project nor the names of its contributors
*    may be used to endorse or promote products derived from this software
*    without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE PROJECT AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED.  IN NO EVENT SHALL THE PROJECT OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/

/*
* Copyright (c) 1982, 1986, 1991, 1993
*      The Regents of the University of California.  All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
*    notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
*    notice, this list of conditions and the following disclaimer in the
*    documentation and/or other materials provided with the distribution.
* 3. All advertising materials mentioning features or use of this software
*    must display the following acknowledgement:
*      This product includes software developed by the University of
*      California, Berkeley and its contributors.
* 4. Neither the name of the University nor the names of its contributors
*    may be used to endorse or promote products derived from this software
*    without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*
*      @(#)in_pcb.c    8.2 (Berkeley) 1/4/94
*/

#include <sys/cdefs.h>
__KERNEL_RCSID(0, "$NetBSD: in6_src.c,v 1.92 2023/08/03 04:24:55 ozaki-r Exp $");

#ifdef _KERNEL_OPT
#include "opt_inet.h"
#endif

#include <sys/param.h>
#include <sys/systm.h>
#include <sys/malloc.h>
#include <sys/mbuf.h>
#include <sys/protosw.h>
#include <sys/socket.h>
#include <sys/socketvar.h>
#include <sys/ioctl.h>
#include <sys/errno.h>
#include <sys/time.h>
#include <sys/kernel.h>
#include <sys/proc.h>
#include <sys/kauth.h>

#include <net/if.h>
#include <net/if_types.h>
#include <net/route.h>

#include <netinet/in.h>
#include <netinet/in_var.h>
#include <netinet/in_systm.h>
#include <netinet/ip.h>
#include <netinet/in_pcb.h>
#include <netinet/portalgo.h>
#include <netinet6/in6_var.h>
#include <netinet/ip6.h>
#include <netinet6/in6_pcb.h>
#include <netinet6/ip6_var.h>
#include <netinet6/ip6_private.h>
#include <netinet6/nd6.h>
#include <netinet6/scope6_var.h>

#ifdef MIP6
#include <netinet6/mip6.h>
#include <netinet6/mip6_var.h>
#include "mip.h"
#if NMIP > 0
#include <net/if_mip.h>
#endif /* NMIP > 0 */
#endif /* MIP6 */

#include <netinet/tcp_vtw.h>

#define ADDR_LABEL_NOTAPP (-1)
struct in6_addrpolicy defaultaddrpolicy;

int ip6_prefer_tempaddr = 0;

static int in6_selectif(struct sockaddr_in6 *, struct ip6_pktopts *,
       struct ip6_moptions *, struct route *, struct ifnet **, struct psref *);

static struct in6_addrpolicy *lookup_addrsel_policy(struct sockaddr_in6 *);

static void init_policy_queue(void);
static int add_addrsel_policyent(struct in6_addrpolicy *);
static int delete_addrsel_policyent(struct in6_addrpolicy *);
static int walk_addrsel_policy(int (*)(struct in6_addrpolicy *, void *),
                                   void *);
static int dump_addrsel_policyent(struct in6_addrpolicy *, void *);
static struct in6_addrpolicy *match_addrsel_policy(struct sockaddr_in6 *);

#define IFA6_IS_VALIDATED(ia) \
       (((ia)->ia6_flags & (IN6_IFF_TENTATIVE | IN6_IFF_DETACHED)) == 0)

/*
* Return an IPv6 address, which is the most appropriate for a given
* destination and user specified options.
* If necessary, this function lookups the routing table and returns
* an entry to the caller for later use.
*/
#if 0                           /* disabled ad-hoc */
#define REPLACE(r) do {\
       char _buf1[INET6_ADDRSTRLEN], _buf2[INET6_ADDRSTRLEN]; \
       if ((r) < sizeof(ip6stat.ip6s_sources_rule) / \
               sizeof(ip6stat.ip6s_sources_rule[0])) /* check for safety */ \
               ip6stat.ip6s_sources_rule[(r)]++; \
       printf("%s: replace %s with %s by %d\n", __func__, ia_best ? \
           IN6_PRINT(_buf1, &ia_best->ia_addr.sin6_addr) : "none", \
           IN6_PRINT(_buf2, &ia->ia_addr.sin6_addr), (r)); \
       goto replace; \
} while(/*CONSTCOND*/0)
#define NEXT(r) do {\
       if ((r) < sizeof(ip6stat.ip6s_sources_rule) / \
               sizeof(ip6stat.ip6s_sources_rule[0])) /* check for safety */ \
               ip6stat.ip6s_sources_rule[(r)]++; \
       printf("%s: keep %s against %s by %d\n", ia_best ? \
           IN6_PRINT(_buf1, &ia_best->ia_addr.sin6_addr) : "none", \
           IN6_PRINT(_buf2, &ia->ia_addr.sin6_addr), (r)); \
       goto next;              /* XXX: we can't use 'continue' here */ \
} while(/*CONSTCOND*/0)
#define BREAK(r) do { \
       if ((r) < sizeof(ip6stat.ip6s_sources_rule) / \
               sizeof(ip6stat.ip6s_sources_rule[0])) /* check for safety */ \
               ip6stat.ip6s_sources_rule[(r)]++; \
       goto out;               /* XXX: we can't use 'break' here */ \
} while(/*CONSTCOND*/0)
#else
#define REPLACE(r) goto replace
#define NEXT(r) goto next
#define BREAK(r) goto out
#endif

/*
* Called inside pserialize critical section. Don't sleep/block.
*/
static struct in6_ifaddr *
in6_select_best_ia(struct sockaddr_in6 *dstsock, struct in6_addr *dst,
   const struct ifnet *ifp, const struct ip6_pktopts *opts,
   const u_int32_t odstzone)
{
       struct in6_ifaddr *ia, *ia_best = NULL;
       int dst_scope = -1, best_scope = -1, best_matchlen = -1;
       struct in6_addrpolicy *dst_policy = NULL, *best_policy = NULL;

       IN6_ADDRLIST_READER_FOREACH(ia) {
               int new_scope = -1, new_matchlen = -1;
               struct in6_addrpolicy *new_policy = NULL;
               u_int32_t srczone, osrczone, dstzone;
               struct in6_addr src;
               struct ifnet *ifp1 = ia->ia_ifp;
               int prefer_tempaddr;

               /*
                * We'll never take an address that breaks the scope zone
                * of the destination.  We also skip an address if its zone
                * does not contain the outgoing interface.
                * XXX: we should probably use sin6_scope_id here.
                */
               if (in6_setscope(dst, ifp1, &dstzone) ||
                   odstzone != dstzone) {
                       continue;
               }
               src = ia->ia_addr.sin6_addr;

               /* Skip the scope test in impossible cases */
               if (!(ifp->if_flags & IFF_LOOPBACK) &&
                   IN6_IS_ADDR_LOOPBACK(&src))
                       continue;

               if (in6_setscope(&src, ifp, &osrczone) ||
                   in6_setscope(&src, ifp1, &srczone) ||
                   osrczone != srczone) {
                       continue;
               }

               /* avoid unusable addresses */
               if ((ia->ia6_flags & (IN6_IFF_DUPLICATED | IN6_IFF_ANYCAST)))
                       continue;
               if (!ip6_use_deprecated && IFA6_IS_DEPRECATED(ia))
                       continue;

#if defined(MIP6) && NMIP > 0
               /* avoid unusable home addresses. */
               if ((ia->ia6_flags & IN6_IFF_HOME) &&
                   !mip6_ifa6_is_addr_valid_hoa(ia))
                       continue;
#endif /* MIP6 && NMIP > 0 */

               /* Rule 1: Prefer same address */
               if (IN6_ARE_ADDR_EQUAL(dst, &ia->ia_addr.sin6_addr)) {
                       ia_best = ia;
                       BREAK(1); /* there should be no better candidate */
               }

               if (ia_best == NULL)
                       REPLACE(1);

               /* Rule 2: Prefer appropriate scope */
               if (dst_scope < 0)
                       dst_scope = in6_addrscope(dst);
               new_scope = in6_addrscope(&ia->ia_addr.sin6_addr);
               if (IN6_ARE_SCOPE_CMP(best_scope, new_scope) < 0) {
                       if (IN6_ARE_SCOPE_CMP(best_scope, dst_scope) < 0)
                               REPLACE(2);
                       NEXT(2);
               } else if (IN6_ARE_SCOPE_CMP(new_scope, best_scope) < 0) {
                       if (IN6_ARE_SCOPE_CMP(new_scope, dst_scope) < 0)
                               NEXT(2);
                       REPLACE(2);
               }

               /*
                * Rule 3: Avoid deprecated addresses.  Note that the case of
                * !ip6_use_deprecated is already rejected above.
                * Treat unvalidated addresses as deprecated here.
                */
               if (IFA6_IS_VALIDATED(ia_best) && !IFA6_IS_VALIDATED(ia))
                       NEXT(3);
               if (!IFA6_IS_VALIDATED(ia_best) && IFA6_IS_VALIDATED(ia))
                       REPLACE(3);
               if (!IFA6_IS_DEPRECATED(ia_best) && IFA6_IS_DEPRECATED(ia))
                       NEXT(3);
               if (IFA6_IS_DEPRECATED(ia_best) && !IFA6_IS_DEPRECATED(ia))
                       REPLACE(3);

               /* Rule 4: Prefer home addresses */
#if defined(MIP6) && NMIP > 0
               if (!MIP6_IS_MN)
                       goto skip_rule4;

               if ((ia_best->ia6_flags & IN6_IFF_HOME) == 0 &&
                   (ia->ia6_flags & IN6_IFF_HOME) == 0) {
                       /* both address are not home addresses. */
                       goto skip_rule4;
               }

               /*
                * If SA is simultaneously a home address and care-of
                * address and SB is not, then prefer SA. Similarly,
                * if SB is simultaneously a home address and care-of
                * address and SA is not, then prefer SB.
                */
               if (((ia_best->ia6_flags & IN6_IFF_HOME) != 0 &&
                       ia_best->ia_ifp->if_type != IFT_MIP)
                   &&
                   ((ia->ia6_flags & IN6_IFF_HOME) != 0 &&
                       ia->ia_ifp->if_type == IFT_MIP))
                       NEXT(4);
               if (((ia_best->ia6_flags & IN6_IFF_HOME) != 0 &&
                       ia_best->ia_ifp->if_type == IFT_MIP)
                   &&
                   ((ia->ia6_flags & IN6_IFF_HOME) != 0 &&
                       ia->ia_ifp->if_type != IFT_MIP))
                       REPLACE(4);
               if (ip6po_usecoa == 0) {
                       /*
                        * If SA is just a home address and SB is just
                        * a care-of address, then prefer
                        * SA. Similarly, if SB is just a home address
                        * and SA is just a care-of address, then
                        * prefer SB.
                        */
                       if ((ia_best->ia6_flags & IN6_IFF_HOME) != 0 &&
                           (ia->ia6_flags & IN6_IFF_HOME) == 0) {
                               NEXT(4);
                       }
                       if ((ia_best->ia6_flags & IN6_IFF_HOME) == 0 &&
                           (ia->ia6_flags & IN6_IFF_HOME) != 0) {
                               REPLACE(4);
                       }
               } else {
                       /*
                        * a sender don't want to use a home address
                        * because:
                        *
                        * 1) we cannot use.  (ex. NS or NA to global
                        * addresses.)
                        *
                        * 2) a user specified not to use.
                        * (ex. mip6control -u)
                        */
                       if ((ia_best->ia6_flags & IN6_IFF_HOME) == 0 &&
                           (ia->ia6_flags & IN6_IFF_HOME) != 0) {
                               /* XXX breaks stat */
                               NEXT(0);
                       }
                       if ((ia_best->ia6_flags & IN6_IFF_HOME) != 0 &&
                           (ia->ia6_flags & IN6_IFF_HOME) == 0) {
                               /* XXX breaks stat */
                               REPLACE(0);
                       }
               }
       skip_rule4:
#endif /* MIP6 && NMIP > 0 */

               /* Rule 5: Prefer outgoing interface */
               if (ia_best->ia_ifp == ifp && ia->ia_ifp != ifp)
                       NEXT(5);
               if (ia_best->ia_ifp != ifp && ia->ia_ifp == ifp)
                       REPLACE(5);

               /*
                * Rule 6: Prefer matching label
                * Note that best_policy should be non-NULL here.
                */
               if (dst_policy == NULL)
                       dst_policy = lookup_addrsel_policy(dstsock);
               if (dst_policy->label != ADDR_LABEL_NOTAPP) {
                       new_policy = lookup_addrsel_policy(&ia->ia_addr);
                       if (dst_policy->label == best_policy->label &&
                           dst_policy->label != new_policy->label)
                               NEXT(6);
                       if (dst_policy->label != best_policy->label &&
                           dst_policy->label == new_policy->label)
                               REPLACE(6);
               }

               /*
                * Rule 7: Prefer public addresses.
                * We allow users to reverse the logic by configuring
                * a sysctl variable, so that privacy conscious users can
                * always prefer temporary addresses.
                */
               if (opts == NULL ||
                   opts->ip6po_prefer_tempaddr == IP6PO_TEMPADDR_SYSTEM) {
                       prefer_tempaddr = ip6_prefer_tempaddr;
               } else if (opts->ip6po_prefer_tempaddr ==
                   IP6PO_TEMPADDR_NOTPREFER) {
                       prefer_tempaddr = 0;
               } else
                       prefer_tempaddr = 1;
               if (!(ia_best->ia6_flags & IN6_IFF_TEMPORARY) &&
                   (ia->ia6_flags & IN6_IFF_TEMPORARY)) {
                       if (prefer_tempaddr)
                               REPLACE(7);
                       else
                               NEXT(7);
               }
               if ((ia_best->ia6_flags & IN6_IFF_TEMPORARY) &&
                   !(ia->ia6_flags & IN6_IFF_TEMPORARY)) {
                       if (prefer_tempaddr)
                               NEXT(7);
                       else
                               REPLACE(7);
               }

               /*
                * Rule 8: prefer addresses on alive interfaces.
                * This is a KAME specific rule.
                */
               if ((ia_best->ia_ifp->if_flags & IFF_UP) &&
                   !(ia->ia_ifp->if_flags & IFF_UP))
                       NEXT(8);
               if (!(ia_best->ia_ifp->if_flags & IFF_UP) &&
                   (ia->ia_ifp->if_flags & IFF_UP))
                       REPLACE(8);

               /*
                * Rule 9: prefer addresses on "preferred" interfaces.
                * This is a KAME specific rule.
                */
#ifdef notyet                   /* until introducing address selection */
#define NDI_BEST ND_IFINFO(ia_best->ia_ifp)
#define NDI_NEW  ND_IFINFO(ia->ia_ifp)
               if ((NDI_BEST->flags & ND6_IFF_PREFER_SOURCE) &&
                   !(NDI_NEW->flags & ND6_IFF_PREFER_SOURCE))
                       NEXT(9);
               if (!(NDI_BEST->flags & ND6_IFF_PREFER_SOURCE) &&
                   (NDI_NEW->flags & ND6_IFF_PREFER_SOURCE))
                       REPLACE(9);
#undef NDI_BEST
#undef NDI_NEW
#endif

               /*
                * Rule 14: Use longest matching prefix.
                * Note: in the address selection draft, this rule is
                * documented as "Rule 8".  However, since it is also
                * documented that this rule can be overridden, we assign
                * a large number so that it is easy to assign smaller numbers
                * to more preferred rules.
                */
               new_matchlen = in6_matchlen(&ia->ia_addr.sin6_addr, dst);
               if (best_matchlen < new_matchlen)
                       REPLACE(14);
               if (new_matchlen < best_matchlen)
                       NEXT(14);

               /* Rule 15 is reserved. */

               /*
                * Last resort: just keep the current candidate.
                * Or, do we need more rules?
                */
               continue;

         replace:
               ia_best = ia;
               best_scope = (new_scope >= 0 ? new_scope :
                             in6_addrscope(&ia_best->ia_addr.sin6_addr));
               best_policy = (new_policy ? new_policy :
                              lookup_addrsel_policy(&ia_best->ia_addr));
               best_matchlen = (new_matchlen >= 0 ? new_matchlen :
                                in6_matchlen(&ia_best->ia_addr.sin6_addr,
                                             dst));

         next:
               continue;

         out:
               break;
       }

       return ia_best;
}
#undef REPLACE
#undef BREAK
#undef NEXT

int
in6_selectsrc(struct sockaddr_in6 *dstsock, struct ip6_pktopts *opts,
       struct ip6_moptions *mopts, struct route *ro, struct in6_addr *laddr,
       struct ifnet **ifpp, struct psref *psref, struct in6_addr *ret_ia6)
{
       struct in6_addr dst;
       struct ifnet *ifp = NULL;
       struct in6_ifaddr *ia = NULL;
       struct in6_pktinfo *pi = NULL;
       u_int32_t odstzone;
       int error = 0, iferror;
#if defined(MIP6) && NMIP > 0
       u_int8_t ip6po_usecoa = 0;
#endif /* MIP6 && NMIP > 0 */
       struct psref local_psref;
       int bound = curlwp_bind();
#define PSREF (psref == NULL) ? &local_psref : psref
       int s;

       KASSERT((ifpp != NULL && psref != NULL) ||
               (ifpp == NULL && psref == NULL));

       dst = dstsock->sin6_addr; /* make a copy for local operation */
       if (ifpp)
               *ifpp = NULL;

       /*
        * Try to determine the outgoing interface for the given destination.
        * We do this regardless of whether the socket is bound, since the
        * caller may need this information as a side effect of the call
        * to this function (e.g., for identifying the appropriate scope zone
        * ID).
        */
       iferror = in6_selectif(dstsock, opts, mopts, ro, &ifp, PSREF);
       if (ifpp != NULL)
               *ifpp = ifp;

       /*
        * If the source address is explicitly specified by the caller,
        * check if the requested source address is indeed a unicast address
        * assigned to the node, and can be used as the packet's source
        * address.  If everything is okay, use the address as source.
        */
       if (opts && (pi = opts->ip6po_pktinfo) &&
           !IN6_IS_ADDR_UNSPECIFIED(&pi->ipi6_addr)) {
               struct sockaddr_in6 srcsock;
               struct in6_ifaddr *ia6;
               int _s;
               struct ifaddr *ifa;

               /*
                * Determine the appropriate zone id of the source based on
                * the zone of the destination and the outgoing interface.
                * If the specified address is ambiguous wrt the scope zone,
                * the interface must be specified; otherwise, ifa_ifwithaddr()
                * will fail matching the address.
                */
               memset(&srcsock, 0, sizeof(srcsock));
               srcsock.sin6_family = AF_INET6;
               srcsock.sin6_len = sizeof(srcsock);
               srcsock.sin6_addr = pi->ipi6_addr;
               if (ifp) {
                       error = in6_setscope(&srcsock.sin6_addr, ifp, NULL);
                       if (error != 0)
                               goto exit;
               }

               _s = pserialize_read_enter();
               ifa = ifa_ifwithaddr(sin6tosa(&srcsock));
               if ((ia6 = ifatoia6(ifa)) == NULL ||
                   ia6->ia6_flags &
                   (IN6_IFF_ANYCAST | IN6_IFF_NOTREADY)) {
                       pserialize_read_exit(_s);
                       error = EADDRNOTAVAIL;
                       goto exit;
               }
               pi->ipi6_addr = srcsock.sin6_addr; /* XXX: this overrides pi */
               if (ifpp)
                       *ifpp = ifp;
               *ret_ia6 = ia6->ia_addr.sin6_addr;
               pserialize_read_exit(_s);
               goto exit;
       }

       /*
        * If the socket has already bound the source, just use it.  We don't
        * care at the moment whether in6_selectif() succeeded above, even
        * though it would eventually cause an error.
        */
       if (laddr && !IN6_IS_ADDR_UNSPECIFIED(laddr)) {
               *ret_ia6 = *laddr;
               goto exit;
       }

       /*
        * The outgoing interface is crucial in the general selection procedure
        * below.  If it is not known at this point, we fail.
        */
       if (ifp == NULL) {
               error = iferror;
               goto exit;
       }

       /*
        * If the address is not yet determined, choose the best one based on
        * the outgoing interface and the destination address.
        */

#if defined(MIP6) && NMIP > 0
       /*
        * a caller can specify IP6PO_USECOA to not to use a home
        * address.  for example, the case that the neighbour
        * unreachability detection to the global address.
        */
       if (opts != NULL &&
           (opts->ip6po_flags & IP6PO_USECOA) != 0) {
               ip6po_usecoa = 1;
       }
#endif /* MIP6 && NMIP > 0 */

       error = in6_setscope(&dst, ifp, &odstzone);
       if (error != 0)
               goto exit;

       s = pserialize_read_enter();

       ia = in6_select_best_ia(dstsock, &dst, ifp, opts, odstzone);
       if (ia == NULL) {
               pserialize_read_exit(s);
               error = EADDRNOTAVAIL;
               goto exit;
       }
       *ret_ia6 = ia->ia_addr.sin6_addr;

       pserialize_read_exit(s);
exit:
       if (ifpp == NULL)
               if_put(ifp, PSREF);
       curlwp_bindx(bound);
       return error;
#undef PSREF
}

int
in6_selectroute(struct sockaddr_in6 *dstsock, struct ip6_pktopts *opts,
   struct route **ro, struct rtentry **retrt, bool count_discard)
{
       int error = 0;
       struct rtentry *rt = NULL;
       union {
               struct sockaddr         dst;
               struct sockaddr_in      dst4;
               struct sockaddr_in6     dst6;
       } u;

       KASSERT(ro != NULL);
       KASSERT(*ro != NULL);
       KASSERT(retrt != NULL);

#if 0
       if (dstsock->sin6_addr.s6_addr32[0] == 0 &&
           dstsock->sin6_addr.s6_addr32[1] == 0 &&
           !IN6_IS_ADDR_LOOPBACK(&dstsock->sin6_addr)) {
               char ip6buf[INET6_ADDRSTRLEN];
               printf("%s: strange destination %s\n", __func__,
                      IN6_PRINT(ip6buf, &dstsock->sin6_addr));
       } else {
               char ip6buf[INET6_ADDRSTRLEN];
               printf("%s: destination = %s%%%d\n", __func__,
                      IN6_PRINT(ip6buf, &dstsock->sin6_addr),
                      dstsock->sin6_scope_id); /* for debug */
       }
#endif

       /*
        * If the next hop address for the packet is specified by the caller,
        * use it as the gateway.
        */
       if (opts && opts->ip6po_nexthop) {
               struct route *ron;
               struct sockaddr_in6 *sin6_next;

               sin6_next = satosin6(opts->ip6po_nexthop);

               /* at this moment, we only support AF_INET6 next hops */
               if (sin6_next->sin6_family != AF_INET6) {
                       IP6_STATINC(IP6_STAT_ODROPPED);
                       error = EAFNOSUPPORT; /* or should we proceed? */
                       goto done;
               }

               /*
                * If the next hop is an IPv6 address, then the node identified
                * by that address must be a neighbor of the sending host.
                */
               ron = &opts->ip6po_nextroute;
               rt = rtcache_lookup(ron, sin6tosa(sin6_next));
               if (rt == NULL || (rt->rt_flags & RTF_GATEWAY) != 0 ||
                   !nd6_is_addr_neighbor(sin6_next, rt->rt_ifp)) {
                       if (rt != NULL) {
                               if (count_discard)
                                       in6_ifstat_inc(rt->rt_ifp,
                                           ifs6_out_discard);
                               rtcache_unref(rt, ron);
                               rt = NULL;
                       }
                       rtcache_free(ron);
                       error = EHOSTUNREACH;
                       goto done;
               }
               *ro = ron;

               goto done;
       }

       /*
        * Use a cached route if it exists and is valid, else try to allocate
        * a new one.  Note that we should check the address family of the
        * cached destination, in case of sharing the cache with IPv4.
        *
        * for V4 mapped addresses we want to pick up the v4 route
        * see PR kern/56348
        */
       if (IN6_IS_ADDR_V4MAPPED(&dstsock->sin6_addr)) {
               in6_sin6_2_sin(&u.dst4, dstsock);
       } else {
               u.dst6 = *dstsock;
               u.dst6.sin6_scope_id = 0;
       }

       rt = rtcache_lookup1(*ro, &u.dst, 1);

       if (rt == NULL)
               error = EHOSTUNREACH;

       /*
        * Check if the outgoing interface conflicts with
        * the interface specified by ipi6_ifindex (if specified).
        * Note that loopback interface is always okay.
        * (this may happen when we are sending a packet to one of
        *  our own addresses.)
        */
       if (opts && opts->ip6po_pktinfo && opts->ip6po_pktinfo->ipi6_ifindex) {
               if (rt != NULL && !(rt->rt_ifp->if_flags & IFF_LOOPBACK) &&
                   rt->rt_ifp->if_index != opts->ip6po_pktinfo->ipi6_ifindex) {
                       if (count_discard)
                               in6_ifstat_inc(rt->rt_ifp, ifs6_out_discard);
                       error = EHOSTUNREACH;
                       rtcache_unref(rt, *ro);
                       rt = NULL;
               }
       }

done:
       if (error == EHOSTUNREACH)
               IP6_STATINC(IP6_STAT_NOROUTE);
       *retrt = rt;
       return error;
}

static int
in6_selectif(struct sockaddr_in6 *dstsock, struct ip6_pktopts *opts,
       struct ip6_moptions *mopts, struct route *ro, struct ifnet **retifp,
       struct psref *psref)
{
       int error = 0;
       struct rtentry *rt = NULL;
       struct in6_addr *dst;
       struct in6_pktinfo *pi = NULL;

       KASSERT(retifp != NULL);
       *retifp = NULL;
       dst = &dstsock->sin6_addr;

       /* If the caller specify the outgoing interface explicitly, use it. */
       if (opts && (pi = opts->ip6po_pktinfo) != NULL && pi->ipi6_ifindex) {
               /* XXX boundary check is assumed to be already done. */
               *retifp = if_get_byindex(pi->ipi6_ifindex, psref);
               if (*retifp != NULL)
                       return 0;
               goto getroute;
       }

       /*
        * If the destination address is a multicast address and the outgoing
        * interface for the address is specified by the caller, use it.
        */
       if (IN6_IS_ADDR_MULTICAST(dst) && mopts != NULL) {
               *retifp = if_get_byindex(mopts->im6o_multicast_if_index, psref);
               if (*retifp != NULL)
                       return 0; /* we do not need a route for multicast. */
       }

getroute:
       error = in6_selectroute(dstsock, opts, &ro, &rt, false);
       if (error != 0)
               return error;

       *retifp = if_get_byindex(rt->rt_ifp->if_index, psref);

       /*
        * do not use a rejected or black hole route.
        * XXX: this check should be done in the L2 output routine.
        * However, if we skipped this check here, we'd see the following
        * scenario:
        * - install a rejected route for a scoped address prefix
        *   (like fe80::/10)
        * - send a packet to a destination that matches the scoped prefix,
        *   with ambiguity about the scope zone.
        * - pick the outgoing interface from the route, and disambiguate the
        *   scope zone with the interface.
        * - ip6_output() would try to get another route with the "new"
        *   destination, which may be valid.
        * - we'd see no error on output.
        * Although this may not be very harmful, it should still be confusing.
        * We thus reject the case here.
        */
       if ((rt->rt_flags & (RTF_REJECT | RTF_BLACKHOLE))) {
               error = (rt->rt_flags & RTF_HOST ? EHOSTUNREACH : ENETUNREACH);
               /* XXX: ifp can be returned with psref even if error */
               goto out;
       }

       /*
        * Adjust the "outgoing" interface.  If we're going to loop the packet
        * back to ourselves, the ifp would be the loopback interface.
        * However, we'd rather know the interface associated to the
        * destination address (which should probably be one of our own
        * addresses.)
        */
       if (rt->rt_ifa->ifa_ifp != *retifp &&
           !if_is_deactivated(rt->rt_ifa->ifa_ifp)) {
               if_put(*retifp, psref);
               *retifp = rt->rt_ifa->ifa_ifp;
               if_acquire(*retifp, psref);
       }
out:
       rtcache_unref(rt, ro);
       return error;
}

/*
* Default hop limit selection. The precedence is as follows:
* 1. Hoplimit value specified via ioctl.
* 2. (If the outgoing interface is detected) the current
*     hop limit of the interface specified by router advertisement.
* 3. The system default hoplimit.
*/
int
in6pcb_selecthlim(struct inpcb *inp, struct ifnet *ifp)
{
       if (inp && in6p_hops6(inp) >= 0)
               return in6p_hops6(inp);
       else if (ifp)
               return (ND_IFINFO(ifp)->chlim);
       else
               return (ip6_defhlim);
}

int
in6pcb_selecthlim_rt(struct inpcb *inp)
{
       struct rtentry *rt;

       if (inp == NULL)
               return in6pcb_selecthlim(inp, NULL);

       rt = rtcache_validate(&inp->inp_route);
       if (rt != NULL) {
               int ret = in6pcb_selecthlim(inp, rt->rt_ifp);
               rtcache_unref(rt, &inp->inp_route);
               return ret;
       } else
               return in6pcb_selecthlim(inp, NULL);
}

/*
* Find an empty port and set it to the specified PCB.
*/
int
in6pcb_set_port(struct sockaddr_in6 *sin6, struct inpcb *inp, struct lwp *l)
{
       struct socket *so = inp->inp_socket;
       struct inpcbtable *table = inp->inp_table;
       u_int16_t lport, *lastport;
       enum kauth_network_req req;
       int error = 0;

       if (inp->inp_flags & IN6P_LOWPORT) {
#ifndef IPNOPRIVPORTS
               req = KAUTH_REQ_NETWORK_BIND_PRIVPORT;
#else
               req = KAUTH_REQ_NETWORK_BIND_PORT;
#endif
               lastport = &table->inpt_lastlow;
       } else {
               req = KAUTH_REQ_NETWORK_BIND_PORT;

               lastport = &table->inpt_lastport;
       }

       /* XXX-kauth: KAUTH_REQ_NETWORK_BIND_AUTOASSIGN_{,PRIV}PORT */
       error = kauth_authorize_network(l->l_cred, KAUTH_NETWORK_BIND, req, so,
           sin6, NULL);
       if (error)
               return (EACCES);

      /*
       * Use RFC6056 randomized port selection
       */
       error = portalgo_randport(&lport, inp, l->l_cred);
       if (error)
               return error;

       inp->inp_flags |= IN6P_ANONPORT;
       *lastport = lport;
       inp->inp_lport = htons(lport);
       in6pcb_set_state(inp, INP_BOUND);
       return (0);             /* success */
}

void
addrsel_policy_init(void)
{
       init_policy_queue();

       /* initialize the "last resort" policy */
       memset(&defaultaddrpolicy, 0, sizeof(defaultaddrpolicy));
       defaultaddrpolicy.label = ADDR_LABEL_NOTAPP;
}

/*
* XXX: NOMPSAFE if a policy is set
*/
static struct in6_addrpolicy *
lookup_addrsel_policy(struct sockaddr_in6 *key)
{
       struct in6_addrpolicy *match = NULL;

       match = match_addrsel_policy(key);

       if (match == NULL)
               match = &defaultaddrpolicy;
       else
               match->use++;

       return (match);
}

/*
* Subroutines to manage the address selection policy table via sysctl.
*/
struct sel_walkarg {
       size_t  w_total;
       size_t  w_given;
       void *  w_where;
       void *w_limit;
};

int sysctl_net_inet6_addrctlpolicy(SYSCTLFN_ARGS);
int
sysctl_net_inet6_addrctlpolicy(SYSCTLFN_ARGS)
{
       int error = 0;
       int s;

       s = splsoftnet();

       if (newp) {
               error = EPERM;
               goto end;
       }
       if (oldp && oldlenp == NULL) {
               error = EINVAL;
               goto end;
       }
       if (oldp || oldlenp) {
               struct sel_walkarg w;
               size_t oldlen = *oldlenp;

               memset(&w, 0, sizeof(w));
               w.w_given = oldlen;
               w.w_where = oldp;
               if (oldp)
                       w.w_limit = (char *)oldp + oldlen;

               error = walk_addrsel_policy(dump_addrsel_policyent, &w);

               *oldlenp = w.w_total;
               if (oldp && w.w_total > oldlen && error == 0)
                       error = ENOMEM;
       }

 end:
       splx(s);

       return (error);
}

int
in6_src_ioctl(u_long cmd, void *data)
{
       int i;
       struct in6_addrpolicy ent0;

       if (cmd != SIOCAADDRCTL_POLICY && cmd != SIOCDADDRCTL_POLICY)
               return (EOPNOTSUPP); /* check for safety */

       ent0 = *(struct in6_addrpolicy *)data;

       if (ent0.label == ADDR_LABEL_NOTAPP)
               return (EINVAL);
       /* check if the prefix mask is consecutive. */
       if (in6_mask2len(&ent0.addrmask.sin6_addr, NULL) < 0)
               return (EINVAL);
       /* clear trailing garbages (if any) of the prefix address. */
       for (i = 0; i < 4; i++) {
               ent0.addr.sin6_addr.s6_addr32[i] &=
                       ent0.addrmask.sin6_addr.s6_addr32[i];
       }
       ent0.use = 0;

       switch (cmd) {
       case SIOCAADDRCTL_POLICY:
               return (add_addrsel_policyent(&ent0));
       case SIOCDADDRCTL_POLICY:
               return (delete_addrsel_policyent(&ent0));
       }

       return (0);             /* XXX: compromise compilers */
}

/*
* The followings are implementation of the policy table using a
* simple tail queue.
* XXX such details should be hidden.
* XXX implementation using binary tree should be more efficient.
*/
struct addrsel_policyent {
       TAILQ_ENTRY(addrsel_policyent) ape_entry;
       struct in6_addrpolicy ape_policy;
};

TAILQ_HEAD(addrsel_policyhead, addrsel_policyent);

struct addrsel_policyhead addrsel_policytab;

static void
init_policy_queue(void)
{
       TAILQ_INIT(&addrsel_policytab);
}

static int
add_addrsel_policyent(struct in6_addrpolicy *newpolicy)
{
       struct addrsel_policyent *newpol, *pol;

       /* duplication check */
       TAILQ_FOREACH(pol, &addrsel_policytab, ape_entry) {
               if (IN6_ARE_ADDR_EQUAL(&newpolicy->addr.sin6_addr,
                   &pol->ape_policy.addr.sin6_addr) &&
                   IN6_ARE_ADDR_EQUAL(&newpolicy->addrmask.sin6_addr,
                   &pol->ape_policy.addrmask.sin6_addr)) {
                       return (EEXIST);        /* or override it? */
               }
       }

       newpol = malloc(sizeof(*newpol), M_IFADDR, M_WAITOK|M_ZERO);

       /* XXX: should validate entry */
       newpol->ape_policy = *newpolicy;

       TAILQ_INSERT_TAIL(&addrsel_policytab, newpol, ape_entry);

       return (0);
}

static int
delete_addrsel_policyent(struct in6_addrpolicy *key)
{
       struct addrsel_policyent *pol;

       /* search for the entry in the table */
       for (pol = TAILQ_FIRST(&addrsel_policytab); pol;
            pol = TAILQ_NEXT(pol, ape_entry)) {
               if (IN6_ARE_ADDR_EQUAL(&key->addr.sin6_addr,
                   &pol->ape_policy.addr.sin6_addr) &&
                   IN6_ARE_ADDR_EQUAL(&key->addrmask.sin6_addr,
                   &pol->ape_policy.addrmask.sin6_addr)) {
                       break;
               }
       }
       if (pol == NULL) {
               return (ESRCH);
       }

       TAILQ_REMOVE(&addrsel_policytab, pol, ape_entry);

       return (0);
}

static int
walk_addrsel_policy(int (*callback)(struct in6_addrpolicy *, void *), void *w)
{
       struct addrsel_policyent *pol;
       int error = 0;

       TAILQ_FOREACH(pol, &addrsel_policytab, ape_entry) {
               if ((error = (*callback)(&pol->ape_policy, w)) != 0)
                       return error;
       }

       return error;
}

static int
dump_addrsel_policyent(struct in6_addrpolicy *pol, void *arg)
{
       int error = 0;
       struct sel_walkarg *w = arg;

       if (w->w_where && (char *)w->w_where + sizeof(*pol) <= (char *)w->w_limit) {
               if ((error = copyout(pol, w->w_where, sizeof(*pol))) != 0)
                       return error;
               w->w_where = (char *)w->w_where + sizeof(*pol);
       }
       w->w_total += sizeof(*pol);

       return error;
}

static struct in6_addrpolicy *
match_addrsel_policy(struct sockaddr_in6 *key)
{
       struct addrsel_policyent *pent;
       struct in6_addrpolicy *bestpol = NULL, *pol;
       int matchlen, bestmatchlen = -1;
       u_char *mp, *ep, *k, *p, m;

       for (pent = TAILQ_FIRST(&addrsel_policytab); pent;
            pent = TAILQ_NEXT(pent, ape_entry)) {
               matchlen = 0;

               pol = &pent->ape_policy;
               mp = (u_char *)&pol->addrmask.sin6_addr;
               ep = mp + 16;   /* XXX: scope field? */
               k = (u_char *)&key->sin6_addr;
               p = (u_char *)&pol->addr.sin6_addr;
               for (; mp < ep && *mp; mp++, k++, p++) {
                       m = *mp;
                       if ((*k & m) != *p)
                               goto next; /* not match */
                       if (m == 0xff) /* short cut for a typical case */
                               matchlen += 8;
                       else {
                               while (m >= 0x80) {
                                       matchlen++;
                                       m <<= 1;
                               }
                       }
               }

               /* matched.  check if this is better than the current best. */
               if (bestpol == NULL ||
                   matchlen > bestmatchlen) {
                       bestpol = pol;
                       bestmatchlen = matchlen;
               }

         next:
               continue;
       }

       return (bestpol);
}