kernel - Fix excessive call stack depth on stuck interrupt

[dragonfly.git] / sys / kern / uipc_usrreq.c

/*
 * Copyright (c) 1982, 1986, 1989, 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. 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.
 *
 *      From: @(#)uipc_usrreq.c 8.3 (Berkeley) 1/4/94
 * $FreeBSD: src/sys/kern/uipc_usrreq.c,v 1.54.2.10 2003/03/04 17:28:09 nectar Exp $
 */

#include <sys/param.h>
#include <sys/systm.h>
#include <sys/kernel.h>
#include <sys/domain.h>
#include <sys/fcntl.h>
#include <sys/malloc.h>         /* XXX must be before <sys/file.h> */
#include <sys/proc.h>
#include <sys/file.h>
#include <sys/filedesc.h>
#include <sys/mbuf.h>
#include <sys/nlookup.h>
#include <sys/protosw.h>
#include <sys/socket.h>
#include <sys/socketvar.h>
#include <sys/resourcevar.h>
#include <sys/stat.h>
#include <sys/mount.h>
#include <sys/sysctl.h>
#include <sys/un.h>
#include <sys/unpcb.h>
#include <sys/vnode.h>
#include <sys/kern_syscall.h>
#include <sys/taskqueue.h>

#include <sys/file2.h>
#include <sys/spinlock2.h>
#include <sys/socketvar2.h>
#include <sys/msgport2.h>

/*
 * Unix communications domain.
 *
 * TODO:
 *      RDM
 *      rethink name space problems
 *      need a proper out-of-band
 *      lock pushdown
 *
 *
 * Unix domain sockets GC.
 *
 * It was originally designed to address following three cases:
 * 1) Receiving unix domain socket can not accept the rights, e.g.
 *    when the so_rcv is full.
 * 2) Caller of recvmsg(2) does not pass buffer to receive rights.
 * 3) Unix domain sockets loop reference, e.g. s1 is on s2.so_rcv,
 *    while s2 on s1.so_rcv.
 *
 * Code under UNP_GC_ALLFILES is intended to address all above three
 * cases.  However, 1) was addressed a long time ago in uipc_send()
 * (we inheritted the fix from FreeBSD when DragonFly forked).  2)
 * was addressed in soreceive() by git-e62cfe62.  3) is the only
 * case that needs GC.  The new code (!UNP_GC_ALLFILES) addresses
 * case 3) in the following way:
 * - Record the struct file in unpcb, if the Unix domain socket is
 *   passed as one of the rights.
 * - At GC time, only unpcbs are scanned, and only Unix domain sockets
 *   that are still used as rights are potential GC targets.
 */

#define UNP_DETACHED            UNP_PRIVATE1
#define UNP_CONNECTING          UNP_PRIVATE2
#define UNP_DROPPED             UNP_PRIVATE3
#define UNP_MARKER              UNP_PRIVATE4

#define UNPGC_REF               0x1     /* unpcb has external ref. */
#define UNPGC_DEAD              0x2     /* unpcb might be dead. */
#define UNPGC_SCANNED           0x4     /* Has been scanned. */

#define UNP_GCFILE_MAX          256

/* For unp_internalize() and unp_externalize() */
CTASSERT(sizeof(struct file *) >= sizeof(int));

#define UNP_ISATTACHED(unp)     \
    ((unp) != NULL && ((unp)->unp_flags & UNP_DETACHED) == 0)

#ifdef INVARIANTS
#define UNP_ASSERT_TOKEN_HELD(unp) \
    ASSERT_LWKT_TOKEN_HELD(lwkt_token_pool_lookup((unp)))
#else   /* !INVARIANTS */
#define UNP_ASSERT_TOKEN_HELD(unp)
#endif  /* INVARIANTS */

struct unp_defdiscard {
        SLIST_ENTRY(unp_defdiscard) next;
        struct file *fp;
};
SLIST_HEAD(unp_defdiscard_list, unp_defdiscard);

TAILQ_HEAD(unpcb_qhead, unpcb);
struct unp_global_head {
        struct unpcb_qhead      list;
        int                     count;
};

static  MALLOC_DEFINE(M_UNPCB, "unpcb", "unpcb struct");
static  unp_gen_t unp_gencnt;

static struct unp_global_head unp_stream_head;
static struct unp_global_head unp_dgram_head;
static struct unp_global_head unp_seqpkt_head;

static struct unp_global_head * const unp_heads[] =
    { &unp_stream_head, &unp_dgram_head, &unp_seqpkt_head, NULL };

static struct lwkt_token unp_token = LWKT_TOKEN_INITIALIZER(unp_token);
static struct taskqueue *unp_taskqueue;

static struct unp_defdiscard_list unp_defdiscard_head;
static struct spinlock unp_defdiscard_spin;
static struct task unp_defdiscard_task;

static struct   sockaddr sun_noname = { sizeof(sun_noname), AF_LOCAL };
static ino_t    unp_ino = 1;            /* prototype for fake inode numbers */

static int     unp_attach (struct socket *, struct pru_attach_info *);
static void    unp_detach (struct unpcb *);
static int     unp_bind (struct unpcb *,struct sockaddr *, struct thread *);
static int     unp_connect (struct socket *,struct sockaddr *,
                                struct thread *);
static void    unp_disconnect(struct unpcb *, int);
static void    unp_shutdown (struct unpcb *);
static void    unp_gc(void *, int);
#ifdef UNP_GC_ALLFILES
static int     unp_gc_clearmarks(struct file *, void *);
static int     unp_gc_checkmarks(struct file *, void *);
static int     unp_gc_checkrefs(struct file *, void *);
static void    unp_mark(struct file *, void *data);
#endif
static void    unp_scan (struct mbuf *, void (*)(struct file *, void *),
                                void *data);
static void    unp_discard (struct file *, void *);
static int     unp_internalize (struct mbuf *, struct thread *);
static int     unp_listen (struct unpcb *, struct thread *);
static void    unp_fp_externalize(struct lwp *lp, struct file *fp, int fd,
                   int flags);
static int     unp_find_lockref(struct sockaddr *nam, struct thread *td,
                   short type, struct unpcb **unp_ret);
static int     unp_connect_pair(struct unpcb *unp, struct unpcb *unp2);
static void    unp_drop(struct unpcb *unp, int error);
static void    unp_defdiscard_taskfunc(void *, int);

static int      unp_rights;                     /* file descriptors in flight */
static struct lwkt_token unp_rights_token =
    LWKT_TOKEN_INITIALIZER(unp_rights_token);
static struct task unp_gc_task;
static struct unpcb *unp_gc_marker;

SYSCTL_DECL(_net_local);
SYSCTL_INT(_net_local, OID_AUTO, inflight, CTLFLAG_RD, &unp_rights, 0,
   "File descriptors in flight");

/*
 * SMP Considerations:
 *
 *      Since unp_token will be automaticly released upon execution of
 *      blocking code, we need to reference unp_conn before any possible
 *      blocking code to prevent it from being ripped behind our back.
 *
 *      Any adjustment to unp->unp_conn requires both the global unp_token
 *      AND the per-unp token (lwkt_token_pool_lookup(unp)) to be held.
 *
 *      Any access to so_pcb to obtain unp requires the pool token for
 *      unp to be held.
 */

static __inline void
unp_reference(struct unpcb *unp)
{
        /* 0->1 transition will not work */
        KKASSERT(unp->unp_refcnt > 0);
        atomic_add_int(&unp->unp_refcnt, 1);
}

static __inline void
unp_free(struct unpcb *unp)
{
        KKASSERT(unp->unp_refcnt > 0);
        if (atomic_fetchadd_int(&unp->unp_refcnt, -1) == 1)
                unp_detach(unp);
}

static __inline struct unpcb *
unp_getsocktoken(struct socket *so)
{
        struct unpcb *unp;

        /*
         * The unp pointer is invalid until we verify that it is
         * good by re-checking so_pcb AFTER obtaining the token.
         */
        while ((unp = so->so_pcb) != NULL) {
                lwkt_getpooltoken(unp);
                if (unp == so->so_pcb)
                        break;
                lwkt_relpooltoken(unp);
        }
        return unp;
}

static __inline void
unp_reltoken(struct unpcb *unp)
{
        if (unp != NULL)
                lwkt_relpooltoken(unp);
}

static __inline void
unp_setflags(struct unpcb *unp, int flags)
{
        atomic_set_int(&unp->unp_flags, flags);
}

static __inline void
unp_clrflags(struct unpcb *unp, int flags)
{
        atomic_clear_int(&unp->unp_flags, flags);
}

static __inline struct unp_global_head *
unp_globalhead(short type)
{
        switch (type) {
        case SOCK_STREAM:
                return &unp_stream_head;
        case SOCK_DGRAM:
                return &unp_dgram_head;
        case SOCK_SEQPACKET:
                return &unp_seqpkt_head;
        default:
                panic("unknown socket type %d", type);
        }
}

static __inline struct unpcb *
unp_fp2unpcb(struct file *fp)
{
        struct socket *so;

        if (fp->f_type != DTYPE_SOCKET)
                return NULL;

        so = fp->f_data;
        if (so == NULL)
                return NULL;

        if (so->so_proto->pr_domain != &localdomain)
                return NULL;

        return so->so_pcb;
}

static __inline void
unp_add_right(struct file *fp)
{
        struct unpcb *unp;

        ASSERT_LWKT_TOKEN_HELD(&unp_rights_token);
        KASSERT(fp->f_count > 0, ("invalid f_count %d", fp->f_count));

        unp = unp_fp2unpcb(fp);
        if (unp != NULL) {
                unp->unp_fp = fp;
                unp->unp_msgcount++;
        }
        fp->f_msgcount++;
        unp_rights++;
}

static __inline void
unp_del_right(struct file *fp)
{
        struct unpcb *unp;

        ASSERT_LWKT_TOKEN_HELD(&unp_rights_token);
        KASSERT(fp->f_count > 0, ("invalid f_count %d", fp->f_count));

        unp = unp_fp2unpcb(fp);
        if (unp != NULL) {
                KASSERT(unp->unp_msgcount > 0,
                    ("invalid unp msgcount %d", unp->unp_msgcount));
                unp->unp_msgcount--;
                if (unp->unp_msgcount == 0)
                        unp->unp_fp = NULL;
        }
        fp->f_msgcount--;
        unp_rights--;
}

/*
 * NOTE: (so) is referenced from soabort*() and netmsg_pru_abort()
 *       will sofree() it when we return.
 */
static void
uipc_abort(netmsg_t msg)
{
        struct unpcb *unp;
        int error;

        lwkt_gettoken(&unp_token);
        unp = unp_getsocktoken(msg->base.nm_so);

        if (UNP_ISATTACHED(unp)) {
                unp_drop(unp, ECONNABORTED);
                error = 0;
        } else {
                error = EINVAL;
        }

        unp_reltoken(unp);
        lwkt_reltoken(&unp_token);

        lwkt_replymsg(&msg->lmsg, error);
}

static void
uipc_accept(netmsg_t msg)
{
        struct unpcb *unp;
        int error;

        lwkt_gettoken(&unp_token);
        unp = unp_getsocktoken(msg->base.nm_so);

        if (!UNP_ISATTACHED(unp)) {
                error = EINVAL;
        } else {
                struct unpcb *unp2 = unp->unp_conn;

                /*
                 * Pass back name of connected socket,
                 * if it was bound and we are still connected
                 * (our peer may have closed already!).
                 */
                if (unp2 && unp2->unp_addr) {
                        unp_reference(unp2);
                        *msg->accept.nm_nam = dup_sockaddr(
                                (struct sockaddr *)unp2->unp_addr);
                        unp_free(unp2);
                } else {
                        *msg->accept.nm_nam = dup_sockaddr(&sun_noname);
                }
                error = 0;
        }

        unp_reltoken(unp);
        lwkt_reltoken(&unp_token);

        lwkt_replymsg(&msg->lmsg, error);
}

static void
uipc_attach(netmsg_t msg)
{
        int error;

        lwkt_gettoken(&unp_token);

        KASSERT(msg->base.nm_so->so_pcb == NULL, ("double unp attach"));
        error = unp_attach(msg->base.nm_so, msg->attach.nm_ai);

        lwkt_reltoken(&unp_token);
        lwkt_replymsg(&msg->lmsg, error);
}

static void
uipc_bind(netmsg_t msg)
{
        struct unpcb *unp;
        int error;

        lwkt_gettoken(&unp_token);
        unp = unp_getsocktoken(msg->base.nm_so);

        if (UNP_ISATTACHED(unp))
                error = unp_bind(unp, msg->bind.nm_nam, msg->bind.nm_td);
        else
                error = EINVAL;

        unp_reltoken(unp);
        lwkt_reltoken(&unp_token);

        lwkt_replymsg(&msg->lmsg, error);
}

static void
uipc_connect(netmsg_t msg)
{
        int error;

        error = unp_connect(msg->base.nm_so, msg->connect.nm_nam,
            msg->connect.nm_td);
        lwkt_replymsg(&msg->lmsg, error);
}

static void
uipc_connect2(netmsg_t msg)
{
        int error;

        error = unp_connect2(msg->connect2.nm_so1, msg->connect2.nm_so2);
        lwkt_replymsg(&msg->lmsg, error);
}

/* control is EOPNOTSUPP */

static void
uipc_detach(netmsg_t msg)
{
        struct unpcb *unp;
        int error;

        lwkt_gettoken(&unp_token);
        unp = unp_getsocktoken(msg->base.nm_so);

        if (UNP_ISATTACHED(unp)) {
                unp_drop(unp, 0);
                error = 0;
        } else {
                error = EINVAL;
        }

        unp_reltoken(unp);
        lwkt_reltoken(&unp_token);

        lwkt_replymsg(&msg->lmsg, error);
}

static void
uipc_disconnect(netmsg_t msg)
{
        struct unpcb *unp;
        int error;

        lwkt_gettoken(&unp_token);
        unp = unp_getsocktoken(msg->base.nm_so);

        if (UNP_ISATTACHED(unp)) {
                unp_disconnect(unp, 0);
                error = 0;
        } else {
                error = EINVAL;
        }

        unp_reltoken(unp);
        lwkt_reltoken(&unp_token);

        lwkt_replymsg(&msg->lmsg, error);
}

static void
uipc_listen(netmsg_t msg)
{
        struct unpcb *unp;
        int error;

        lwkt_gettoken(&unp_token);
        unp = unp_getsocktoken(msg->base.nm_so);

        if (!UNP_ISATTACHED(unp) || unp->unp_vnode == NULL)
                error = EINVAL;
        else
                error = unp_listen(unp, msg->listen.nm_td);

        unp_reltoken(unp);
        lwkt_reltoken(&unp_token);

        lwkt_replymsg(&msg->lmsg, error);
}

static void
uipc_peeraddr(netmsg_t msg)
{
        struct unpcb *unp;
        int error;

        lwkt_gettoken(&unp_token);
        unp = unp_getsocktoken(msg->base.nm_so);

        if (!UNP_ISATTACHED(unp)) {
                error = EINVAL;
        } else if (unp->unp_conn && unp->unp_conn->unp_addr) {
                struct unpcb *unp2 = unp->unp_conn;

                unp_reference(unp2);
                *msg->peeraddr.nm_nam = dup_sockaddr(
                                (struct sockaddr *)unp2->unp_addr);
                unp_free(unp2);
                error = 0;
        } else {
                /*
                 * XXX: It seems that this test always fails even when
                 * connection is established.  So, this else clause is
                 * added as workaround to return PF_LOCAL sockaddr.
                 */
                *msg->peeraddr.nm_nam = dup_sockaddr(&sun_noname);
                error = 0;
        }

        unp_reltoken(unp);
        lwkt_reltoken(&unp_token);

        lwkt_replymsg(&msg->lmsg, error);
}

static void
uipc_rcvd(netmsg_t msg)
{
        struct unpcb *unp, *unp2;
        struct socket *so;
        struct socket *so2;
        int error;

        /*
         * so_pcb is only modified with both the global and the unp
         * pool token held.
         */
        so = msg->base.nm_so;
        unp = unp_getsocktoken(so);

        if (!UNP_ISATTACHED(unp)) {
                error = EINVAL;
                goto done;
        }

        switch (so->so_type) {
        case SOCK_DGRAM:
                panic("uipc_rcvd DGRAM?");
                /*NOTREACHED*/
        case SOCK_STREAM:
        case SOCK_SEQPACKET:
                if (unp->unp_conn == NULL)
                        break;
                unp2 = unp->unp_conn;   /* protected by pool token */

                /*
                 * Because we are transfering mbufs directly to the
                 * peer socket we have to use SSB_STOP on the sender
                 * to prevent it from building up infinite mbufs.
                 *
                 * As in several places in this module w ehave to ref unp2
                 * to ensure that it does not get ripped out from under us
                 * if we block on the so2 token or in sowwakeup().
                 */
                so2 = unp2->unp_socket;
                unp_reference(unp2);
                lwkt_gettoken(&so2->so_rcv.ssb_token);
                if (so->so_rcv.ssb_cc < so2->so_snd.ssb_hiwat &&
                    so->so_rcv.ssb_mbcnt < so2->so_snd.ssb_mbmax
                ) {
                        atomic_clear_int(&so2->so_snd.ssb_flags, SSB_STOP);

                        sowwakeup(so2);
                }
                lwkt_reltoken(&so2->so_rcv.ssb_token);
                unp_free(unp2);
                break;
        default:
                panic("uipc_rcvd unknown socktype");
                /*NOTREACHED*/
        }
        error = 0;
done:
        unp_reltoken(unp);
        lwkt_replymsg(&msg->lmsg, error);
}

/* pru_rcvoob is EOPNOTSUPP */

static void
uipc_send(netmsg_t msg)
{
        struct unpcb *unp, *unp2;
        struct socket *so;
        struct socket *so2;
        struct mbuf *control;
        struct mbuf *m;
        int error = 0;

        so = msg->base.nm_so;
        control = msg->send.nm_control;
        m = msg->send.nm_m;

        /*
         * so_pcb is only modified with both the global and the unp
         * pool token held.
         */
        so = msg->base.nm_so;
        unp = unp_getsocktoken(so);

        if (!UNP_ISATTACHED(unp)) {
                error = EINVAL;
                goto release;
        }

        if (msg->send.nm_flags & PRUS_OOB) {
                error = EOPNOTSUPP;
                goto release;
        }

        wakeup_start_delayed();

        if (control && (error = unp_internalize(control, msg->send.nm_td)))
                goto release;

        switch (so->so_type) {
        case SOCK_DGRAM: 
        {
                struct sockaddr *from;

                if (msg->send.nm_addr) {
                        if (unp->unp_conn) {
                                error = EISCONN;
                                break;
                        }
                        lwkt_gettoken(&unp_token);
                        error = unp_find_lockref(msg->send.nm_addr,
                            msg->send.nm_td, so->so_type, &unp2);
                        if (error) {
                                lwkt_reltoken(&unp_token);
                                break;
                        }
                        /*
                         * NOTE:
                         * unp2 is locked and referenced.
                         *
                         * We could unlock unp2 now, since it was checked
                         * and referenced.
                         */
                        unp_reltoken(unp2);
                        lwkt_reltoken(&unp_token);
                } else {
                        if (unp->unp_conn == NULL) {
                                error = ENOTCONN;
                                break;
                        }
                        unp2 = unp->unp_conn;
                        unp_reference(unp2);
                }
                /* NOTE: unp2 is referenced. */
                so2 = unp2->unp_socket;

                if (unp->unp_addr)
                        from = (struct sockaddr *)unp->unp_addr;
                else
                        from = &sun_noname;

                lwkt_gettoken(&so2->so_rcv.ssb_token);
                if (ssb_appendaddr(&so2->so_rcv, from, m, control)) {
                        sorwakeup(so2);
                        m = NULL;
                        control = NULL;
                } else {
                        error = ENOBUFS;
                }
                lwkt_reltoken(&so2->so_rcv.ssb_token);

                unp_free(unp2);
                break;
        }

        case SOCK_STREAM:
        case SOCK_SEQPACKET:
                /* Connect if not connected yet. */
                /*
                 * Note: A better implementation would complain
                 * if not equal to the peer's address.
                 */
                if (unp->unp_conn == NULL) {
                        if (msg->send.nm_addr) {
                                error = unp_connect(so,
                                                    msg->send.nm_addr,
                                                    msg->send.nm_td);
                                if (error)
                                        break;  /* XXX */
                        }
                        /*
                         * NOTE:
                         * unp_conn still could be NULL, even if the
                         * above unp_connect() succeeds; since the
                         * current unp's token could be released due
                         * to blocking operations after unp_conn is
                         * assigned.
                         */
                        if (unp->unp_conn == NULL) {
                                error = ENOTCONN;
                                break;
                        }
                }
                if (so->so_state & SS_CANTSENDMORE) {
                        error = EPIPE;
                        break;
                }

                unp2 = unp->unp_conn;
                KASSERT(unp2 != NULL, ("unp is not connected"));
                so2 = unp2->unp_socket;

                unp_reference(unp2);

                /*
                 * Send to paired receive port, and then reduce
                 * send buffer hiwater marks to maintain backpressure.
                 * Wake up readers.
                 */
                lwkt_gettoken(&so2->so_rcv.ssb_token);
                if (control) {
                        if (ssb_appendcontrol(&so2->so_rcv, m, control)) {
                                control = NULL;
                                m = NULL;
                        }
                } else if (so->so_type == SOCK_SEQPACKET) {
                        sbappendrecord(&so2->so_rcv.sb, m);
                        m = NULL;
                } else {
                        sbappend(&so2->so_rcv.sb, m);
                        m = NULL;
                }

                /*
                 * Because we are transfering mbufs directly to the
                 * peer socket we have to use SSB_STOP on the sender
                 * to prevent it from building up infinite mbufs.
                 */
                if (so2->so_rcv.ssb_cc >= so->so_snd.ssb_hiwat ||
                    so2->so_rcv.ssb_mbcnt >= so->so_snd.ssb_mbmax
                ) {
                        atomic_set_int(&so->so_snd.ssb_flags, SSB_STOP);
                }
                lwkt_reltoken(&so2->so_rcv.ssb_token);
                sorwakeup(so2);

                unp_free(unp2);
                break;

        default:
                panic("uipc_send unknown socktype");
        }

        /*
         * SEND_EOF is equivalent to a SEND followed by a SHUTDOWN.
         */
        if (msg->send.nm_flags & PRUS_EOF) {
                socantsendmore(so);
                unp_shutdown(unp);
        }

        if (control && error != 0)
                unp_dispose(control);
release:
        unp_reltoken(unp);
        wakeup_end_delayed();

        if (control)
                m_freem(control);
        if (m)
                m_freem(m);
        lwkt_replymsg(&msg->lmsg, error);
}

/*
 * MPSAFE
 */
static void
uipc_sense(netmsg_t msg)
{
        struct unpcb *unp;
        struct socket *so;
        struct stat *sb;
        int error;

        so = msg->base.nm_so;
        sb = msg->sense.nm_stat;

        /*
         * so_pcb is only modified with both the global and the unp
         * pool token held.
         */
        unp = unp_getsocktoken(so);

        if (!UNP_ISATTACHED(unp)) {
                error = EINVAL;
                goto done;
        }

        sb->st_blksize = so->so_snd.ssb_hiwat;
        sb->st_dev = NOUDEV;
        if (unp->unp_ino == 0) {        /* make up a non-zero inode number */
                unp->unp_ino = atomic_fetchadd_long(&unp_ino, 1);
                if (__predict_false(unp->unp_ino == 0))
                        unp->unp_ino = atomic_fetchadd_long(&unp_ino, 1);
        }
        sb->st_ino = unp->unp_ino;
        error = 0;
done:
        unp_reltoken(unp);
        lwkt_replymsg(&msg->lmsg, error);
}

static void
uipc_shutdown(netmsg_t msg)
{
        struct socket *so;
        struct unpcb *unp;
        int error;

        /*
         * so_pcb is only modified with both the global and the unp
         * pool token held.
         */
        so = msg->base.nm_so;
        unp = unp_getsocktoken(so);

        if (UNP_ISATTACHED(unp)) {
                socantsendmore(so);
                unp_shutdown(unp);
                error = 0;
        } else {
                error = EINVAL;
        }

        unp_reltoken(unp);
        lwkt_replymsg(&msg->lmsg, error);
}

static void
uipc_sockaddr(netmsg_t msg)
{
        struct unpcb *unp;
        int error;

        /*
         * so_pcb is only modified with both the global and the unp
         * pool token held.
         */
        unp = unp_getsocktoken(msg->base.nm_so);

        if (UNP_ISATTACHED(unp)) {
                if (unp->unp_addr) {
                        *msg->sockaddr.nm_nam =
                                dup_sockaddr((struct sockaddr *)unp->unp_addr);
                }
                error = 0;
        } else {
                error = EINVAL;
        }

        unp_reltoken(unp);
        lwkt_replymsg(&msg->lmsg, error);
}

struct pr_usrreqs uipc_usrreqs = {
        .pru_abort = uipc_abort,
        .pru_accept = uipc_accept,
        .pru_attach = uipc_attach,
        .pru_bind = uipc_bind,
        .pru_connect = uipc_connect,
        .pru_connect2 = uipc_connect2,
        .pru_control = pr_generic_notsupp,
        .pru_detach = uipc_detach,
        .pru_disconnect = uipc_disconnect,
        .pru_listen = uipc_listen,
        .pru_peeraddr = uipc_peeraddr,
        .pru_rcvd = uipc_rcvd,
        .pru_rcvoob = pr_generic_notsupp,
        .pru_send = uipc_send,
        .pru_sense = uipc_sense,
        .pru_shutdown = uipc_shutdown,
        .pru_sockaddr = uipc_sockaddr,
        .pru_sosend = sosend,
        .pru_soreceive = soreceive
};

void
uipc_ctloutput(netmsg_t msg)
{
        struct socket *so;
        struct sockopt *sopt;
        struct unpcb *unp;
        int error = 0;

        so = msg->base.nm_so;
        sopt = msg->ctloutput.nm_sopt;

        lwkt_gettoken(&unp_token);
        unp = unp_getsocktoken(so);

        if (!UNP_ISATTACHED(unp)) {
                error = EINVAL;
                goto done;
        }

        switch (sopt->sopt_dir) {
        case SOPT_GET:
                switch (sopt->sopt_name) {
                case LOCAL_PEERCRED:
                        if (unp->unp_flags & UNP_HAVEPC)
                                soopt_from_kbuf(sopt, &unp->unp_peercred,
                                                sizeof(unp->unp_peercred));
                        else {
                                if (so->so_type == SOCK_STREAM)
                                        error = ENOTCONN;
                                else if (so->so_type == SOCK_SEQPACKET)
                                        error = ENOTCONN;
                                else
                                        error = EINVAL;
                        }
                        break;
                default:
                        error = EOPNOTSUPP;
                        break;
                }
                break;
        case SOPT_SET:
        default:
                error = EOPNOTSUPP;
                break;
        }

done:
        unp_reltoken(unp);
        lwkt_reltoken(&unp_token);

        lwkt_replymsg(&msg->lmsg, error);
}
        
/*
 * Both send and receive buffers are allocated PIPSIZ bytes of buffering
 * for stream sockets, although the total for sender and receiver is
 * actually only PIPSIZ.
 *
 * Datagram sockets really use the sendspace as the maximum datagram size,
 * and don't really want to reserve the sendspace.  Their recvspace should
 * be large enough for at least one max-size datagram plus address.
 *
 * We want the local send/recv space to be significant larger then lo0's
 * mtu of 16384.
 */
#ifndef PIPSIZ
#define PIPSIZ  57344
#endif
static u_long   unpst_sendspace = PIPSIZ;
static u_long   unpst_recvspace = PIPSIZ;
static u_long   unpdg_sendspace = 2*1024;       /* really max datagram size */
static u_long   unpdg_recvspace = 4*1024;

SYSCTL_DECL(_net_local_seqpacket);
SYSCTL_DECL(_net_local_stream);
SYSCTL_INT(_net_local_stream, OID_AUTO, sendspace, CTLFLAG_RW, 
    &unpst_sendspace, 0, "Size of stream socket send buffer");
SYSCTL_INT(_net_local_stream, OID_AUTO, recvspace, CTLFLAG_RW,
    &unpst_recvspace, 0, "Size of stream socket receive buffer");

SYSCTL_DECL(_net_local_dgram);
SYSCTL_INT(_net_local_dgram, OID_AUTO, maxdgram, CTLFLAG_RW,
    &unpdg_sendspace, 0, "Max datagram socket size");
SYSCTL_INT(_net_local_dgram, OID_AUTO, recvspace, CTLFLAG_RW,
    &unpdg_recvspace, 0, "Size of datagram socket receive buffer");

static int
unp_attach(struct socket *so, struct pru_attach_info *ai)
{
        struct unp_global_head *head;
        struct unpcb *unp;
        int error;

        lwkt_gettoken(&unp_token);

        if (so->so_snd.ssb_hiwat == 0 || so->so_rcv.ssb_hiwat == 0) {
                switch (so->so_type) {
                case SOCK_STREAM:
                case SOCK_SEQPACKET:
                        error = soreserve(so, unpst_sendspace, unpst_recvspace,
                                          ai->sb_rlimit);
                        break;

                case SOCK_DGRAM:
                        error = soreserve(so, unpdg_sendspace, unpdg_recvspace,
                                          ai->sb_rlimit);
                        break;

                default:
                        panic("unp_attach");
                }
                if (error)
                        goto failed;
        }

        /*
         * In order to support sendfile we have to set either SSB_STOPSUPP
         * or SSB_PREALLOC.  Unix domain sockets use the SSB_STOP flow
         * control mechanism.
         */
        if (so->so_type == SOCK_STREAM) {
                atomic_set_int(&so->so_rcv.ssb_flags, SSB_STOPSUPP);
                atomic_set_int(&so->so_snd.ssb_flags, SSB_STOPSUPP);
        }

        unp = kmalloc(sizeof(*unp), M_UNPCB, M_WAITOK | M_ZERO | M_NULLOK);
        if (unp == NULL) {
                error = ENOBUFS;
                goto failed;
        }
        unp->unp_refcnt = 1;
        unp->unp_gencnt = ++unp_gencnt;
        LIST_INIT(&unp->unp_refs);
        unp->unp_socket = so;
        unp->unp_rvnode = ai->fd_rdir;          /* jail cruft XXX JH */
        so->so_pcb = (caddr_t)unp;
        soreference(so);

        head = unp_globalhead(so->so_type);
        TAILQ_INSERT_TAIL(&head->list, unp, unp_link);
        head->count++;
        error = 0;
failed:
        lwkt_reltoken(&unp_token);
        return error;
}

static void
unp_detach(struct unpcb *unp)
{
        struct socket *so;

        lwkt_gettoken(&unp_token);
        lwkt_getpooltoken(unp);

        so = unp->unp_socket;

        unp->unp_gencnt = ++unp_gencnt;
        if (unp->unp_vnode) {
                unp->unp_vnode->v_socket = NULL;
                vrele(unp->unp_vnode);
                unp->unp_vnode = NULL;
        }
        soisdisconnected(so);
        KKASSERT(so->so_pcb == unp);
        so->so_pcb = NULL;              /* both tokens required */
        unp->unp_socket = NULL;

        lwkt_relpooltoken(unp);
        lwkt_reltoken(&unp_token);

        sofree(so);

        KASSERT(unp->unp_conn == NULL, ("unp is still connected"));
        KASSERT(LIST_EMPTY(&unp->unp_refs), ("unp still has references"));

        if (unp->unp_addr)
                kfree(unp->unp_addr, M_SONAME);
        kfree(unp, M_UNPCB);

        if (unp_rights)
                taskqueue_enqueue(unp_taskqueue, &unp_gc_task);
}

static int
unp_bind(struct unpcb *unp, struct sockaddr *nam, struct thread *td)
{
        struct proc *p = td->td_proc;
        struct sockaddr_un *soun = (struct sockaddr_un *)nam;
        struct vnode *vp;
        struct vattr vattr;
        int error, namelen;
        struct nlookupdata nd;
        char buf[SOCK_MAXADDRLEN];

        ASSERT_LWKT_TOKEN_HELD(&unp_token);
        UNP_ASSERT_TOKEN_HELD(unp);

        if (unp->unp_vnode != NULL)
                return EINVAL;

        namelen = soun->sun_len - offsetof(struct sockaddr_un, sun_path);
        if (namelen <= 0)
                return EINVAL;
        strncpy(buf, soun->sun_path, namelen);
        buf[namelen] = 0;       /* null-terminate the string */
        error = nlookup_init(&nd, buf, UIO_SYSSPACE,
                             NLC_LOCKVP | NLC_CREATE | NLC_REFDVP);
        if (error == 0)
                error = nlookup(&nd);
        if (error == 0 && nd.nl_nch.ncp->nc_vp != NULL)
                error = EADDRINUSE;
        if (error)
                goto done;

        VATTR_NULL(&vattr);
        vattr.va_type = VSOCK;
        vattr.va_mode = (ACCESSPERMS & ~p->p_fd->fd_cmask);
        error = VOP_NCREATE(&nd.nl_nch, nd.nl_dvp, &vp, nd.nl_cred, &vattr);
        if (error == 0) {
                if (unp->unp_vnode == NULL) {
                        vp->v_socket = unp->unp_socket;
                        unp->unp_vnode = vp;
                        unp->unp_addr = (struct sockaddr_un *)dup_sockaddr(nam);
                        vn_unlock(vp);
                } else {
                        vput(vp);               /* late race */
                        error = EINVAL;
                }
        }
done:
        nlookup_done(&nd);
        return (error);
}

static int
unp_connect(struct socket *so, struct sockaddr *nam, struct thread *td)
{
        struct unpcb *unp, *unp2;
        int error, flags = 0;

        lwkt_gettoken(&unp_token);

        unp = unp_getsocktoken(so);
        if (!UNP_ISATTACHED(unp)) {
                error = EINVAL;
                goto failed;
        }

        if ((unp->unp_flags & UNP_CONNECTING) || unp->unp_conn != NULL) {
                error = EISCONN;
                goto failed;
        }

        flags = UNP_CONNECTING;
        unp_setflags(unp, flags);

        error = unp_find_lockref(nam, td, so->so_type, &unp2);
        if (error)
                goto failed;
        /*
         * NOTE:
         * unp2 is locked and referenced.
         */

        if (so->so_proto->pr_flags & PR_CONNREQUIRED) {
                struct socket *so2, *so3;
                struct unpcb *unp3;

                so2 = unp2->unp_socket;
                if (!(so2->so_options & SO_ACCEPTCONN) ||
                    /* listen is not completed yet */
                    !(unp2->unp_flags & UNP_HAVEPCCACHED) ||
                    (so3 = sonewconn_faddr(so2, 0, NULL,
                     TRUE /* keep ref */)) == NULL) {
                        error = ECONNREFUSED;
                        goto done;
                }
                /* so3 has a socket reference. */

                unp3 = unp_getsocktoken(so3);
                if (!UNP_ISATTACHED(unp3)) {
                        unp_reltoken(unp3);
                        /*
                         * Already aborted; we only need to drop the
                         * socket reference held by sonewconn_faddr().
                         */
                        sofree(so3);
                        error = ECONNREFUSED;
                        goto done;
                }
                unp_reference(unp3);
                /*
                 * NOTE:
                 * unp3 is locked and referenced.
                 */

                /*
                 * Release so3 socket reference held by sonewconn_faddr().
                 * Since we have referenced unp3, neither unp3 nor so3 will
                 * be destroyed here.
                 */
                sofree(so3);

                if (unp2->unp_addr != NULL) {
                        unp3->unp_addr = (struct sockaddr_un *)
                            dup_sockaddr((struct sockaddr *)unp2->unp_addr);
                }

                /*
                 * unp_peercred management:
                 *
                 * The connecter's (client's) credentials are copied
                 * from its process structure at the time of connect()
                 * (which is now).
                 */
                cru2x(td->td_proc->p_ucred, &unp3->unp_peercred);
                unp_setflags(unp3, UNP_HAVEPC);
                /*
                 * The receiver's (server's) credentials are copied
                 * from the unp_peercred member of socket on which the
                 * former called listen(); unp_listen() cached that
                 * process's credentials at that time so we can use
                 * them now.
                 */
                KASSERT(unp2->unp_flags & UNP_HAVEPCCACHED,
                    ("unp_connect: listener without cached peercred"));
                memcpy(&unp->unp_peercred, &unp2->unp_peercred,
                    sizeof(unp->unp_peercred));
                unp_setflags(unp, UNP_HAVEPC);

                error = unp_connect_pair(unp, unp3);
                if (error)
                        soabort_direct(so3);

                /* Done with unp3 */
                unp_free(unp3);
                unp_reltoken(unp3);
        } else {
                error = unp_connect_pair(unp, unp2);
        }
done:
        unp_free(unp2);
        unp_reltoken(unp2);
failed:
        if (flags)
                unp_clrflags(unp, flags);
        unp_reltoken(unp);

        lwkt_reltoken(&unp_token);
        return (error);
}

/*
 * Connect two unix domain sockets together.
 *
 * NOTE: Semantics for any change to unp_conn requires that the per-unp
 *       pool token also be held.
 */
int
unp_connect2(struct socket *so, struct socket *so2)
{
        struct unpcb *unp, *unp2;
        int error;

        lwkt_gettoken(&unp_token);
        if (so2->so_type != so->so_type) {
                lwkt_reltoken(&unp_token);
                return (EPROTOTYPE);
        }
        unp = unp_getsocktoken(so);
        unp2 = unp_getsocktoken(so2);

        if (!UNP_ISATTACHED(unp)) {
                error = EINVAL;
                goto done;
        }
        if (!UNP_ISATTACHED(unp2)) {
                error = ECONNREFUSED;
                goto done;
        }

        if (unp->unp_conn != NULL) {
                error = EISCONN;
                goto done;
        }
        if ((so->so_type == SOCK_STREAM || so->so_type == SOCK_SEQPACKET) &&
            unp2->unp_conn != NULL) {
                error = EISCONN;
                goto done;
        }

        error = unp_connect_pair(unp, unp2);
done:
        unp_reltoken(unp2);
        unp_reltoken(unp);
        lwkt_reltoken(&unp_token);
        return (error);
}

/*
 * Disconnect a unix domain socket pair.
 *
 * NOTE: Semantics for any change to unp_conn requires that the per-unp
 *       pool token also be held.
 */
static void
unp_disconnect(struct unpcb *unp, int error)
{
        struct socket *so = unp->unp_socket;
        struct unpcb *unp2;

        ASSERT_LWKT_TOKEN_HELD(&unp_token);
        UNP_ASSERT_TOKEN_HELD(unp);

        if (error)
                so->so_error = error;

        while ((unp2 = unp->unp_conn) != NULL) {
                lwkt_getpooltoken(unp2);
                if (unp2 == unp->unp_conn)
                        break;
                lwkt_relpooltoken(unp2);
        }
        if (unp2 == NULL)
                return;
        /* unp2 is locked. */

        KASSERT((unp2->unp_flags & UNP_DROPPED) == 0, ("unp2 was dropped"));

        unp->unp_conn = NULL;

        switch (so->so_type) {
        case SOCK_DGRAM:
                LIST_REMOVE(unp, unp_reflink);
                soclrstate(so, SS_ISCONNECTED);
                break;

        case SOCK_STREAM:
        case SOCK_SEQPACKET:
                /*
                 * Keep a reference before clearing the unp_conn
                 * to avoid racing uipc_detach()/uipc_abort() in
                 * other thread.
                 */
                unp_reference(unp2);
                KASSERT(unp2->unp_conn == unp, ("unp_conn mismatch"));
                unp2->unp_conn = NULL;

                soisdisconnected(so);
                soisdisconnected(unp2->unp_socket);

                unp_free(unp2);
                break;
        }

        lwkt_relpooltoken(unp2);
}

#ifdef notdef
void
unp_abort(struct unpcb *unp)
{
        lwkt_gettoken(&unp_token);
        unp_free(unp);
        lwkt_reltoken(&unp_token);
}
#endif

static int
prison_unpcb(struct thread *td, struct unpcb *unp)
{
        struct proc *p;

        if (td == NULL)
                return (0);
        if ((p = td->td_proc) == NULL)
                return (0);
        if (!p->p_ucred->cr_prison)
                return (0);
        if (p->p_fd->fd_rdir == unp->unp_rvnode)
                return (0);
        return (1);
}

static int
unp_pcblist(SYSCTL_HANDLER_ARGS)
{
        struct unp_global_head *head = arg1;
        int error, i, n;
        struct unpcb *unp, *marker;

        KKASSERT(curproc != NULL);

        /*
         * The process of preparing the PCB list is too time-consuming and
         * resource-intensive to repeat twice on every request.
         */
        if (req->oldptr == NULL) {
                n = head->count;
                req->oldidx = (n + n/8) * sizeof(struct xunpcb);
                return 0;
        }

        if (req->newptr != NULL)
                return EPERM;

        marker = kmalloc(sizeof(*marker), M_UNPCB, M_WAITOK | M_ZERO);
        marker->unp_flags |= UNP_MARKER;

        lwkt_gettoken(&unp_token);

        n = head->count;
        i = 0;
        error = 0;

        TAILQ_INSERT_HEAD(&head->list, marker, unp_link);
        while ((unp = TAILQ_NEXT(marker, unp_link)) != NULL && i < n) {
                struct xunpcb xu;

                TAILQ_REMOVE(&head->list, marker, unp_link);
                TAILQ_INSERT_AFTER(&head->list, unp, marker, unp_link);

                if (unp->unp_flags & UNP_MARKER)
                        continue;
                if (prison_unpcb(req->td, unp))
                        continue;

                xu.xu_len = sizeof(xu);
                xu.xu_unpp = unp;

                /*
                 * NOTE:
                 * unp->unp_addr and unp->unp_conn are protected by
                 * unp_token.  So if we want to get rid of unp_token
                 * or reduce the coverage of unp_token, care must be
                 * taken.
                 */
                if (unp->unp_addr) {
                        bcopy(unp->unp_addr, &xu.xu_addr, 
                              unp->unp_addr->sun_len);
                }
                if (unp->unp_conn && unp->unp_conn->unp_addr) {
                        bcopy(unp->unp_conn->unp_addr,
                              &xu.xu_caddr,
                              unp->unp_conn->unp_addr->sun_len);
                }
                bcopy(unp, &xu.xu_unp, sizeof(*unp));
                sotoxsocket(unp->unp_socket, &xu.xu_socket);

                /* NOTE: This could block and temporarily release unp_token */
                error = SYSCTL_OUT(req, &xu, sizeof(xu));
                if (error)
                        break;
                ++i;
        }
        TAILQ_REMOVE(&head->list, marker, unp_link);

        lwkt_reltoken(&unp_token);

        kfree(marker, M_UNPCB);
        return error;
}

SYSCTL_PROC(_net_local_dgram, OID_AUTO, pcblist, CTLFLAG_RD, 
            &unp_dgram_head, 0, unp_pcblist, "S,xunpcb",
            "List of active local datagram sockets");
SYSCTL_PROC(_net_local_stream, OID_AUTO, pcblist, CTLFLAG_RD, 
            &unp_stream_head, 0, unp_pcblist, "S,xunpcb",
            "List of active local stream sockets");
SYSCTL_PROC(_net_local_seqpacket, OID_AUTO, pcblist, CTLFLAG_RD, 
            &unp_seqpkt_head, 0, unp_pcblist, "S,xunpcb",
            "List of active local seqpacket sockets");

static void
unp_shutdown(struct unpcb *unp)
{
        struct socket *so;

        if ((unp->unp_socket->so_type == SOCK_STREAM ||
             unp->unp_socket->so_type == SOCK_SEQPACKET) &&
            unp->unp_conn != NULL && (so = unp->unp_conn->unp_socket)) {
                socantrcvmore(so);
        }
}

#ifdef notdef
void
unp_drain(void)
{
        lwkt_gettoken(&unp_token);
        lwkt_reltoken(&unp_token);
}
#endif

int
unp_externalize(struct mbuf *rights, int flags)
{
        struct thread *td = curthread;
        struct proc *p = td->td_proc;           /* XXX */
        struct lwp *lp = td->td_lwp;
        struct cmsghdr *cm = mtod(rights, struct cmsghdr *);
        int *fdp;
        int i;
        struct file **rp;
        struct file *fp;
        int newfds = (cm->cmsg_len - (CMSG_DATA(cm) - (u_char *)cm))
                / sizeof(struct file *);
        int f;

        lwkt_gettoken(&unp_rights_token);

        /*
         * if the new FD's will not fit, then we free them all
         */
        if (!fdavail(p, newfds)) {
                rp = (struct file **)CMSG_DATA(cm);
                for (i = 0; i < newfds; i++) {
                        fp = *rp;
                        /*
                         * zero the pointer before calling unp_discard,
                         * since it may end up in unp_gc()..
                         */
                        *rp++ = NULL;
                        unp_discard(fp, NULL);
                }
                lwkt_reltoken(&unp_rights_token);
                return (EMSGSIZE);
        }

        /*
         * now change each pointer to an fd in the global table to 
         * an integer that is the index to the local fd table entry
         * that we set up to point to the global one we are transferring.
         * Since the sizeof(struct file *) is bigger than or equal to
         * the sizeof(int), we do it in forward order.  In that case,
         * an integer will always come in the same place or before its
         * corresponding struct file pointer.
         *
         * Hold revoke_token in 'shared' mode, so that we won't miss
         * the FREVOKED update on fps being externalized (fsetfd).
         */
        lwkt_gettoken_shared(&revoke_token);
        fdp = (int *)CMSG_DATA(cm);
        rp = (struct file **)CMSG_DATA(cm);
        for (i = 0; i < newfds; i++) {
                if (fdalloc(p, 0, &f)) {
                        int j;

                        /*
                         * Previous fdavail() can't garantee
                         * fdalloc() success due to SMP race.
                         * Just clean up and return the same
                         * error value as if fdavail() failed.
                         */
                        lwkt_reltoken(&revoke_token);

                        /* Close externalized files */
                        for (j = 0; j < i; j++)
                                kern_close(fdp[j]);
                        /* Discard the rest of internal files */
                        for (; i < newfds; i++)
                                unp_discard(rp[i], NULL);
                        /* Wipe out the control message */
                        for (i = 0; i < newfds; i++)
                                rp[i] = NULL;

                        lwkt_reltoken(&unp_rights_token);
                        return (EMSGSIZE);
                }
                fp = rp[i];
                unp_fp_externalize(lp, fp, f, flags);
                fdp[i] = f;
        }
        lwkt_reltoken(&revoke_token);

        lwkt_reltoken(&unp_rights_token);

        /*
         * Adjust length, in case sizeof(struct file *) and sizeof(int)
         * differs.
         */
        cm->cmsg_len = CMSG_LEN(newfds * sizeof(int));
        rights->m_len = cm->cmsg_len;

        return (0);
}

static void
unp_fp_externalize(struct lwp *lp, struct file *fp, int fd, int flags)
{
        if (lp) {
                struct filedesc *fdp = lp->lwp_proc->p_fd;

                KKASSERT(fd >= 0);
                if (fp->f_flag & FREVOKED) {
                        struct file *fx;
                        int error;

                        kprintf("Warning: revoked fp exiting unix socket\n");
                        error = falloc(lp, &fx, NULL);
                        if (error == 0) {
                                if (flags & MSG_CMSG_CLOEXEC)
                                        fdp->fd_files[fd].fileflags |= UF_EXCLOSE;
                                fsetfd(fdp, fx, fd);
                                fdrop(fx);
                        } else {
                                fsetfd(fdp, NULL, fd);
                        }
                } else {
                        if (flags & MSG_CMSG_CLOEXEC)
                                fdp->fd_files[fd].fileflags |= UF_EXCLOSE;
                        fsetfd(fdp, fp, fd);
                }
        }
        unp_del_right(fp);
        fdrop(fp);
}

void
unp_init(void)
{
        TAILQ_INIT(&unp_stream_head.list);
        TAILQ_INIT(&unp_dgram_head.list);
        TAILQ_INIT(&unp_seqpkt_head.list);

        SLIST_INIT(&unp_defdiscard_head);
        spin_init(&unp_defdiscard_spin, "unpdisc");
        TASK_INIT(&unp_defdiscard_task, 0, unp_defdiscard_taskfunc, NULL);

        /*
         * This implies that only one gc can be in-progress at any
         * given moment.
         */
        TASK_INIT(&unp_gc_task, 0, unp_gc, NULL);

        unp_gc_marker = kmalloc(sizeof(*unp_gc_marker), M_UNPCB,
            M_WAITOK | M_ZERO);
        unp_gc_marker->unp_flags |= UNP_MARKER;

        /*
         * Create taskqueue for defered discard, and stick it to
         * the last CPU.
         */
        unp_taskqueue = taskqueue_create("unp_taskq", M_WAITOK,
            taskqueue_thread_enqueue, &unp_taskqueue);
        taskqueue_start_threads(&unp_taskqueue, 1, TDPRI_KERN_DAEMON,
            ncpus - 1, "unp taskq");
}

static int
unp_internalize(struct mbuf *control, struct thread *td)
{
        struct proc *p = td->td_proc;
        struct filedesc *fdescp;
        struct cmsghdr *cm = mtod(control, struct cmsghdr *);
        struct file **rp;
        struct file *fp;
        int i, fd, *fdp;
        struct cmsgcred *cmcred;
        int oldfds;
        u_int newlen;
        int error;

        KKASSERT(p);

        if ((cm->cmsg_type != SCM_RIGHTS && cm->cmsg_type != SCM_CREDS) ||
            cm->cmsg_level != SOL_SOCKET ||
            CMSG_ALIGN(cm->cmsg_len) != control->m_len)
                return EINVAL;

        /*
         * Fill in credential information.
         */
        if (cm->cmsg_type == SCM_CREDS) {
                cmcred = (struct cmsgcred *)CMSG_DATA(cm);
                cmcred->cmcred_pid = p->p_pid;
                cmcred->cmcred_uid = p->p_ucred->cr_ruid;
                cmcred->cmcred_gid = p->p_ucred->cr_rgid;
                cmcred->cmcred_euid = p->p_ucred->cr_uid;
                cmcred->cmcred_ngroups = MIN(p->p_ucred->cr_ngroups,
                                                        CMGROUP_MAX);
                for (i = 0; i < cmcred->cmcred_ngroups; i++)
                        cmcred->cmcred_groups[i] = p->p_ucred->cr_groups[i];
                return 0;
        }

        /*
         * cmsghdr may not be aligned, do not allow calculation(s) to
         * go negative.
         */
        if (cm->cmsg_len < CMSG_LEN(0))
                return EINVAL;

        oldfds = (cm->cmsg_len - CMSG_LEN(0)) / sizeof(int);

        /*
         * Now replace the integer FDs with pointers to
         * the associated global file table entry..
         * Allocate a bigger buffer as necessary. But if an cluster is not
         * enough, return E2BIG.
         */
        newlen = CMSG_LEN(oldfds * sizeof(struct file *));
        if (newlen > MCLBYTES)
                return E2BIG;
        if (newlen - control->m_len > M_TRAILINGSPACE(control)) {
                if (control->m_flags & M_EXT)
                        return E2BIG;
                MCLGET(control, M_WAITOK);
                if (!(control->m_flags & M_EXT))
                        return ENOBUFS;

                /* copy the data to the cluster */
                memcpy(mtod(control, char *), cm, cm->cmsg_len);
                cm = mtod(control, struct cmsghdr *);
        }

        lwkt_gettoken(&unp_rights_token);

        fdescp = p->p_fd;
        spin_lock_shared(&fdescp->fd_spin);

        /*
         * check that all the FDs passed in refer to legal OPEN files
         * If not, reject the entire operation.
         */
        fdp = (int *)CMSG_DATA(cm);
        for (i = 0; i < oldfds; i++) {
                fd = *fdp++;
                if ((unsigned)fd >= fdescp->fd_nfiles ||
                    fdescp->fd_files[fd].fp == NULL) {
                        error = EBADF;
                        goto done;
                }
                if (fdescp->fd_files[fd].fp->f_type == DTYPE_KQUEUE) {
                        error = EOPNOTSUPP;
                        goto done;
                }
        }

        /*
         * Adjust length, in case sizeof(struct file *) and sizeof(int)
         * differs.
         */
        cm->cmsg_len = newlen;
        control->m_len = CMSG_ALIGN(newlen);

        /*
         * Transform the file descriptors into struct file pointers.
         * Since the sizeof(struct file *) is bigger than or equal to
         * the sizeof(int), we do it in reverse order so that the int
         * won't get trashed until we're done.
         */
        fdp = (int *)CMSG_DATA(cm) + oldfds - 1;
        rp = (struct file **)CMSG_DATA(cm) + oldfds - 1;
        for (i = 0; i < oldfds; i++) {
                fp = fdescp->fd_files[*fdp--].fp;
                *rp-- = fp;
                fhold(fp);
                unp_add_right(fp);
        }
        error = 0;
done:
        spin_unlock_shared(&fdescp->fd_spin);
        lwkt_reltoken(&unp_rights_token);
        return error;
}

#ifdef UNP_GC_ALLFILES

/*
 * Garbage collect in-transit file descriptors that get lost due to
 * loops (i.e. when a socket is sent to another process over itself,
 * and more complex situations).
 *
 * NOT MPSAFE - TODO socket flush code and maybe fdrop.  Rest is MPSAFE.
 */

struct unp_gc_info {
        struct file **extra_ref;
        struct file *locked_fp;
        int defer;
        int index;
        int maxindex;
};

static void
unp_gc(void *arg __unused, int pending __unused)
{
        struct unp_gc_info info;
        struct file **fpp;
        int i;

        lwkt_gettoken(&unp_rights_token);

        /* 
         * Before going through all this, set all FDs to be NOT defered
         * and NOT externally accessible (not marked).  During the scan
         * a fd can be marked externally accessible but we may or may not
         * be able to immediately process it (controlled by FDEFER).
         *
         * If we loop sleep a bit.  The complexity of the topology can cause
         * multiple loops.  Also failure to acquire the socket's so_rcv
         * token can cause us to loop.
         */
        allfiles_scan_exclusive(unp_gc_clearmarks, NULL);
        do {
                info.defer = 0;
                allfiles_scan_exclusive(unp_gc_checkmarks, &info);
                if (info.defer)
                        tsleep(&info, 0, "gcagain", 1);
        } while (info.defer);

        /*
         * We grab an extra reference to each of the file table entries
         * that are not otherwise accessible and then free the rights
         * that are stored in messages on them.
         *
         * The bug in the orginal code is a little tricky, so I'll describe
         * what's wrong with it here.
         *
         * It is incorrect to simply unp_discard each entry for f_msgcount
         * times -- consider the case of sockets A and B that contain
         * references to each other.  On a last close of some other socket,
         * we trigger a gc since the number of outstanding rights (unp_rights)
         * is non-zero.  If during the sweep phase the gc code unp_discards,
         * we end up doing a (full) fdrop on the descriptor.  A fdrop on A
         * results in the following chain.  Closef calls soo_close, which
         * calls soclose.   Soclose calls first (through the switch
         * uipc_usrreq) unp_detach, which re-invokes unp_gc.  Unp_gc simply
         * returns because the previous instance had set unp_gcing, and
         * we return all the way back to soclose, which marks the socket
         * with SS_NOFDREF, and then calls sofree.  Sofree calls sorflush
         * to free up the rights that are queued in messages on the socket A,
         * i.e., the reference on B.  The sorflush calls via the dom_dispose
         * switch unp_dispose, which unp_scans with unp_discard.  This second
         * instance of unp_discard just calls fdrop on B.
         *
         * Well, a similar chain occurs on B, resulting in a sorflush on B,
         * which results in another fdrop on A.  Unfortunately, A is already
         * being closed, and the descriptor has already been marked with
         * SS_NOFDREF, and soclose panics at this point.
         *
         * Here, we first take an extra reference to each inaccessible
         * descriptor.  Then, we call sorflush ourself, since we know
         * it is a Unix domain socket anyhow.  After we destroy all the
         * rights carried in messages, we do a last fdrop to get rid
         * of our extra reference.  This is the last close, and the
         * unp_detach etc will shut down the socket.
         *
         * 91/09/19, bsy@cs.cmu.edu
         */
        info.extra_ref = kmalloc(256 * sizeof(struct file *), M_FILE, M_WAITOK);
        info.maxindex = 256;

        do {
                /*
                 * Look for matches
                 */
                info.index = 0;
                allfiles_scan_exclusive(unp_gc_checkrefs, &info);

                /* 
                 * For each FD on our hit list, do the following two things
                 */
                for (i = info.index, fpp = info.extra_ref; --i >= 0; ++fpp) {
                        struct file *tfp = *fpp;
                        if (tfp->f_type == DTYPE_SOCKET && tfp->f_data != NULL)
                                sorflush((struct socket *)(tfp->f_data));
                }
                for (i = info.index, fpp = info.extra_ref; --i >= 0; ++fpp)
                        fdrop(*fpp);
        } while (info.index == info.maxindex);

        kfree((caddr_t)info.extra_ref, M_FILE);

        lwkt_reltoken(&unp_rights_token);
}

/*
 * MPSAFE - NOTE: filehead list and file pointer spinlocked on entry
 */
static int
unp_gc_checkrefs(struct file *fp, void *data)
{
        struct unp_gc_info *info = data;

        if (fp->f_count == 0)
                return(0);
        if (info->index == info->maxindex)
                return(-1);

        /* 
         * If all refs are from msgs, and it's not marked accessible
         * then it must be referenced from some unreachable cycle
         * of (shut-down) FDs, so include it in our
         * list of FDs to remove
         */
        if (fp->f_count == fp->f_msgcount && !(fp->f_flag & FMARK)) {
                info->extra_ref[info->index++] = fp;
                fhold(fp);
        }
        return(0);
}

/*
 * MPSAFE - NOTE: filehead list and file pointer spinlocked on entry
 */
static int
unp_gc_clearmarks(struct file *fp, void *data __unused)
{
        atomic_clear_int(&fp->f_flag, FMARK | FDEFER);
        return(0);
}

/*
 * MPSAFE - NOTE: filehead list and file pointer spinlocked on entry
 */
static int
unp_gc_checkmarks(struct file *fp, void *data)
{
        struct unp_gc_info *info = data;
        struct socket *so;

        /*
         * If the file is not open, skip it.  Make sure it isn't marked
         * defered or we could loop forever, in case we somehow race
         * something.
         */
        if (fp->f_count == 0) {
                if (fp->f_flag & FDEFER)
                        atomic_clear_int(&fp->f_flag, FDEFER);
                return(0);
        }
        /*
         * If we already marked it as 'defer'  in a
         * previous pass, then try process it this time
         * and un-mark it
         */
        if (fp->f_flag & FDEFER) {
                atomic_clear_int(&fp->f_flag, FDEFER);
        } else {
                /*
                 * if it's not defered, then check if it's
                 * already marked.. if so skip it
                 */
                if (fp->f_flag & FMARK)
                        return(0);
                /* 
                 * If all references are from messages
                 * in transit, then skip it. it's not 
                 * externally accessible.
                 */ 
                if (fp->f_count == fp->f_msgcount)
                        return(0);
                /* 
                 * If it got this far then it must be
                 * externally accessible.
                 */
                atomic_set_int(&fp->f_flag, FMARK);
        }

        /*
         * either it was defered, or it is externally 
         * accessible and not already marked so.
         * Now check if it is possibly one of OUR sockets.
         */ 
        if (fp->f_type != DTYPE_SOCKET ||
            (so = (struct socket *)fp->f_data) == NULL) {
                return(0);
        }
        if (so->so_proto->pr_domain != &localdomain ||
            !(so->so_proto->pr_flags & PR_RIGHTS)) {
                return(0);
        }

        /*
         * So, Ok, it's one of our sockets and it IS externally accessible
         * (or was defered).  Now we look to see if we hold any file
         * descriptors in its message buffers.  Follow those links and mark
         * them as accessible too.
         *
         * We are holding multiple spinlocks here, if we cannot get the
         * token non-blocking defer until the next loop.
         */
        info->locked_fp = fp;
        if (lwkt_trytoken(&so->so_rcv.ssb_token)) {
                unp_scan(so->so_rcv.ssb_mb, unp_mark, info);
                lwkt_reltoken(&so->so_rcv.ssb_token);
        } else {
                atomic_set_int(&fp->f_flag, FDEFER);
                ++info->defer;
        }
        return (0);
}

/*
 * Mark visibility.  info->defer is recalculated on every pass.
 */
static void
unp_mark(struct file *fp, void *data)
{
        struct unp_gc_info *info = data;

        if ((fp->f_flag & FMARK) == 0) {
                ++info->defer;
                atomic_set_int(&fp->f_flag, FMARK | FDEFER);
        } else if (fp->f_flag & FDEFER) {
                ++info->defer;
        }
}

#else   /* !UNP_GC_ALLFILES */

/*
 * They are thread local and do not require explicit synchronization.
 */
static int      unp_marked;
static int      unp_unreachable;

static void
unp_accessable(struct file *fp, void *data __unused)
{
        struct unpcb *unp;

        if ((unp = unp_fp2unpcb(fp)) == NULL)
                return;
        if (unp->unp_gcflags & UNPGC_REF)
                return;
        unp->unp_gcflags &= ~UNPGC_DEAD;
        unp->unp_gcflags |= UNPGC_REF;
        unp_marked++;
}

static void
unp_gc_process(struct unpcb *unp)
{
        struct file *fp;

        /* Already processed. */
        if (unp->unp_gcflags & UNPGC_SCANNED)
                return;
        fp = unp->unp_fp;

        /*
         * Check for a socket potentially in a cycle.  It must be in a
         * queue as indicated by msgcount, and this must equal the file
         * reference count.  Note that when msgcount is 0 the file is NULL.
         */
        if ((unp->unp_gcflags & UNPGC_REF) == 0 && fp &&
            unp->unp_msgcount != 0 && fp->f_count == unp->unp_msgcount) {
                unp->unp_gcflags |= UNPGC_DEAD;
                unp_unreachable++;
                return;
        }

        /*
         * Mark all sockets we reference with RIGHTS.
         */
        if (UNP_ISATTACHED(unp)) {
                struct signalsockbuf *ssb = &unp->unp_socket->so_rcv;

                unp_reference(unp);
                lwkt_gettoken(&ssb->ssb_token);
                /*
                 * unp_token would be temporarily dropped, if getting
                 * so_rcv token blocks, so we need to check unp state
                 * here again.
                 */
                if (UNP_ISATTACHED(unp))
                        unp_scan(ssb->ssb_mb, unp_accessable, NULL);
                lwkt_reltoken(&ssb->ssb_token);
                unp->unp_gcflags |= UNPGC_SCANNED;
                unp_free(unp);
        } else {
                unp->unp_gcflags |= UNPGC_SCANNED;
        }
}

static void
unp_gc(void *arg __unused, int pending __unused)
{
        struct unp_global_head *head;
        int h, filemax, fileidx, filetot;
        struct file **unref;
        struct unpcb *unp;

        lwkt_gettoken(&unp_rights_token);
        lwkt_gettoken(&unp_token);

        /*
         * First clear all gc flags from previous runs.
         */
        for (h = 0; unp_heads[h] != NULL; ++h) {
                /* 
                 * NOTE: This loop does not block, so it is safe
                 * to use TAILQ_FOREACH here.
                 */
                head = unp_heads[h];
                TAILQ_FOREACH(unp, &head->list, unp_link)
                        unp->unp_gcflags = 0;
        }

        /*
         * Scan marking all reachable sockets with UNPGC_REF.  Once a socket
         * is reachable all of the sockets it references are reachable.
         * Stop the scan once we do a complete loop without discovering
         * a new reachable socket.
         */
        do {
                unp_unreachable = 0;
                unp_marked = 0;
                for (h = 0; unp_heads[h] != NULL; ++h) {
                        head = unp_heads[h];
                        TAILQ_INSERT_HEAD(&head->list, unp_gc_marker, unp_link);
                        while ((unp = TAILQ_NEXT(unp_gc_marker, unp_link))
                            != NULL) {
                                TAILQ_REMOVE(&head->list, unp_gc_marker,
                                    unp_link);
                                TAILQ_INSERT_AFTER(&head->list, unp,
                                    unp_gc_marker, unp_link);

                                if (unp->unp_flags & UNP_MARKER)
                                        continue;
                                unp_gc_process(unp);
                        }
                        TAILQ_REMOVE(&head->list, unp_gc_marker, unp_link);
                }
        } while (unp_marked);

        if (unp_unreachable == 0)
                goto done;

        /*
         * We grab an extra reference to each of the file table entries
         * that are not otherwise accessible and then free the rights
         * that are stored in messages on them.
         *
         * The bug in the orginal code is a little tricky, so I'll describe
         * what's wrong with it here.
         *
         * It is incorrect to simply unp_discard each entry for f_msgcount
         * times -- consider the case of sockets A and B that contain
         * references to each other.  On a last close of some other socket,
         * we trigger a gc since the number of outstanding rights (unp_rights)
         * is non-zero.  If during the sweep phase the gc code unp_discards,
         * we end up doing a (full) fdrop on the descriptor.  A fdrop on A
         * results in the following chain.  Closef calls soo_close, which
         * calls soclose.   Soclose calls first (through the switch
         * uipc_usrreq) unp_detach, which re-invokes unp_gc.  Unp_gc simply
         * returns because the previous instance had set unp_gcing, and
         * we return all the way back to soclose, which marks the socket
         * with SS_NOFDREF, and then calls sofree.  Sofree calls sorflush
         * to free up the rights that are queued in messages on the socket A,
         * i.e., the reference on B.  The sorflush calls via the dom_dispose
         * switch unp_dispose, which unp_scans with unp_discard.  This second
         * instance of unp_discard just calls fdrop on B.
         *
         * Well, a similar chain occurs on B, resulting in a sorflush on B,
         * which results in another fdrop on A.  Unfortunately, A is already
         * being closed, and the descriptor has already been marked with
         * SS_NOFDREF, and soclose panics at this point.
         *
         * Here, we first take an extra reference to each inaccessible
         * descriptor.  Then, we call sorflush ourself, since we know
         * it is a Unix domain socket anyhow.  After we destroy all the
         * rights carried in messages, we do a last fdrop to get rid
         * of our extra reference.  This is the last close, and the
         * unp_detach etc will shut down the socket.
         *
         * 91/09/19, bsy@cs.cmu.edu
         */

        filemax = unp_unreachable;
        if (filemax > UNP_GCFILE_MAX)
                filemax = UNP_GCFILE_MAX;
        unref = kmalloc(filemax * sizeof(struct file *), M_TEMP, M_WAITOK);

        filetot = 0;
        do {
                int i;

                /*
                 * Iterate looking for sockets which have been specifically
                 * marked as as unreachable and store them locally.
                 */
                fileidx = 0;
                for (h = 0; unp_heads[h] != NULL; ++h) {
                        /*
                         * NOTE: This loop does not block, so it is safe
                         * to use TAILQ_FOREACH here.
                         */
                        head = unp_heads[h];
                        TAILQ_FOREACH(unp, &head->list, unp_link) {
                                struct file *fp;

                                if ((unp->unp_gcflags & UNPGC_DEAD) == 0)
                                        continue;
                                unp->unp_gcflags &= ~UNPGC_DEAD;

                                fp = unp->unp_fp;
                                if (unp->unp_msgcount == 0 || fp == NULL ||
                                    fp->f_count != unp->unp_msgcount)
                                        continue;
                                fhold(fp);

                                KASSERT(fileidx < filemax,
                                    ("invalid fileidx %d, filemax %d",
                                     fileidx, filemax));
                                unref[fileidx++] = fp;

                                KASSERT(filetot < unp_unreachable,
                                    ("invalid filetot %d and "
                                     "unp_unreachable %d",
                                     filetot, unp_unreachable));
                                ++filetot;

                                if (fileidx == filemax ||
                                    filetot == unp_unreachable)
                                        goto dogc;
                        }
                }
dogc:
                /*
                 * For each Unix domain socket on our hit list, do the
                 * following two things.
                 */
                for (i = 0; i < fileidx; ++i)
                        sorflush(unref[i]->f_data);
                for (i = 0; i < fileidx; ++i)
                        fdrop(unref[i]);
        } while (fileidx == filemax && filetot < unp_unreachable);
        kfree(unref, M_TEMP);
done:
        lwkt_reltoken(&unp_token);
        lwkt_reltoken(&unp_rights_token);
}

#endif  /* UNP_GC_ALLFILES */

/*
 * Dispose of the fp's stored in a mbuf.
 *
 * The dds loop can cause additional fps to be entered onto the
 * list while it is running, flattening out the operation and avoiding
 * a deep kernel stack recursion.
 */
void
unp_dispose(struct mbuf *m)
{
        lwkt_gettoken(&unp_rights_token);
        if (m)
                unp_scan(m, unp_discard, NULL);
        lwkt_reltoken(&unp_rights_token);
}

static int
unp_listen(struct unpcb *unp, struct thread *td)
{
        struct proc *p = td->td_proc;

        ASSERT_LWKT_TOKEN_HELD(&unp_token);
        UNP_ASSERT_TOKEN_HELD(unp);

        KKASSERT(p);
        cru2x(p->p_ucred, &unp->unp_peercred);
        unp_setflags(unp, UNP_HAVEPCCACHED);
        return (0);
}

static void
unp_scan(struct mbuf *m0, void (*op)(struct file *, void *), void *data)
{
        struct mbuf *m;
        struct file **rp;
        struct cmsghdr *cm;
        int i;
        int qfds;

        while (m0) {
                for (m = m0; m; m = m->m_next) {
                        if (m->m_type == MT_CONTROL &&
                            m->m_len >= sizeof(*cm)) {
                                cm = mtod(m, struct cmsghdr *);
                                if (cm->cmsg_level != SOL_SOCKET ||
                                    cm->cmsg_type != SCM_RIGHTS)
                                        continue;
                                qfds = (cm->cmsg_len - CMSG_LEN(0)) /
                                        sizeof(void *);
                                rp = (struct file **)CMSG_DATA(cm);
                                for (i = 0; i < qfds; i++)
                                        (*op)(*rp++, data);
                                break;          /* XXX, but saves time */
                        }
                }
                m0 = m0->m_nextpkt;
        }
}

/*
 * Discard a fp previously held in a unix domain socket mbuf.  To
 * avoid blowing out the kernel stack due to contrived chain-reactions
 * we may have to defer the operation to a dedicated taskqueue.
 *
 * Caller holds unp_rights_token.
 */
static void
unp_discard(struct file *fp, void *data __unused)
{
        unp_del_right(fp);
        if (unp_fp2unpcb(fp) != NULL) {
                struct unp_defdiscard *d;

                /*
                 * This fp is a Unix domain socket itself and fdrop()
                 * it here directly may cause deep unp_discard()
                 * recursion, so the fdrop() is defered to the
                 * dedicated taskqueue.
                 */
                d = kmalloc(sizeof(*d), M_UNPCB, M_WAITOK);
                d->fp = fp;

                spin_lock(&unp_defdiscard_spin);
                SLIST_INSERT_HEAD(&unp_defdiscard_head, d, next);
                spin_unlock(&unp_defdiscard_spin);

                taskqueue_enqueue(unp_taskqueue, &unp_defdiscard_task);
        } else {
                /* This fp is not a Unix domain socket */
                fdrop(fp);
        }
}

/*
 * NOTE:
 * unp_token must be held before calling this function to avoid name
 * resolution and v_socket accessing races, especially racing against
 * the unp_detach().
 *
 * NOTE:
 * For anyone caring about unconnected Unix domain socket sending
 * performance, other approach could be taken...
 */
static int
unp_find_lockref(struct sockaddr *nam, struct thread *td, short type,
    struct unpcb **unp_ret)
{
        struct proc *p = td->td_proc;
        struct sockaddr_un *soun = (struct sockaddr_un *)nam;
        struct vnode *vp = NULL;
        struct socket *so;
        struct unpcb *unp;
        int error, len;
        struct nlookupdata nd;
        char buf[SOCK_MAXADDRLEN];

        ASSERT_LWKT_TOKEN_HELD(&unp_token);

        *unp_ret = NULL;

        len = nam->sa_len - offsetof(struct sockaddr_un, sun_path);
        if (len <= 0) {
                error = EINVAL;
                goto failed;
        }
        strncpy(buf, soun->sun_path, len);
        buf[len] = 0;

        error = nlookup_init(&nd, buf, UIO_SYSSPACE, NLC_FOLLOW);
        if (error == 0)
                error = nlookup(&nd);
        if (error == 0)
                error = cache_vget(&nd.nl_nch, nd.nl_cred, LK_EXCLUSIVE, &vp);
        nlookup_done(&nd);
        if (error) {
                vp = NULL;
                goto failed;
        }

        if (vp->v_type != VSOCK) {
                error = ENOTSOCK;
                goto failed;
        }
        error = VOP_EACCESS(vp, VWRITE, p->p_ucred);
        if (error)
                goto failed;
        so = vp->v_socket;
        if (so == NULL) {
                error = ECONNREFUSED;
                goto failed;
        }
        if (so->so_type != type) {
                error = EPROTOTYPE;
                goto failed;
        }

        /* Lock this unp. */
        unp = unp_getsocktoken(so);
        if (!UNP_ISATTACHED(unp)) {
                unp_reltoken(unp);
                error = ECONNREFUSED;
                goto failed;
        }
        /* And keep this unp referenced. */
        unp_reference(unp);

        /* Done! */
        *unp_ret = unp;
        error = 0;
failed:
        if (vp != NULL)
                vput(vp);
        return error;
}

static int
unp_connect_pair(struct unpcb *unp, struct unpcb *unp2)
{
        struct socket *so = unp->unp_socket;
        struct socket *so2 = unp2->unp_socket;

        ASSERT_LWKT_TOKEN_HELD(&unp_token);
        UNP_ASSERT_TOKEN_HELD(unp);
        UNP_ASSERT_TOKEN_HELD(unp2);

        KASSERT(so->so_type == so2->so_type,
            ("socket type mismatch, so %d, so2 %d", so->so_type, so2->so_type));

        if (!UNP_ISATTACHED(unp))
                return EINVAL;
        if (!UNP_ISATTACHED(unp2))
                return ECONNREFUSED;

        KASSERT(unp->unp_conn == NULL, ("unp is already connected"));
        unp->unp_conn = unp2;

        switch (so->so_type) {
        case SOCK_DGRAM:
                LIST_INSERT_HEAD(&unp2->unp_refs, unp, unp_reflink);
                soisconnected(so);
                break;

        case SOCK_STREAM:
        case SOCK_SEQPACKET:
                KASSERT(unp2->unp_conn == NULL, ("unp2 is already connected"));
                unp2->unp_conn = unp;
                soisconnected(so);
                soisconnected(so2);
                break;

        default:
                panic("unp_connect_pair: unknown socket type %d", so->so_type);
        }
        return 0;
}

static void
unp_drop(struct unpcb *unp, int error)
{
        struct unp_global_head *head;
        struct unpcb *unp2;

        ASSERT_LWKT_TOKEN_HELD(&unp_token);
        UNP_ASSERT_TOKEN_HELD(unp);

        KASSERT((unp->unp_flags & (UNP_DETACHED | UNP_DROPPED)) == 0,
            ("unp is dropped"));

        /* Mark this unp as detached. */
        unp_setflags(unp, UNP_DETACHED);

        /* Remove this unp from the global unp list. */
        head = unp_globalhead(unp->unp_socket->so_type);
        KASSERT(head->count > 0, ("invalid unp count"));
        TAILQ_REMOVE(&head->list, unp, unp_link);
        head->count--;

        /* Disconnect all. */
        unp_disconnect(unp, error);
        while ((unp2 = LIST_FIRST(&unp->unp_refs)) != NULL) {
                lwkt_getpooltoken(unp2);
                unp_disconnect(unp2, ECONNRESET);
                lwkt_relpooltoken(unp2);
        }
        unp_setflags(unp, UNP_DROPPED);

        /* Try freeing this unp. */
        unp_free(unp);
}

static void
unp_defdiscard_taskfunc(void *arg __unused, int pending __unused)
{
        struct unp_defdiscard *d;

        spin_lock(&unp_defdiscard_spin);
        while ((d = SLIST_FIRST(&unp_defdiscard_head)) != NULL) {
                SLIST_REMOVE_HEAD(&unp_defdiscard_head, next);
                spin_unlock(&unp_defdiscard_spin);

                fdrop(d->fp);
                kfree(d, M_UNPCB);

                spin_lock(&unp_defdiscard_spin);
        }
        spin_unlock(&unp_defdiscard_spin);
}