[apple/xnu.git] / bsd / security / audit / audit_session.c

/*-
 * Copyright (c) 2008-2009 Apple Inc.
 * 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 Apple Inc. ("Apple") 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 APPLE AND ITS 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 APPLE OR ITS 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.
 */

#include <stdarg.h>

#include <sys/kernel.h>
#include <sys/fcntl.h>
#include <sys/kauth.h>
#include <sys/conf.h>
#include <sys/poll.h>
#include <sys/priv.h>
#include <sys/queue.h>
#include <sys/signalvar.h>
#include <sys/syscall.h>
#include <sys/sysent.h>
#include <sys/sysproto.h>
#include <sys/systm.h>
#include <sys/ucred.h>
#include <sys/user.h>

#include <miscfs/devfs/devfs.h>

#include <libkern/OSAtomic.h>

#include <bsm/audit.h>
#include <bsm/audit_internal.h>
#include <bsm/audit_kevents.h>

#include <security/audit/audit.h>
#include <security/audit/audit_bsd.h>
#include <security/audit/audit_ioctl.h>
#include <security/audit/audit_private.h>

#include <vm/vm_protos.h>
#include <mach/mach_port.h>
#include <kern/audit_sessionport.h>

#include <libkern/OSDebug.h>

/*
 * Audit Session Entry.  This is treated as an object with public and private
 * data.   The se_auinfo field is the only information that is public and
 * needs to be the first entry.
 */
struct au_sentry {
        auditinfo_addr_t        se_auinfo;      /* Public audit session data. */
#define se_asid         se_auinfo.ai_asid
#define se_auid         se_auinfo.ai_auid
#define se_mask         se_auinfo.ai_mask
#define se_termid       se_auinfo.ai_termid
#define se_flags        se_auinfo.ai_flags

        long                    se_refcnt;      /* Reference count. */
        long                    se_procnt;      /* Processes in session. */
        ipc_port_t              se_port;        /* Session port. */
        LIST_ENTRY(au_sentry)   se_link;        /* Hash bucket link list (1) */
};
typedef struct au_sentry au_sentry_t;

#define AU_SENTRY_PTR(aia_p)    ((au_sentry_t *)(aia_p))

/*
 * The default au_sentry/auditinfo_addr entry for ucred.
 */

static au_sentry_t audit_default_se = {
        .se_auinfo = {
                .ai_auid = AU_DEFAUDITID,
                .ai_asid = AU_DEFAUDITSID,
                .ai_termid = { .at_type = AU_IPv4, },
        },
        .se_refcnt = 1,
        .se_procnt = 1,
};

struct auditinfo_addr * const audit_default_aia_p = &audit_default_se.se_auinfo;

/* Copied from <ipc/ipc_object.h> */
#define IPC_OBJECT_COPYIN_FLAGS_ALLOW_IMMOVABLE_SEND 0x1
kern_return_t ipc_object_copyin(ipc_space_t, mach_port_name_t,
    mach_msg_type_name_t, ipc_port_t *, mach_port_context_t, mach_msg_guard_flags_t *, uint32_t);
void ipc_port_release_send(ipc_port_t);

#if CONFIG_AUDIT


/*
 * Currently the hash table is a fixed size.
 */
#define HASH_TABLE_SIZE         97
#define HASH_ASID(asid)         (audit_session_hash(asid) % HASH_TABLE_SIZE)

static struct rwlock    se_entry_lck;           /* (1) lock for se_link above */

LIST_HEAD(au_sentry_head, au_sentry);
static struct au_sentry_head *au_sentry_bucket = NULL;

#define AU_HISTORY_LOGGING 0
#if AU_HISTORY_LOGGING
typedef enum au_history_event {
        AU_HISTORY_EVENT_UNKNOWN = 0,
        AU_HISTORY_EVENT_REF     = 1,
        AU_HISTORY_EVENT_UNREF   = 2,
        AU_HISTORY_EVENT_BIRTH   = 3,
        AU_HISTORY_EVENT_DEATH   = 4,
        AU_HISTORY_EVENT_FIND    = 5
} au_history_event_t;

#define AU_HISTORY_MAX_STACK_DEPTH 8

struct au_history {
        struct au_sentry        *ptr;
        struct au_sentry         se;
        void                    *stack[AU_HISTORY_MAX_STACK_DEPTH];
        unsigned int             stack_depth;
        au_history_event_t       event;
};

static struct au_history *au_history;
static size_t             au_history_size = 65536;
static unsigned int       au_history_index;

static inline unsigned int
au_history_entries(void)
{
        if (au_history_index >= au_history_size) {
                return au_history_size;
        } else {
                return au_history_index;
        }
}

static inline void
au_history_record(au_sentry_t *se, au_history_event_t event)
{
        struct au_history *p;
        unsigned int i;

        i = OSAddAtomic(1, &au_history_index);
        p = &au_history[i % au_history_size];

        bzero(p, sizeof(*p));
        p->event = event;
        bcopy(se, &p->se, sizeof(p->se));
        p->stack_depth = OSBacktrace(&p->stack[0], AU_HISTORY_MAX_STACK_DEPTH);
        p->ptr = se;
}
#else
#define au_history_record(se, event) do {} while (0)
#endif

MALLOC_DEFINE(M_AU_SESSION, "audit_session", "Audit session data");

static void     audit_ref_session(au_sentry_t *se);
static void     audit_unref_session(au_sentry_t *se);

static void     audit_session_event(int event, auditinfo_addr_t *aia_p);

/*
 * Audit session device.
 */

static MALLOC_DEFINE(M_AUDIT_SDEV, "audit_sdev", "Audit sdevs");
static MALLOC_DEFINE(M_AUDIT_SDEV_ENTRY, "audit_sdevent",
    "Audit sdev entries and buffers");

/*
 * Default audit sdev buffer parameters.
 */
#define AUDIT_SDEV_QLIMIT_DEFAULT       128
#define AUDIT_SDEV_QLIMIT_MIN           1
#define AUDIT_SDEV_QLIMIT_MAX           1024

/*
 * Entry structure.
 */
struct  audit_sdev_entry {
        void                            *ase_record;
        u_int                            ase_record_len;
        TAILQ_ENTRY(audit_sdev_entry)    ase_queue;
};

/*
 * Per audit sdev structure.
 */

struct audit_sdev {
        int             asdev_open;

#define AUDIT_SDEV_ASYNC        0x00000001
#define AUDIT_SDEV_NBIO         0x00000002

#define AUDIT_SDEV_ALLSESSIONS  0x00010000
        u_int           asdev_flags;

        struct selinfo  asdev_selinfo;
        pid_t           asdev_sigio;

        au_id_t         asdev_auid;
        au_asid_t       asdev_asid;

        /* Per-sdev mutex for most fields in this struct. */
        struct mtx      asdev_mtx;

        /*
         * Per-sdev sleep lock serializing user-generated reads and
         * flushes. uiomove() is called to copy out the current head
         * record's data whie the record remains in the queue, so we
         * prevent other threads from removing it using this lock.
         */
        struct slck     asdev_sx;

        /*
         * Condition variable to signal when data has been delivered to
         * a sdev.
         */
        struct cv       asdev_cv;

        /* Count and bound of records in the queue. */
        u_int           asdev_qlen;
        u_int           asdev_qlimit;

        /* The number of bytes of data across all records. */
        u_int           asdev_qbyteslen;

        /*
         * The amount read so far of the first record in the queue.
         * (The number of bytes available for reading in the queue is
         * qbyteslen - qoffset.)
         */
        u_int           asdev_qoffset;

        /*
         * Per-sdev operation statistics.
         */
        u_int64_t       asdev_inserts;  /* Records added. */
        u_int64_t       asdev_reads;    /* Records read. */
        u_int64_t       asdev_drops;    /* Records dropped. */

        /*
         * Current pending record list.  This is protected by a
         * combination of asdev_mtx and asdev_sx.  Note that both
         * locks are required to remove a record from the head of the
         * queue, as an in-progress read may sleep while copying and,
         * therefore, cannot hold asdev_mtx.
         */
        TAILQ_HEAD(, audit_sdev_entry)  asdev_queue;

        /* Global sdev list. */
        TAILQ_ENTRY(audit_sdev)         asdev_list;
};

#define AUDIT_SDEV_LOCK(asdev)          mtx_lock(&(asdev)->asdev_mtx)
#define AUDIT_SDEV_LOCK_ASSERT(asdev)   mtx_assert(&(asdev)->asdev_mtx, \
                                            MA_OWNED)
#define AUDIT_SDEV_LOCK_DESTROY(asdev)  mtx_destroy(&(asdev)->asdev_mtx)
#define AUDIT_SDEV_LOCK_INIT(asdev)     mtx_init(&(asdev)->asdev_mtx, \
                                            "audit_sdev_mtx", NULL, MTX_DEF)
#define AUDIT_SDEV_UNLOCK(asdev)        mtx_unlock(&(asdev)->asdev_mtx)
#define AUDIT_SDEV_MTX(asdev)           (&(asdev)->asdev_mtx)

#define AUDIT_SDEV_SX_LOCK_DESTROY(asd) slck_destroy(&(asd)->asdev_sx)
#define AUDIT_SDEV_SX_LOCK_INIT(asd)    slck_init(&(asd)->asdev_sx, \
                                            "audit_sdev_sx")
#define AUDIT_SDEV_SX_XLOCK_ASSERT(asd) slck_assert(&(asd)->asdev_sx, \
                                            SA_XLOCKED)
#define AUDIT_SDEV_SX_XLOCK_SIG(asd)    slck_lock_sig(&(asd)->asdev_sx)
#define AUDIT_SDEV_SX_XUNLOCK(asd)      slck_unlock(&(asd)->asdev_sx)

/*
 * Cloning variables and constants.
 */
#define AUDIT_SDEV_NAME         "auditsessions"
#define MAX_AUDIT_SDEVS         32

static int audit_sdev_major;
static void *devnode;

/*
 * Global list of audit sdevs.  The list is protected by a rw lock.
 * Individaul record queues are protected by  per-sdev locks.  These
 * locks synchronize between threads walking the list to deliver to
 * individual sdevs and adds/removes of sdevs.
 */
static TAILQ_HEAD(, audit_sdev) audit_sdev_list;
static struct rwlock            audit_sdev_lock;

#define AUDIT_SDEV_LIST_LOCK_INIT()     rw_init(&audit_sdev_lock, \
                                            "audit_sdev_list_lock")
#define AUDIT_SDEV_LIST_RLOCK()         rw_rlock(&audit_sdev_lock)
#define AUDIT_SDEV_LIST_RUNLOCK()       rw_runlock(&audit_sdev_lock)
#define AUDIT_SDEV_LIST_WLOCK()         rw_wlock(&audit_sdev_lock)
#define AUDIT_SDEV_LIST_WLOCK_ASSERT()  rw_assert(&audit_sdev_lock, \
                                            RA_WLOCKED)
#define AUDIT_SDEV_LIST_WUNLOCK()       rw_wunlock(&audit_sdev_lock)

/*
 * dev_t doesn't have a pointer for "softc" data so we have to keep track of
 * it with the following global array (indexed by the minor number).
 *
 * XXX We may want to dynamically grow this as need.
 */
static struct audit_sdev        *audit_sdev_dtab[MAX_AUDIT_SDEVS];

/*
 * Special device methods and definition.
 */
static open_close_fcn_t         audit_sdev_open;
static open_close_fcn_t         audit_sdev_close;
static read_write_fcn_t         audit_sdev_read;
static ioctl_fcn_t              audit_sdev_ioctl;
static select_fcn_t             audit_sdev_poll;

static const struct cdevsw audit_sdev_cdevsw = {
        .d_open      =          audit_sdev_open,
        .d_close     =          audit_sdev_close,
        .d_read      =          audit_sdev_read,
        .d_write     =          eno_rdwrt,
        .d_ioctl     =          audit_sdev_ioctl,
        .d_stop      =          eno_stop,
        .d_reset     =          eno_reset,
        .d_ttys      =          NULL,
        .d_select    =          audit_sdev_poll,
        .d_mmap      =          eno_mmap,
        .d_strategy  =          eno_strat,
        .d_type      =          0
};

/*
 * Global statistics on audit sdevs.
 */
static int              audit_sdev_count;       /* Current number of sdevs. */
static u_int64_t        audit_sdev_ever;        /* Sdevs ever allocated. */
static u_int64_t        audit_sdev_records;     /* Total records seen. */
static u_int64_t        audit_sdev_drops;       /* Global record drop count. */

static int audit_sdev_init(void);

#define AUDIT_SENTRY_RWLOCK_INIT()      rw_init(&se_entry_lck, \
                                            "se_entry_lck")
#define AUDIT_SENTRY_RLOCK()            rw_rlock(&se_entry_lck)
#define AUDIT_SENTRY_WLOCK()            rw_wlock(&se_entry_lck)
#define AUDIT_SENTRY_RWLOCK_ASSERT()    rw_assert(&se_entry_lck, RA_LOCKED)
#define AUDIT_SENTRY_RUNLOCK()          rw_runlock(&se_entry_lck)
#define AUDIT_SENTRY_WUNLOCK()          rw_wunlock(&se_entry_lck)

/* Access control on the auditinfo_addr.ai_flags member. */
static uint64_t audit_session_superuser_set_sflags_mask;
static uint64_t audit_session_superuser_clear_sflags_mask;
static uint64_t audit_session_member_set_sflags_mask;
static uint64_t audit_session_member_clear_sflags_mask;
SYSCTL_NODE(, OID_AUTO, audit, CTLFLAG_RW | CTLFLAG_LOCKED, 0, "Audit controls");
SYSCTL_NODE(_audit, OID_AUTO, session, CTLFLAG_RW | CTLFLAG_LOCKED, 0, "Audit sessions");
SYSCTL_QUAD(_audit_session, OID_AUTO, superuser_set_sflags_mask, CTLFLAG_RW | CTLFLAG_LOCKED,
    &audit_session_superuser_set_sflags_mask,
    "Audit session flags settable by superuser");
SYSCTL_QUAD(_audit_session, OID_AUTO, superuser_clear_sflags_mask, CTLFLAG_RW | CTLFLAG_LOCKED,
    &audit_session_superuser_clear_sflags_mask,
    "Audit session flags clearable by superuser");
SYSCTL_QUAD(_audit_session, OID_AUTO, member_set_sflags_mask, CTLFLAG_RW | CTLFLAG_LOCKED,
    &audit_session_member_set_sflags_mask,
    "Audit session flags settable by a session member");
SYSCTL_QUAD(_audit_session, OID_AUTO, member_clear_sflags_mask, CTLFLAG_RW | CTLFLAG_LOCKED,
    &audit_session_member_clear_sflags_mask,
    "Audit session flags clearable by a session member");

extern int set_security_token_task_internal(proc_t p, void *task);

#define AUDIT_SESSION_DEBUG     0
#if     AUDIT_SESSION_DEBUG
/*
 * The following is debugging code that can be used to get a snapshot of the
 * session state.  The audit session information is read out using sysctl:
 *
 * error = sysctlbyname("kern.audit_session_debug", buffer_ptr, &buffer_len,
 *              NULL, 0);
 */
#include <kern/kalloc.h>

/*
 * The per session record structure for the snapshot data.
 */
struct au_sentry_debug {
        auditinfo_addr_t        se_auinfo;
        int64_t                 se_refcnt;      /* refereence count */
        int64_t                 se_procnt;      /* process count */
        int64_t                 se_ptcnt;       /* process count from
                                                 *  proc table */
};
typedef struct au_sentry_debug au_sentry_debug_t;

static int audit_sysctl_session_debug(struct sysctl_oid *oidp, void *arg1,
    int arg2, struct sysctl_req *req);

SYSCTL_PROC(_kern, OID_AUTO, audit_session_debug, CTLFLAG_RD | CTLFLAG_LOCKED,
    NULL, 0, audit_sysctl_session_debug, "S,audit_session_debug",
    "Current session debug info for auditing.");

/*
 * Callouts for proc_interate() which is used to reconcile the audit session
 * proc state information with the proc table.  We get everything we need
 * in the filterfn while the proc_lock() is held so we really don't need the
 * callout() function.
 */
static int
audit_session_debug_callout(__unused proc_t p, __unused void *arg)
{
        return PROC_RETURNED_DONE;
}

static int
audit_session_debug_filterfn(proc_t p, void *st)
{
        kauth_cred_t cred = p->p_ucred;
        auditinfo_addr_t *aia_p = cred->cr_audit.as_aia_p;
        au_sentry_debug_t *sed_tab = (au_sentry_debug_t *) st;
        au_sentry_debug_t  *sdtp;
        au_sentry_t *se;

        if (IS_VALID_SESSION(aia_p)) {
                sdtp = &sed_tab[0];
                do {
                        if (aia_p->ai_asid == sdtp->se_asid) {
                                sdtp->se_ptcnt++;

                                /* Do some santy checks. */
                                se = AU_SENTRY_PTR(aia_p);
                                if (se->se_refcnt != sdtp->se_refcnt) {
                                        sdtp->se_refcnt =
                                            (int64_t)se->se_refcnt;
                                }
                                if (se->se_procnt != sdtp->se_procnt) {
                                        sdtp->se_procnt =
                                            (int64_t)se->se_procnt;
                                }
                                break;
                        }
                        sdtp++;
                } while (sdtp->se_asid != 0 && sdtp->se_auid != 0);
        } else {
                /* add it to the default sesison */
                sed_tab->se_ptcnt++;
        }

        return 0;
}

/*
 * Copy out the session debug info via the sysctl interface.
 *
 */
static int
audit_sysctl_session_debug(__unused struct sysctl_oid *oidp,
    __unused void *arg1, __unused int arg2, struct sysctl_req *req)
{
        au_sentry_t *se;
        au_sentry_debug_t *sed_tab, *next_sed;
        int i, entry_cnt = 0;
        size_t sz;
        int err = 0;

        /*
         * This provides a read-only node.
         */
        if (req->newptr != USER_ADDR_NULL) {
                return EPERM;
        }

        /*
         * Walk the audit session hash table to determine the size.
         */
        AUDIT_SENTRY_RLOCK();
        for (i = 0; i < HASH_TABLE_SIZE; i++) {
                LIST_FOREACH(se, &au_sentry_bucket[i], se_link)
                if (se != NULL) {
                        entry_cnt++;
                }
        }

        entry_cnt++;  /* add one for the default entry */
        /*
         * If just querying then return the space required.  There is an
         * obvious race condition here so we just fudge this by 3 in case
         * the audit session table grows.
         */
        if (req->oldptr == USER_ADDR_NULL) {
                req->oldidx = (entry_cnt + 3) * sizeof(au_sentry_debug_t);
                AUDIT_SENTRY_RUNLOCK();
                return 0;
        }

        /*
         * Alloc a temporary buffer.
         */
        if (req->oldlen < (entry_cnt * sizeof(au_sentry_debug_t))) {
                AUDIT_SENTRY_RUNLOCK();
                return ENOMEM;
        }
        /*
         * We hold the lock over the alloc since we don't want the table to
         * grow on us.   Therefore, use the non-blocking version of kalloc().
         */
        sed_tab = (au_sentry_debug_t *)kheap_alloc(KHEAP_TEMP,
            entry_cnt * sizeof(au_sentry_debug_t), Z_NOWAIT | Z_ZERO);
        if (sed_tab == NULL) {
                AUDIT_SENTRY_RUNLOCK();
                return ENOMEM;
        }

        /*
         * Walk the audit session hash table and build the record array.
         */
        sz = 0;
        next_sed = sed_tab;
        /* add the first entry for processes not tracked in sessions. */
        bcopy(audit_default_aia_p, &next_sed->se_auinfo, sizeof(au_sentry_t));
        next_sed->se_refcnt = (int64_t)audit_default_se.se_refcnt;
        next_sed->se_procnt = (int64_t)audit_default_se.se_procnt;
        next_sed++;
        sz += sizeof(au_sentry_debug_t);
        for (i = 0; i < HASH_TABLE_SIZE; i++) {
                LIST_FOREACH(se, &au_sentry_bucket[i], se_link) {
                        if (se != NULL) {
                                next_sed->se_auinfo = se->se_auinfo;
                                next_sed->se_refcnt = (int64_t)se->se_refcnt;
                                next_sed->se_procnt = (int64_t)se->se_procnt;
                                next_sed++;
                                sz += sizeof(au_sentry_debug_t);
                        }
                }
        }
        AUDIT_SENTRY_RUNLOCK();

        /* Reconcile with the process table. */
        proc_iterate(PROC_ALLPROCLIST | PROC_ZOMBPROCLIST,
            audit_session_debug_callout, NULL,
            audit_session_debug_filterfn, (void *)&sed_tab[0]);


        req->oldlen = sz;
        err = SYSCTL_OUT(req, sed_tab, sz);
        kheap_free(KHEAP_TEMP, sed_tab, entry_cnt * sizeof(au_sentry_debug_t));

        return err;
}

#endif /* AUDIT_SESSION_DEBUG */

/*
 * Create and commit a session audit event. The proc and se arguments needs to
 * be that of the subject and not necessarily the current process.
 */
static void
audit_session_event(int event, auditinfo_addr_t *aia_p)
{
        struct kaudit_record *ar;

        KASSERT(AUE_SESSION_START == event || AUE_SESSION_UPDATE == event ||
            AUE_SESSION_END == event || AUE_SESSION_CLOSE == event,
            ("audit_session_event: invalid event: %d", event));

        if (NULL == aia_p) {
                return;
        }

        /*
         * Create a new audit record.  The record will contain the subject
         * ruid, rgid, egid, pid, auid, asid, amask, and term_addr
         * (implicitly added by audit_new).
         */
        ar = audit_new(event, PROC_NULL, /* Not used */ NULL);
        if (NULL == ar) {
                return;
        }

        /*
         * Audit session events are always generated because they are used
         * by some userland consumers so just set the preselect flag.
         */
        ar->k_ar_commit |= AR_PRESELECT_FILTER;

        /*
         * Populate the subject information.  Note that the ruid, rgid,
         * egid, and pid values are incorrect. We only need the  auditinfo_addr
         * information.
         */
        ar->k_ar.ar_subj_ruid = 0;
        ar->k_ar.ar_subj_rgid = 0;
        ar->k_ar.ar_subj_egid = 0;
        ar->k_ar.ar_subj_pid = 0;
        ar->k_ar.ar_subj_auid = aia_p->ai_auid;
        ar->k_ar.ar_subj_asid = aia_p->ai_asid;
        bcopy(&aia_p->ai_termid, &ar->k_ar.ar_subj_term_addr,
            sizeof(struct au_tid_addr));

        /* Add the audit masks to the record. */
        ar->k_ar.ar_arg_amask.am_success = aia_p->ai_mask.am_success;
        ar->k_ar.ar_arg_amask.am_failure = aia_p->ai_mask.am_failure;
        ARG_SET_VALID(ar, ARG_AMASK);

        /* Add the audit session flags to the record. */
        ar->k_ar.ar_arg_value64 = aia_p->ai_flags;
        ARG_SET_VALID(ar, ARG_VALUE64);


        /* Commit the record to the queue. */
        audit_commit(ar, 0, 0);
}

/*
 * Hash the audit session ID using a simple 32-bit mix.
 */
static inline uint32_t
audit_session_hash(au_asid_t asid)
{
        uint32_t a = (uint32_t) asid;

        a = (a - (a << 6)) ^ (a >> 17);
        a = (a - (a << 9)) ^ (a << 4);
        a = (a - (a << 3)) ^ (a << 10);
        a = a ^ (a >> 15);

        return a;
}

/*
 * Do an hash lookup and find the session entry for a given ASID.  Return NULL
 * if not found. If the session is found then audit_session_find takes a
 * reference.
 */
static au_sentry_t *
audit_session_find(au_asid_t asid)
{
        uint32_t         hkey;
        au_sentry_t     *found_se;

        AUDIT_SENTRY_RWLOCK_ASSERT();

        hkey = HASH_ASID(asid);

        LIST_FOREACH(found_se, &au_sentry_bucket[hkey], se_link)
        if (found_se->se_asid == asid) {
                au_history_record(found_se, AU_HISTORY_EVENT_FIND);
                audit_ref_session(found_se);
                return found_se;
        }
        return NULL;
}

/*
 * Remove the given audit_session entry from the hash table.
 */
static void
audit_session_remove(au_sentry_t *se)
{
        uint32_t         hkey;
        au_sentry_t     *found_se, *tmp_se;

        au_history_record(se, AU_HISTORY_EVENT_DEATH);
        KASSERT(se->se_refcnt == 0, ("audit_session_remove: ref count != 0"));
        KASSERT(se != &audit_default_se,
            ("audit_session_remove: removing default session"));

        hkey = HASH_ASID(se->se_asid);

        AUDIT_SENTRY_WLOCK();
        /*
         * Check and see if someone got a reference before we got the lock.
         */
        if (se->se_refcnt != 0) {
                AUDIT_SENTRY_WUNLOCK();
                return;
        }

        audit_session_portdestroy(&se->se_port);
        LIST_FOREACH_SAFE(found_se, &au_sentry_bucket[hkey], se_link, tmp_se) {
                if (found_se == se) {
                        /*
                         * Generate an audit event to notify userland of the
                         * session close.
                         */
                        audit_session_event(AUE_SESSION_CLOSE,
                            &found_se->se_auinfo);

                        LIST_REMOVE(found_se, se_link);
                        AUDIT_SENTRY_WUNLOCK();
                        free(found_se, M_AU_SESSION);

                        return;
                }
        }
        AUDIT_SENTRY_WUNLOCK();
}

/*
 * Reference the session by incrementing the sentry ref count.
 */
static void
audit_ref_session(au_sentry_t *se)
{
        long old_val;

        if (se == NULL || se == &audit_default_se) {
                return;
        }

        au_history_record(se, AU_HISTORY_EVENT_REF);

        old_val = OSAddAtomicLong(1, &se->se_refcnt);
        KASSERT(old_val < 100000,
            ("audit_ref_session: Too many references on session."));
}

/*
 * Decrement the sentry ref count and remove the session entry if last one.
 */
static void
audit_unref_session(au_sentry_t *se)
{
        long old_val;

        if (se == NULL || se == &audit_default_se) {
                return;
        }

        au_history_record(se, AU_HISTORY_EVENT_UNREF);

        old_val = OSAddAtomicLong(-1, &se->se_refcnt);
        if (old_val == 1) {
                audit_session_remove(se);
        }
        KASSERT(old_val > 0,
            ("audit_unref_session: Too few references on session."));
}

/*
 * Increment the process count in the session.
 */
static void
audit_inc_procount(au_sentry_t *se)
{
        long old_val;

        if (se == NULL || se == &audit_default_se) {
                return;
        }

        old_val = OSAddAtomicLong(1, &se->se_procnt);
        KASSERT(old_val <= PID_MAX,
            ("audit_inc_procount: proc count > PID_MAX"));
}

/*
 * Decrement the process count and add a knote if it is the last process
 * to exit the session.
 */
static void
audit_dec_procount(au_sentry_t *se)
{
        long old_val;

        if (se == NULL || se == &audit_default_se) {
                return;
        }

        old_val = OSAddAtomicLong(-1, &se->se_procnt);
        /*
         * If this was the last process generate an audit event to notify
         * userland of the session ending.
         */
        if (old_val == 1) {
                audit_session_event(AUE_SESSION_END, &se->se_auinfo);
        }
        KASSERT(old_val >= 1,
            ("audit_dec_procount: proc count < 0"));
}

/*
 * Update the session entry and check to see if anything was updated.
 * Returns:
 *    0    Nothing was updated (We don't care about process preselection masks)
 *    1    Something was updated.
 */
static int
audit_update_sentry(au_sentry_t *se, auditinfo_addr_t *new_aia)
{
        auditinfo_addr_t *aia = &se->se_auinfo;
        int update;

        KASSERT(new_aia != audit_default_aia_p,
            ("audit_update_sentry: Trying to update the default aia."));

        update = (aia->ai_auid != new_aia->ai_auid ||
            bcmp(&aia->ai_termid, &new_aia->ai_termid,
            sizeof(new_aia->ai_termid)) ||
            aia->ai_flags != new_aia->ai_flags);

        if (update) {
                bcopy(new_aia, aia, sizeof(*aia));
        }

        return update;
}

/*
 * Return the next session ID.  The range of kernel generated audit session IDs
 * is ASSIGNED_ASID_MIN to ASSIGNED_ASID_MAX.
 */
static uint32_t
audit_session_nextid(void)
{
        static uint32_t next_asid = ASSIGNED_ASID_MIN;

        AUDIT_SENTRY_RWLOCK_ASSERT();

        if (next_asid > ASSIGNED_ASID_MAX) {
                next_asid = ASSIGNED_ASID_MIN;
        }

        return next_asid++;
}

/*
 * Allocated a new audit_session entry and add it to the hash table.  If the
 * given ASID is set to AU_ASSIGN_ASID then audit_session_new() will pick an
 * audit session ID.  Otherwise, it attempts use the one given. It creates a
 * reference to the entry that must be unref'ed.
 */
static auditinfo_addr_t *
audit_session_new(auditinfo_addr_t *new_aia_p, auditinfo_addr_t *old_aia_p)
{
        au_asid_t new_asid;
        au_sentry_t *se = NULL;
        au_sentry_t *found_se = NULL;
        auditinfo_addr_t *aia = NULL;

        KASSERT(new_aia_p != NULL, ("audit_session_new: new_aia_p == NULL"));

        new_asid = new_aia_p->ai_asid;

        /*
         * Alloc a new session entry now so we don't wait holding the lock.
         */
        se = malloc(sizeof(au_sentry_t), M_AU_SESSION, M_WAITOK | M_ZERO);

        /*
         * Find an unique session ID, if desired.
         */
        AUDIT_SENTRY_WLOCK();
        if (new_asid == AU_ASSIGN_ASID) {
                do {
                        new_asid = (au_asid_t)audit_session_nextid();
                        found_se = audit_session_find(new_asid);

                        /*
                         * If the session ID is currently active then drop the
                         * reference and try again.
                         */
                        if (found_se != NULL) {
                                audit_unref_session(found_se);
                        } else {
                                break;
                        }
                } while (1);
        } else {
                /*
                 * Check to see if the requested ASID is already in the
                 * hash table.  If so, update it with the new auditinfo.
                 */
                if ((found_se = audit_session_find(new_asid)) != NULL) {
                        int updated;

                        updated = audit_update_sentry(found_se, new_aia_p);

                        AUDIT_SENTRY_WUNLOCK();
                        free(se, M_AU_SESSION);

                        /* If a different session then add this process in. */
                        if (new_aia_p != old_aia_p) {
                                audit_inc_procount(found_se);
                        }

                        /*
                         * If the session information was updated then
                         * generate an audit event to notify userland.
                         */
                        if (updated) {
                                audit_session_event(AUE_SESSION_UPDATE,
                                    &found_se->se_auinfo);
                        }

                        return &found_se->se_auinfo;
                }
        }

        /*
         * Start the reference and proc count at 1 to account for the process
         * that invoked this via setaudit_addr() (or friends).
         */
        se->se_refcnt = se->se_procnt = 1;

        /*
         * Populate the new session entry.  Note that process masks are stored
         * in kauth ucred so just zero them here.
         */
        se->se_port = IPC_PORT_NULL;
        aia = &se->se_auinfo;
        aia->ai_asid = new_asid;
        aia->ai_auid = new_aia_p->ai_auid;
        bzero(&new_aia_p->ai_mask, sizeof(new_aia_p->ai_mask));
        bcopy(&new_aia_p->ai_termid, &aia->ai_termid, sizeof(aia->ai_termid));
        aia->ai_flags = new_aia_p->ai_flags;

        /*
         * Add it to the hash table.
         */
        LIST_INSERT_HEAD(&au_sentry_bucket[HASH_ASID(new_asid)], se, se_link);
        AUDIT_SENTRY_WUNLOCK();

        /*
         * Generate an audit event to notify userland of the new session.
         */
        audit_session_event(AUE_SESSION_START, aia);
        au_history_record(se, AU_HISTORY_EVENT_BIRTH);
        return aia;
}

/*
 * Lookup an existing session.  A copy of the audit session info for a given
 * ASID is returned in ret_aia.  Returns 0 on success.
 */
int
audit_session_lookup(au_asid_t asid, auditinfo_addr_t *ret_aia)
{
        au_sentry_t *se = NULL;

        if ((uint32_t)asid > ASSIGNED_ASID_MAX) {
                return -1;
        }
        AUDIT_SENTRY_RLOCK();
        if ((se = audit_session_find(asid)) == NULL) {
                AUDIT_SENTRY_RUNLOCK();
                return 1;
        }
        /* We have a reference on the session so it is safe to drop the lock. */
        AUDIT_SENTRY_RUNLOCK();
        if (ret_aia != NULL) {
                bcopy(&se->se_auinfo, ret_aia, sizeof(*ret_aia));
        }
        audit_unref_session(se);

        return 0;
}

void
audit_session_aiaref(auditinfo_addr_t *aia_p)
{
        audit_ref_session(AU_SENTRY_PTR(aia_p));
}

/*
 * Add a reference to the session entry.
 */
void
audit_session_ref(kauth_cred_t cred)
{
        auditinfo_addr_t *aia_p;

        KASSERT(IS_VALID_CRED(cred),
            ("audit_session_ref: Invalid kauth_cred."));

        aia_p = cred->cr_audit.as_aia_p;
        audit_session_aiaref(aia_p);
}

void
audit_session_aiaunref(auditinfo_addr_t *aia_p)
{
        audit_unref_session(AU_SENTRY_PTR(aia_p));
}

/*
 * Remove a reference to the session entry.
 */
void
audit_session_unref(kauth_cred_t cred)
{
        auditinfo_addr_t *aia_p;

        KASSERT(IS_VALID_CRED(cred),
            ("audit_session_unref: Invalid kauth_cred."));

        aia_p = cred->cr_audit.as_aia_p;
        audit_session_aiaunref(aia_p);
}

/*
 * Increment the per audit session process count.  Assumes that the caller has
 * a reference on the process' cred.
 */
void
audit_session_procnew(proc_t p)
{
        kauth_cred_t cred = p->p_ucred;
        auditinfo_addr_t *aia_p;

        KASSERT(IS_VALID_CRED(cred),
            ("audit_session_procnew: Invalid kauth_cred."));

        aia_p = cred->cr_audit.as_aia_p;

        audit_inc_procount(AU_SENTRY_PTR(aia_p));
}

/*
 * Decrement the per audit session process count.  Assumes that the caller has
 * a reference on the cred.
 */
void
audit_session_procexit(proc_t p)
{
        kauth_cred_t cred = p->p_ucred;
        auditinfo_addr_t *aia_p;

        KASSERT(IS_VALID_CRED(cred),
            ("audit_session_procexit: Invalid kauth_cred."));

        aia_p = cred->cr_audit.as_aia_p;

        audit_dec_procount(AU_SENTRY_PTR(aia_p));
}

/*
 * Init the audit session code.
 */
void
audit_session_init(void)
{
        int i;

        KASSERT((ASSIGNED_ASID_MAX - ASSIGNED_ASID_MIN) > PID_MAX,
            ("audit_session_init: ASSIGNED_ASID_MAX is not large enough."));

        AUDIT_SENTRY_RWLOCK_INIT();

        au_sentry_bucket = malloc( sizeof(struct au_sentry) *
            HASH_TABLE_SIZE, M_AU_SESSION, M_WAITOK | M_ZERO);

        for (i = 0; i < HASH_TABLE_SIZE; i++) {
                LIST_INIT(&au_sentry_bucket[i]);
        }

        (void)audit_sdev_init();
#if AU_HISTORY_LOGGING
        au_history = malloc(sizeof(struct au_history) * au_history_size,
            M_AU_SESSION, M_WAITOK | M_ZERO);
#endif
}

static int
audit_session_update_check(kauth_cred_t cred, auditinfo_addr_t *old,
    auditinfo_addr_t *new)
{
        uint64_t n;

        /* If the current audit ID is not the default then it is immutable. */
        if (old->ai_auid != AU_DEFAUDITID && old->ai_auid != new->ai_auid) {
                return EINVAL;
        }

        /* If the current termid is not the default then it is immutable. */
        if ((old->ai_termid.at_type != AU_IPv4 ||
            old->ai_termid.at_port != 0 ||
            old->ai_termid.at_addr[0] != 0) &&
            (old->ai_termid.at_port != new->ai_termid.at_port ||
            old->ai_termid.at_type != new->ai_termid.at_type ||
            0 != bcmp(&old->ai_termid.at_addr, &new->ai_termid.at_addr,
            sizeof(old->ai_termid.at_addr)))) {
                return EINVAL;
        }

        /* The flags may be set only according to the
         * audit_session_*_set_sflags_masks.
         */
        n = ~old->ai_flags & new->ai_flags;
        if (0 != n &&
            !((n == (audit_session_superuser_set_sflags_mask & n) &&
            kauth_cred_issuser(cred)) ||
            (n == (audit_session_member_set_sflags_mask & n) &&
            old->ai_asid == new->ai_asid))) {
                return EINVAL;
        }

        /* The flags may be cleared only according to the
         * audit_session_*_clear_sflags_masks.
         */
        n = ~new->ai_flags & old->ai_flags;
        if (0 != n &&
            !((n == (audit_session_superuser_clear_sflags_mask & n) &&
            kauth_cred_issuser(cred)) ||
            (n == (audit_session_member_clear_sflags_mask & n) &&
            old->ai_asid == new->ai_asid))) {
                return EINVAL;
        }

        /* The audit masks are mutable. */
        return 0;
}

/*
 * Safely update kauth cred of the given process with new the given audit info.
 */
int
audit_session_setaia(proc_t p, auditinfo_addr_t *new_aia_p)
{
        kauth_cred_t my_cred, my_new_cred;
        struct au_session  as;
        struct au_session  tmp_as;
        auditinfo_addr_t caia, *old_aia_p;
        int ret;

        /*
         * If this is going to modify an existing session then do some
         * immutable checks.
         */
        if (audit_session_lookup(new_aia_p->ai_asid, &caia) == 0) {
                my_cred = kauth_cred_proc_ref(p);
                ret = audit_session_update_check(my_cred, &caia, new_aia_p);
                kauth_cred_unref(&my_cred);
                if (ret) {
                        return ret;
                }
        }

        my_cred = kauth_cred_proc_ref(p);
        bcopy(&new_aia_p->ai_mask, &as.as_mask, sizeof(as.as_mask));
        old_aia_p = my_cred->cr_audit.as_aia_p;
        /* audit_session_new() adds a reference on the session */
        as.as_aia_p = audit_session_new(new_aia_p, old_aia_p);

        /* If the process left a session then update the process count. */
        if (old_aia_p != new_aia_p) {
                audit_dec_procount(AU_SENTRY_PTR(old_aia_p));
        }


        /*
         * We are modifying the audit info in a credential so we need a new
         * credential (or take another reference on an existing credential that
         * matches our new one).  We must do this because the audit info in the
         * credential is used as part of our hash key.  Get current credential
         * in the target process and take a reference while we muck with it.
         */
        for (;;) {
                /*
                 * Set the credential with new info.  If there is no change,
                 * we get back the same credential we passed in; if there is
                 * a change, we drop the reference on the credential we
                 * passed in.  The subsequent compare is safe, because it is
                 * a pointer compare rather than a contents compare.
                 */
                bcopy(&as, &tmp_as, sizeof(tmp_as));
                my_new_cred = kauth_cred_setauditinfo(my_cred, &tmp_as);

                if (my_cred != my_new_cred) {
                        proc_ucred_lock(p);
                        /* Need to protect for a race where another thread also
                         * changed the credential after we took our reference.
                         * If p_ucred has changed then we should restart this
                         * again with the new cred.
                         */
                        if (p->p_ucred != my_cred) {
                                proc_ucred_unlock(p);
                                audit_session_unref(my_new_cred);
                                kauth_cred_unref(&my_new_cred);
                                /* try again */
                                my_cred = kauth_cred_proc_ref(p);
                                continue;
                        }
                        p->p_ucred = my_new_cred;
                        /* update cred on proc */
                        PROC_UPDATE_CREDS_ONPROC(p);
                        proc_ucred_unlock(p);
                }
                /*
                 * Drop old proc reference or our extra reference.
                 */
                kauth_cred_unref(&my_cred);
                break;
        }

        /* Drop the reference taken by audit_session_new() above. */
        audit_unref_session(AU_SENTRY_PTR(as.as_aia_p));

        /* Propagate the change from the process to the Mach task. */
        set_security_token(p);

        return 0;
}

/*
 * audit_session_self  (system call)
 *
 * Description: Obtain a Mach send right for the current session.
 *
 * Parameters:  p               Process calling audit_session_self().
 *
 * Returns:     *ret_port       Named Mach send right, which may be
 *                              MACH_PORT_NULL in the failure case.
 *
 * Errno:       0               Success
 *              EINVAL          The calling process' session has not be set.
 *              ESRCH           Bad process, can't get valid cred for process.
 *              ENOMEM          Port allocation failed due to no free memory.
 */
int
audit_session_self(proc_t p, __unused struct audit_session_self_args *uap,
    mach_port_name_t *ret_port)
{
        ipc_port_t sendport = IPC_PORT_NULL;
        kauth_cred_t cred = NULL;
        auditinfo_addr_t *aia_p;
        au_sentry_t *se;
        int err = 0;

        cred = kauth_cred_proc_ref(p);
        if (!IS_VALID_CRED(cred)) {
                err = ESRCH;
                goto done;
        }

        aia_p = cred->cr_audit.as_aia_p;
        if (!IS_VALID_SESSION(aia_p)) {
                /* Can't join the default session. */
                err = EINVAL;
                goto done;
        }

        se = AU_SENTRY_PTR(aia_p);

        /*
         * Processes that join using this mach port will inherit this process'
         * pre-selection masks.
         */
        if (se->se_port == IPC_PORT_NULL) {
                bcopy(&cred->cr_audit.as_mask, &se->se_mask,
                    sizeof(se->se_mask));
        }

        /*
         * Get a send right to the session's Mach port and insert it in the
         * process' mach port namespace.
         */
        sendport = audit_session_mksend(aia_p, &se->se_port);
        *ret_port = ipc_port_copyout_send(sendport, get_task_ipcspace(p->task));

done:
        if (cred != NULL) {
                kauth_cred_unref(&cred);
        }
        if (err != 0) {
                *ret_port = MACH_PORT_NULL;
        }
        return err;
}

/*
 * audit_session_port  (system call)
 *
 * Description: Obtain a Mach send right for the given session ID.
 *
 * Parameters:  p               Process calling audit_session_port().
 *              uap->asid       The target audit session ID.  The special
 *                              value -1 can be used to target the process's
 *                              own session.
 *              uap->portnamep  User address at which to place port name.
 *
 * Returns:     0               Success
 *              EINVAL          The calling process' session has not be set.
 *              EINVAL          The given session ID could not be found.
 *              EINVAL          The Mach port right could not be copied out.
 *              ESRCH           Bad process, can't get valid cred for process.
 *              EPERM           Only the superuser can reference sessions other
 *                              than the process's own.
 *              ENOMEM          Port allocation failed due to no free memory.
 */
int
audit_session_port(proc_t p, struct audit_session_port_args *uap,
    __unused int *retval)
{
        ipc_port_t sendport = IPC_PORT_NULL;
        mach_port_name_t portname = MACH_PORT_NULL;
        kauth_cred_t cred = NULL;
        auditinfo_addr_t *aia_p = NULL;
        au_sentry_t *se = NULL;
        int err = 0;

        /* Note: Currently this test will never be true, because
         * ASSIGNED_ASID_MAX is effectively (uint32_t)-2.
         */
        if (uap->asid != -1 && (uint32_t)uap->asid > ASSIGNED_ASID_MAX) {
                err = EINVAL;
                goto done;
        }
        cred = kauth_cred_proc_ref(p);
        if (!IS_VALID_CRED(cred)) {
                err = ESRCH;
                goto done;
        }
        aia_p = cred->cr_audit.as_aia_p;

        /* Find the session corresponding to the requested audit
         * session ID.  If found, take a reference on it so that
         * the session is not dropped until the join is later done.
         */
        if (uap->asid == (au_asid_t)-1 ||
            uap->asid == aia_p->ai_asid) {
                if (!IS_VALID_SESSION(aia_p)) {
                        /* Can't join the default session. */
                        err = EINVAL;
                        goto done;
                }

                /* No privilege is required to obtain a port for our
                 * own session.
                 */
                se = AU_SENTRY_PTR(aia_p);
                audit_ref_session(se);
        } else {
                /*
                 * Only privileged processes may obtain a port for
                 * any existing session.
                 */
                err = priv_check_cred(cred, PRIV_AUDIT_SESSION_PORT, 0);
                if (err != 0) {
                        goto done;
                }
                AUDIT_SENTRY_RLOCK();
                se = audit_session_find(uap->asid);
                AUDIT_SENTRY_RUNLOCK();
                if (NULL == se) {
                        err = EINVAL;
                        goto done;
                }
                aia_p = &se->se_auinfo;
        }

        /*
         * Processes that join using this mach port will inherit this process'
         * pre-selection masks.
         */
        if (se->se_port == IPC_PORT_NULL) {
                bcopy(&cred->cr_audit.as_mask, &se->se_mask,
                    sizeof(se->se_mask));
        }

        /*
         * Use the session reference to create a mach port reference for the
         * session (at which point we are free to drop the session reference)
         * and then copy out the mach port to the process' mach port namespace.
         */
        sendport = audit_session_mksend(aia_p, &se->se_port);
        portname = ipc_port_copyout_send(sendport, get_task_ipcspace(p->task));
        if (!MACH_PORT_VALID(portname)) {
                err = EINVAL;
                goto done;
        }
        err = copyout(&portname, uap->portnamep, sizeof(mach_port_name_t));
done:
        if (cred != NULL) {
                kauth_cred_unref(&cred);
        }
        if (NULL != se) {
                audit_unref_session(se);
        }
        if (MACH_PORT_VALID(portname) && 0 != err) {
                (void)mach_port_deallocate(get_task_ipcspace(p->task),
                    portname);
        }

        return err;
}

static int
audit_session_join_internal(proc_t p, task_t task, ipc_port_t port, au_asid_t *new_asid)
{
        auditinfo_addr_t *new_aia_p, *old_aia_p;
        kauth_cred_t my_cred = NULL;
        au_asid_t old_asid;
        int err = 0;

        *new_asid = AU_DEFAUDITSID;

        if ((new_aia_p = audit_session_porttoaia(port)) == NULL) {
                err = EINVAL;
                goto done;
        }

        proc_ucred_lock(p);
        kauth_cred_ref(p->p_ucred);
        my_cred = p->p_ucred;
        if (!IS_VALID_CRED(my_cred)) {
                kauth_cred_unref(&my_cred);
                proc_ucred_unlock(p);
                err = ESRCH;
                goto done;
        }
        old_aia_p = my_cred->cr_audit.as_aia_p;
        old_asid = old_aia_p->ai_asid;
        *new_asid = new_aia_p->ai_asid;

        /*
         * Add process in if not already in the session.
         */
        if (*new_asid != old_asid) {
                kauth_cred_t my_new_cred;
                struct au_session new_as;

                bcopy(&new_aia_p->ai_mask, &new_as.as_mask,
                    sizeof(new_as.as_mask));
                new_as.as_aia_p = new_aia_p;

                my_new_cred = kauth_cred_setauditinfo(my_cred, &new_as);
                p->p_ucred = my_new_cred;
                PROC_UPDATE_CREDS_ONPROC(p);

                /* Increment the proc count of new session */
                audit_inc_procount(AU_SENTRY_PTR(new_aia_p));

                proc_ucred_unlock(p);

                /* Propagate the change from the process to the Mach task. */
                set_security_token_task_internal(p, task);

                /* Decrement the process count of the former session. */
                audit_dec_procount(AU_SENTRY_PTR(old_aia_p));
        } else {
                proc_ucred_unlock(p);
        }
        kauth_cred_unref(&my_cred);

done:
        if (port != IPC_PORT_NULL) {
                ipc_port_release_send(port);
        }

        return err;
}

/*
 * audit_session_spawnjoin
 *
 * Description: posix_spawn() interface to audit_session_join_internal().
 *
 * Returns:     0               Success
 *              EINVAL          Invalid Mach port name.
 *              ESRCH           Invalid calling process/cred.
 */
int
audit_session_spawnjoin(proc_t p, task_t task, ipc_port_t port)
{
        au_asid_t new_asid;

        return audit_session_join_internal(p, task, port, &new_asid);
}

/*
 * audit_session_join  (system call)
 *
 * Description: Join the session for a given Mach port send right.
 *
 * Parameters:  p               Process calling session join.
 *              uap->port       A Mach send right.
 *
 * Returns:     *ret_asid       Audit session ID of new session.
 *                              In the failure case the return value will be -1
 *                              and 'errno' will be set to a non-zero value
 *                              described below.
 *
 * Errno:       0               Success
 *              EINVAL          Invalid Mach port name.
 *              ESRCH           Invalid calling process/cred.
 */
int
audit_session_join(proc_t p, struct audit_session_join_args *uap,
    au_asid_t *ret_asid)
{
        ipc_port_t port = IPC_PORT_NULL;
        mach_port_name_t send = uap->port;
        int err = 0;


        if (ipc_object_copyin(get_task_ipcspace(p->task), send,
            MACH_MSG_TYPE_COPY_SEND, &port, 0, NULL, IPC_OBJECT_COPYIN_FLAGS_ALLOW_IMMOVABLE_SEND) != KERN_SUCCESS) {
                *ret_asid = AU_DEFAUDITSID;
                err = EINVAL;
        } else {
                err = audit_session_join_internal(p, p->task, port, ret_asid);
        }

        return err;
}

/*
 * Audit session device.
 */

/*
 * Free an audit sdev entry.
 */
static void
audit_sdev_entry_free(struct audit_sdev_entry *ase)
{
        free(ase->ase_record, M_AUDIT_SDEV_ENTRY);
        free(ase, M_AUDIT_SDEV_ENTRY);
}

/*
 * Append individual record to a queue.  Allocate queue-local buffer and
 * add to the queue.  If the queue is full or we can't allocate memory,
 * drop the newest record.
 */
static void
audit_sdev_append(struct audit_sdev *asdev, void *record, u_int record_len)
{
        struct audit_sdev_entry *ase;

        AUDIT_SDEV_LOCK_ASSERT(asdev);

        if (asdev->asdev_qlen >= asdev->asdev_qlimit) {
                asdev->asdev_drops++;
                audit_sdev_drops++;
                return;
        }

        ase = malloc(sizeof(*ase), M_AUDIT_SDEV_ENTRY, M_NOWAIT | M_ZERO);
        if (NULL == ase) {
                asdev->asdev_drops++;
                audit_sdev_drops++;
                return;
        }

        ase->ase_record = malloc(record_len, M_AUDIT_SDEV_ENTRY, M_NOWAIT);
        if (NULL == ase->ase_record) {
                free(ase, M_AUDIT_SDEV_ENTRY);
                asdev->asdev_drops++;
                audit_sdev_drops++;
                return;
        }

        bcopy(record, ase->ase_record, record_len);
        ase->ase_record_len = record_len;

        TAILQ_INSERT_TAIL(&asdev->asdev_queue, ase, ase_queue);
        asdev->asdev_inserts++;
        asdev->asdev_qlen++;
        asdev->asdev_qbyteslen += ase->ase_record_len;
        selwakeup(&asdev->asdev_selinfo);
        if (asdev->asdev_flags & AUDIT_SDEV_ASYNC) {
                pgsigio(asdev->asdev_sigio, SIGIO);
        }

        cv_broadcast(&asdev->asdev_cv);
}

/*
 * Submit an audit record to be queued in the audit session device.
 */
void
audit_sdev_submit(__unused au_id_t auid, __unused au_asid_t asid, void *record,
    u_int record_len)
{
        struct audit_sdev *asdev;

        /*
         * Lockless read to avoid lock overhead if sessio devices are not in
         * use.
         */
        if (NULL == TAILQ_FIRST(&audit_sdev_list)) {
                return;
        }

        AUDIT_SDEV_LIST_RLOCK();
        TAILQ_FOREACH(asdev, &audit_sdev_list, asdev_list) {
                AUDIT_SDEV_LOCK(asdev);

                /*
                 * Only append to the sdev queue if the AUID and ASID match that
                 * of the process that opened this session device or if the
                 * ALLSESSIONS flag is set.
                 */
                if ((/* XXXss auid == asdev->asdev_auid && */
                            asid == asdev->asdev_asid) ||
                    (asdev->asdev_flags & AUDIT_SDEV_ALLSESSIONS) != 0) {
                        audit_sdev_append(asdev, record, record_len);
                }
                AUDIT_SDEV_UNLOCK(asdev);
        }
        AUDIT_SDEV_LIST_RUNLOCK();

        /* Unlocked increment. */
        audit_sdev_records++;
}

/*
 * Allocate a new audit sdev.  Connects the sdev, on succes, to the global
 * list and updates statistics.
 */
static struct audit_sdev *
audit_sdev_alloc(void)
{
        struct audit_sdev *asdev;

        AUDIT_SDEV_LIST_WLOCK_ASSERT();

        asdev = malloc(sizeof(*asdev), M_AUDIT_SDEV, M_WAITOK | M_ZERO);
        if (NULL == asdev) {
                return NULL;
        }

        asdev->asdev_qlimit = AUDIT_SDEV_QLIMIT_DEFAULT;
        TAILQ_INIT(&asdev->asdev_queue);
        AUDIT_SDEV_LOCK_INIT(asdev);
        AUDIT_SDEV_SX_LOCK_INIT(asdev);
        cv_init(&asdev->asdev_cv, "audit_sdev_cv");

        /*
         * Add to global list and update global statistics.
         */
        TAILQ_INSERT_HEAD(&audit_sdev_list, asdev, asdev_list);
        audit_sdev_count++;
        audit_sdev_ever++;

        return asdev;
}

/*
 * Flush all records currently present in an audit sdev.
 */
static void
audit_sdev_flush(struct audit_sdev *asdev)
{
        struct audit_sdev_entry *ase;

        AUDIT_SDEV_LOCK_ASSERT(asdev);

        while ((ase = TAILQ_FIRST(&asdev->asdev_queue)) != NULL) {
                TAILQ_REMOVE(&asdev->asdev_queue, ase, ase_queue);
                asdev->asdev_qbyteslen -= ase->ase_record_len;
                audit_sdev_entry_free(ase);
                asdev->asdev_qlen--;
        }
        asdev->asdev_qoffset = 0;

        KASSERT(0 == asdev->asdev_qlen, ("audit_sdev_flush: asdev_qlen"));
        KASSERT(0 == asdev->asdev_qbyteslen,
            ("audit_sdev_flush: asdev_qbyteslen"));
}

/*
 * Free an audit sdev.
 */
static void
audit_sdev_free(struct audit_sdev *asdev)
{
        AUDIT_SDEV_LIST_WLOCK_ASSERT();
        AUDIT_SDEV_LOCK_ASSERT(asdev);

        /* XXXss - preselect hook here */
        audit_sdev_flush(asdev);
        cv_destroy(&asdev->asdev_cv);
        AUDIT_SDEV_SX_LOCK_DESTROY(asdev);
        AUDIT_SDEV_UNLOCK(asdev);
        AUDIT_SDEV_LOCK_DESTROY(asdev);

        TAILQ_REMOVE(&audit_sdev_list, asdev, asdev_list);
        free(asdev, M_AUDIT_SDEV);
        audit_sdev_count--;
}

/*
 * Get the auditinfo_addr of the proc and check to see if suser.  Will return
 * non-zero if not suser.
 */
static int
audit_sdev_get_aia(proc_t p, struct auditinfo_addr *aia_p)
{
        int error;
        kauth_cred_t scred;

        scred = kauth_cred_proc_ref(p);
        error = suser(scred, &p->p_acflag);

        if (NULL != aia_p) {
                bcopy(scred->cr_audit.as_aia_p, aia_p, sizeof(*aia_p));
        }
        kauth_cred_unref(&scred);

        return error;
}

/*
 * Audit session dev open method.
 */
static int
audit_sdev_open(dev_t dev, __unused int flags, __unused int devtype, proc_t p)
{
        struct audit_sdev *asdev;
        struct auditinfo_addr aia;
        int u;

        u = minor(dev);
        if (u < 0 || u >= MAX_AUDIT_SDEVS) {
                return ENXIO;
        }

        (void) audit_sdev_get_aia(p, &aia);

        AUDIT_SDEV_LIST_WLOCK();
        asdev = audit_sdev_dtab[u];
        if (NULL == asdev) {
                asdev = audit_sdev_alloc();
                if (NULL == asdev) {
                        AUDIT_SDEV_LIST_WUNLOCK();
                        return ENOMEM;
                }
                audit_sdev_dtab[u] = asdev;
        } else {
                KASSERT(asdev->asdev_open, ("audit_sdev_open: Already open"));
                AUDIT_SDEV_LIST_WUNLOCK();
                return EBUSY;
        }
        asdev->asdev_open = 1;
        asdev->asdev_auid = aia.ai_auid;
        asdev->asdev_asid = aia.ai_asid;
        asdev->asdev_flags = 0;

        AUDIT_SDEV_LIST_WUNLOCK();

        return 0;
}

/*
 * Audit session dev close method.
 */
static int
audit_sdev_close(dev_t dev, __unused int flags, __unused int devtype,
    __unused proc_t p)
{
        struct audit_sdev *asdev;
        int u;

        u = minor(dev);
        asdev = audit_sdev_dtab[u];

        KASSERT(asdev != NULL, ("audit_sdev_close: asdev == NULL"));
        KASSERT(asdev->asdev_open, ("audit_sdev_close: !asdev_open"));

        AUDIT_SDEV_LIST_WLOCK();
        AUDIT_SDEV_LOCK(asdev);
        asdev->asdev_open = 0;
        audit_sdev_free(asdev);  /* sdev lock unlocked in audit_sdev_free() */
        audit_sdev_dtab[u] = NULL;
        AUDIT_SDEV_LIST_WUNLOCK();

        return 0;
}

/*
 * Audit session dev ioctl method.
 */
static int
audit_sdev_ioctl(dev_t dev, u_long cmd, caddr_t data,
    __unused int flag, proc_t p)
{
        struct audit_sdev *asdev;
        int error;

        asdev = audit_sdev_dtab[minor(dev)];
        KASSERT(asdev != NULL, ("audit_sdev_ioctl: asdev == NULL"));

        error = 0;

        switch (cmd) {
        case FIONBIO:
                AUDIT_SDEV_LOCK(asdev);
                if (*(int *)data) {
                        asdev->asdev_flags |= AUDIT_SDEV_NBIO;
                } else {
                        asdev->asdev_flags &= ~AUDIT_SDEV_NBIO;
                }
                AUDIT_SDEV_UNLOCK(asdev);
                break;

        case FIONREAD:
                AUDIT_SDEV_LOCK(asdev);
                *(int *)data = asdev->asdev_qbyteslen - asdev->asdev_qoffset;
                AUDIT_SDEV_UNLOCK(asdev);
                break;

        case AUDITSDEV_GET_QLEN:
                *(u_int *)data = asdev->asdev_qlen;
                break;

        case AUDITSDEV_GET_QLIMIT:
                *(u_int *)data = asdev->asdev_qlimit;
                break;

        case AUDITSDEV_SET_QLIMIT:
                if (*(u_int *)data >= AUDIT_SDEV_QLIMIT_MIN ||
                    *(u_int *)data <= AUDIT_SDEV_QLIMIT_MAX) {
                        asdev->asdev_qlimit = *(u_int *)data;
                } else {
                        error = EINVAL;
                }
                break;

        case AUDITSDEV_GET_QLIMIT_MIN:
                *(u_int *)data = AUDIT_SDEV_QLIMIT_MIN;
                break;

        case AUDITSDEV_GET_QLIMIT_MAX:
                *(u_int *)data = AUDIT_SDEV_QLIMIT_MAX;
                break;

        case AUDITSDEV_FLUSH:
                if (AUDIT_SDEV_SX_XLOCK_SIG(asdev) != 0) {
                        return EINTR;
                }
                AUDIT_SDEV_LOCK(asdev);
                audit_sdev_flush(asdev);
                AUDIT_SDEV_UNLOCK(asdev);
                AUDIT_SDEV_SX_XUNLOCK(asdev);
                break;

        case AUDITSDEV_GET_MAXDATA:
                *(u_int *)data = MAXAUDITDATA;
                break;

        /* XXXss these should be 64 bit, maybe. */
        case AUDITSDEV_GET_INSERTS:
                *(u_int *)data = asdev->asdev_inserts;
                break;

        case AUDITSDEV_GET_READS:
                *(u_int *)data = asdev->asdev_reads;
                break;

        case AUDITSDEV_GET_DROPS:
                *(u_int *)data = asdev->asdev_drops;
                break;

        case AUDITSDEV_GET_ALLSESSIONS:
                error = audit_sdev_get_aia(p, NULL);
                if (error) {
                        break;
                }
                *(u_int *)data = (asdev->asdev_flags & AUDIT_SDEV_ALLSESSIONS) ?
                    1 : 0;
                break;

        case AUDITSDEV_SET_ALLSESSIONS:
                error = audit_sdev_get_aia(p, NULL);
                if (error) {
                        break;
                }

                AUDIT_SDEV_LOCK(asdev);
                if (*(int *)data) {
                        asdev->asdev_flags |= AUDIT_SDEV_ALLSESSIONS;
                } else {
                        asdev->asdev_flags &= ~AUDIT_SDEV_ALLSESSIONS;
                }
                AUDIT_SDEV_UNLOCK(asdev);
                break;

        default:
                error = ENOTTY;
        }

        return error;
}

/*
 * Audit session dev read method.
 */
static int
audit_sdev_read(dev_t dev, struct uio *uio, __unused int flag)
{
        struct audit_sdev_entry *ase;
        struct audit_sdev *asdev;
        u_int toread;
        int error;

        asdev = audit_sdev_dtab[minor(dev)];
        KASSERT(NULL != asdev, ("audit_sdev_read: asdev == NULL"));

        /*
         * We hold a sleep lock over read and flush because we rely on the
         * stability of a record in the queue during uiomove.
         */
        if (0 != AUDIT_SDEV_SX_XLOCK_SIG(asdev)) {
                return EINTR;
        }
        AUDIT_SDEV_LOCK(asdev);
        while (TAILQ_EMPTY(&asdev->asdev_queue)) {
                if (asdev->asdev_flags & AUDIT_SDEV_NBIO) {
                        AUDIT_SDEV_UNLOCK(asdev);
                        AUDIT_SDEV_SX_XUNLOCK(asdev);
                        return EAGAIN;
                }
                error = cv_wait_sig(&asdev->asdev_cv, AUDIT_SDEV_MTX(asdev));
                if (error) {
                        AUDIT_SDEV_UNLOCK(asdev);
                        AUDIT_SDEV_SX_XUNLOCK(asdev);
                        return error;
                }
        }

        /*
         * Copy as many remaining bytes from the current record to userspace
         * as we can. Keep processing records until we run out of records in
         * the queue or until the user buffer runs out of space.
         *
         * We rely on the sleep lock to maintain ase's stability here.
         */
        asdev->asdev_reads++;
        while ((ase = TAILQ_FIRST(&asdev->asdev_queue)) != NULL &&
            uio_resid(uio) > 0) {
                AUDIT_SDEV_LOCK_ASSERT(asdev);

                KASSERT(ase->ase_record_len > asdev->asdev_qoffset,
                    ("audit_sdev_read: record_len > qoffset (1)"));
                toread = MIN((int)(ase->ase_record_len - asdev->asdev_qoffset),
                    uio_resid(uio));
                AUDIT_SDEV_UNLOCK(asdev);
                error = uiomove((char *) ase->ase_record + asdev->asdev_qoffset,
                    toread, uio);
                if (error) {
                        AUDIT_SDEV_SX_XUNLOCK(asdev);
                        return error;
                }

                /*
                 * If the copy succeeded then update book-keeping, and if no
                 * bytes remain in the current record then free it.
                 */
                AUDIT_SDEV_LOCK(asdev);
                KASSERT(TAILQ_FIRST(&asdev->asdev_queue) == ase,
                    ("audit_sdev_read: queue out of sync after uiomove"));
                asdev->asdev_qoffset += toread;
                KASSERT(ase->ase_record_len >= asdev->asdev_qoffset,
                    ("audit_sdev_read: record_len >= qoffset (2)"));
                if (asdev->asdev_qoffset == ase->ase_record_len) {
                        TAILQ_REMOVE(&asdev->asdev_queue, ase, ase_queue);
                        asdev->asdev_qbyteslen -= ase->ase_record_len;
                        audit_sdev_entry_free(ase);
                        asdev->asdev_qlen--;
                        asdev->asdev_qoffset = 0;
                }
        }
        AUDIT_SDEV_UNLOCK(asdev);
        AUDIT_SDEV_SX_XUNLOCK(asdev);
        return 0;
}

/*
 * Audit session device poll method.
 */
static int
audit_sdev_poll(dev_t dev, int events, void *wql, struct proc *p)
{
        struct audit_sdev *asdev;
        int revents;

        revents = 0;
        asdev = audit_sdev_dtab[minor(dev)];
        KASSERT(NULL != asdev, ("audit_sdev_poll: asdev == NULL"));

        if (events & (POLLIN | POLLRDNORM)) {
                AUDIT_SDEV_LOCK(asdev);
                if (NULL != TAILQ_FIRST(&asdev->asdev_queue)) {
                        revents |= events & (POLLIN | POLLRDNORM);
                } else {
                        selrecord(p, &asdev->asdev_selinfo, wql);
                }
                AUDIT_SDEV_UNLOCK(asdev);
        }
        return revents;
}

/*
 * Audit sdev clone routine.  Provides a new minor number or returns -1.
 * This called with DEVFS_LOCK held.
 */
static int
audit_sdev_clone(__unused dev_t dev, int action)
{
        int i;

        if (DEVFS_CLONE_ALLOC == action) {
                for (i = 0; i < MAX_AUDIT_SDEVS; i++) {
                        if (NULL == audit_sdev_dtab[i]) {
                                return i;
                        }
                }

                /*
                 * This really should return -1 here but that seems to
                 * hang things in devfs.  We instead return 0 and let
                 * audit_sdev_open tell userland the bad news.
                 */
                return 0;
        }

        return -1;
}

static int
audit_sdev_init(void)
{
        dev_t dev;

        TAILQ_INIT(&audit_sdev_list);
        AUDIT_SDEV_LIST_LOCK_INIT();

        audit_sdev_major = cdevsw_add(-1, &audit_sdev_cdevsw);
        if (audit_sdev_major < 0) {
                return KERN_FAILURE;
        }

        dev = makedev(audit_sdev_major, 0);
        devnode = devfs_make_node_clone(dev, DEVFS_CHAR, UID_ROOT, GID_WHEEL,
            0644, audit_sdev_clone, AUDIT_SDEV_NAME, 0);

        if (NULL == devnode) {
                return KERN_FAILURE;
        }

        return KERN_SUCCESS;
}

/* XXXss
 *  static int
 *  audit_sdev_shutdown(void)
 *  {
 *
 *       devfs_remove(devnode);
 *       (void) cdevsw_remove(audit_sdev_major, &audit_sdev_cdevsw);
 *
 *       return (KERN_SUCCESS);
 *  }
 */

#else

int
audit_session_self(proc_t p, struct audit_session_self_args *uap,
    mach_port_name_t *ret_port)
{
#pragma unused(p, uap, ret_port)

        return ENOSYS;
}

int
audit_session_join(proc_t p, struct audit_session_join_args *uap,
    au_asid_t *ret_asid)
{
#pragma unused(p, uap, ret_asid)

        return ENOSYS;
}

int
audit_session_port(proc_t p, struct audit_session_port_args *uap, int *retval)
{
#pragma unused(p, uap, retval)

        return ENOSYS;
}

#endif /* CONFIG_AUDIT */