usr/src/uts/common/fs/nfs/nfs4_srv.c

/*
 * CDDL HEADER START
 *
 * The contents of this file are subject to the terms of the
 * Common Development and Distribution License (the "License").
 * You may not use this file except in compliance with the License.
 *
 * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
 * or http://www.opensolaris.org/os/licensing.
 * See the License for the specific language governing permissions
 * and limitations under the License.
 *
 * When distributing Covered Code, include this CDDL HEADER in each
 * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
 * If applicable, add the following below this CDDL HEADER, with the
 * fields enclosed by brackets "[]" replaced with your own identifying
 * information: Portions Copyright [yyyy] [name of copyright owner]
 *
 * CDDL HEADER END
 */
/*
 * Copyright 2006 Sun Microsystems, Inc.  All rights reserved.
 * Use is subject to license terms.
 */

/*
 *	Copyright (c) 1983,1984,1985,1986,1987,1988,1989  AT&T.
 *	All Rights Reserved
 */

#pragma ident	"%Z%%M%	%I%	%E% SMI"

#include <sys/param.h>
#include <sys/types.h>
#include <sys/systm.h>
#include <sys/cred.h>
#include <sys/buf.h>
#include <sys/vfs.h>
#include <sys/vnode.h>
#include <sys/uio.h>
#include <sys/errno.h>
#include <sys/sysmacros.h>
#include <sys/statvfs.h>
#include <sys/kmem.h>
#include <sys/dirent.h>
#include <sys/cmn_err.h>
#include <sys/debug.h>
#include <sys/systeminfo.h>
#include <sys/flock.h>
#include <sys/pathname.h>
#include <sys/nbmlock.h>
#include <sys/share.h>
#include <sys/atomic.h>
#include <sys/policy.h>
#include <sys/fem.h>
#include <sys/sdt.h>

#include <rpc/types.h>
#include <rpc/auth.h>
#include <rpc/rpcsec_gss.h>
#include <rpc/svc.h>

#include <nfs/nfs.h>
#include <nfs/export.h>
#include <nfs/lm.h>
#include <nfs/nfs4.h>

#include <sys/strsubr.h>
#include <sys/strsun.h>

#include <inet/common.h>
#include <inet/ip.h>
#include <inet/ip6.h>

#include <sys/tsol/label.h>
#include <sys/tsol/tndb.h>

#define	RFS4_MAXLOCK_TRIES 4	/* Try to get the lock this many times */
static int rfs4_maxlock_tries = RFS4_MAXLOCK_TRIES;
#define	RFS4_LOCK_DELAY 10	/* Milliseconds */
static clock_t rfs4_lock_delay = RFS4_LOCK_DELAY;

/* End of Tunables */

/*
 * Used to bump the stateid4.seqid value and show changes in the stateid
 */
#define	next_stateid(sp) (++(sp)->bits.chgseq)

/*
 * RFS4_MINLEN_ENTRY4: XDR-encoded size of smallest possible dirent.
 *	This is used to return NFS4ERR_TOOSMALL when clients specify
 *	maxcount that isn't large enough to hold the smallest possible
 *	XDR encoded dirent.
 *
 *	    sizeof cookie (8 bytes) +
 *	    sizeof name_len (4 bytes) +
 *	    sizeof smallest (padded) name (4 bytes) +
 *	    sizeof bitmap4_len (12 bytes) +   NOTE: we always encode len=2 bm4
 *	    sizeof attrlist4_len (4 bytes) +
 *	    sizeof next boolean (4 bytes)
 *
 * RFS4_MINLEN_RDDIR4: XDR-encoded size of READDIR op reply containing
 * the smallest possible entry4 (assumes no attrs requested).
 *	sizeof nfsstat4 (4 bytes) +
 *	sizeof verifier4 (8 bytes) +
 *	sizeof entry4list bool (4 bytes) +
 *	sizeof entry4 	(36 bytes) +
 *	sizeof eof bool  (4 bytes)
 *
 * RFS4_MINLEN_RDDIR_BUF: minimum length of buffer server will provide to
 *	VOP_READDIR.  Its value is the size of the maximum possible dirent
 *	for solaris.  The DIRENT64_RECLEN macro returns	the size of dirent
 *	required for a given name length.  MAXNAMELEN is the maximum
 *	filename length allowed in Solaris.  The first two DIRENT64_RECLEN()
 *	macros are to allow for . and .. entries -- just a minor tweak to try
 *	and guarantee that buffer we give to VOP_READDIR will be large enough
 *	to hold ., .., and the largest possible solaris dirent64.
 */
#define	RFS4_MINLEN_ENTRY4 36
#define	RFS4_MINLEN_RDDIR4 (4 + NFS4_VERIFIER_SIZE + 4 + RFS4_MINLEN_ENTRY4 + 4)
#define	RFS4_MINLEN_RDDIR_BUF \
	(DIRENT64_RECLEN(1) + DIRENT64_RECLEN(2) + DIRENT64_RECLEN(MAXNAMELEN))

/*
 * It would be better to pad to 4 bytes since that's what XDR would do,
 * but the dirents UFS gives us are already padded to 8, so just take
 * what we're given.  Dircount is only a hint anyway.  Currently the
 * solaris kernel is ASCII only, so there's no point in calling the
 * UTF8 functions.
 *
 * dirent64: named padded to provide 8 byte struct alignment
 *	d_ino(8) + d_off(8) + d_reclen(2) + d_name(namelen + null(1) + pad)
 *
 * cookie: uint64_t   +  utf8namelen: uint_t  +   utf8name padded to 8 bytes
 *
 */
#define	DIRENT64_TO_DIRCOUNT(dp) \
	(3 * BYTES_PER_XDR_UNIT + DIRENT64_NAMELEN((dp)->d_reclen))

/*
 * types of label comparison
 */
#define	EQUALITY_CHECK	0
#define	DOMINANCE_CHECK	1

time_t rfs4_start_time;			/* Initialized in rfs4_srvrinit */

static sysid_t lockt_sysid;		/* dummy sysid for all LOCKT calls */

u_longlong_t nfs4_srv_caller_id;

verifier4	Write4verf;
verifier4	Readdir4verf;

void		rfs4_init_compound_state(struct compound_state *);

static void	nullfree(caddr_t);
static void	rfs4_op_inval(nfs_argop4 *, nfs_resop4 *, struct svc_req *,
			struct compound_state *);
static void	rfs4_op_access(nfs_argop4 *, nfs_resop4 *, struct svc_req *,
			struct compound_state *);
static void	rfs4_op_close(nfs_argop4 *, nfs_resop4 *, struct svc_req *,
			struct compound_state *);
static void	rfs4_op_commit(nfs_argop4 *, nfs_resop4 *, struct svc_req *,
			struct compound_state *);
static void	rfs4_op_create(nfs_argop4 *, nfs_resop4 *, struct svc_req *,
			struct compound_state *);
static void	rfs4_op_create_free(nfs_resop4 *resop);
static void	rfs4_op_delegreturn(nfs_argop4 *, nfs_resop4 *,
				    struct svc_req *, struct compound_state *);
static void	rfs4_op_getattr(nfs_argop4 *, nfs_resop4 *, struct svc_req *,
			struct compound_state *);
static void	rfs4_op_getattr_free(nfs_resop4 *);
static void	rfs4_op_getfh(nfs_argop4 *, nfs_resop4 *, struct svc_req *,
			struct compound_state *);
static void	rfs4_op_getfh_free(nfs_resop4 *);
static void	rfs4_op_illegal(nfs_argop4 *, nfs_resop4 *, struct svc_req *,
			struct compound_state *);
static void	rfs4_op_link(nfs_argop4 *, nfs_resop4 *, struct svc_req *,
			struct compound_state *);
static void	rfs4_op_lock(nfs_argop4 *, nfs_resop4 *, struct svc_req *,
			struct compound_state *);
static void	lock_denied_free(nfs_resop4 *);
static void	rfs4_op_locku(nfs_argop4 *, nfs_resop4 *, struct svc_req *,
			struct compound_state *);
static void	rfs4_op_lockt(nfs_argop4 *, nfs_resop4 *, struct svc_req *,
			struct compound_state *);
static void	rfs4_op_lookup(nfs_argop4 *, nfs_resop4 *, struct svc_req *,
			struct compound_state *);
static void	rfs4_op_lookupp(nfs_argop4 *, nfs_resop4 *, struct svc_req *,
			struct compound_state *);
static void	rfs4_op_openattr(nfs_argop4 *argop, nfs_resop4 *resop,
				struct svc_req *req, struct compound_state *cs);
static void	rfs4_op_nverify(nfs_argop4 *, nfs_resop4 *, struct svc_req *,
			struct compound_state *);
static void	rfs4_op_open(nfs_argop4 *, nfs_resop4 *, struct svc_req *,
			struct compound_state *);
static void	rfs4_op_open_confirm(nfs_argop4 *, nfs_resop4 *,
			struct svc_req *, struct compound_state *);
static void	rfs4_op_open_downgrade(nfs_argop4 *, nfs_resop4 *,
			struct svc_req *, struct compound_state *);
static void	rfs4_op_putfh(nfs_argop4 *, nfs_resop4 *, struct svc_req *,
			struct compound_state *);
static void	rfs4_op_putpubfh(nfs_argop4 *, nfs_resop4 *, struct svc_req *,
			struct compound_state *);
static void	rfs4_op_putrootfh(nfs_argop4 *, nfs_resop4 *, struct svc_req *,
			struct compound_state *);
static void	rfs4_op_read(nfs_argop4 *, nfs_resop4 *, struct svc_req *,
			struct compound_state *);
static void	rfs4_op_read_free(nfs_resop4 *);
static void	rfs4_op_readdir_free(nfs_resop4 *resop);
static void	rfs4_op_readlink(nfs_argop4 *, nfs_resop4 *, struct svc_req *,
			struct compound_state *);
static void	rfs4_op_readlink_free(nfs_resop4 *);
static void	rfs4_op_release_lockowner(nfs_argop4 *, nfs_resop4 *,
			struct svc_req *, struct compound_state *);
static void	rfs4_op_remove(nfs_argop4 *, nfs_resop4 *, struct svc_req *,
			struct compound_state *);
static void	rfs4_op_rename(nfs_argop4 *, nfs_resop4 *, struct svc_req *,
			struct compound_state *);
static void	rfs4_op_renew(nfs_argop4 *, nfs_resop4 *, struct svc_req *,
			struct compound_state *);
static void	rfs4_op_restorefh(nfs_argop4 *, nfs_resop4 *, struct svc_req *,
			struct compound_state *);
static void	rfs4_op_savefh(nfs_argop4 *, nfs_resop4 *, struct svc_req *,
			struct compound_state *);
static void	rfs4_op_setattr(nfs_argop4 *, nfs_resop4 *, struct svc_req *,
			struct compound_state *);
static void	rfs4_op_verify(nfs_argop4 *, nfs_resop4 *, struct svc_req *,
			struct compound_state *);
static void	rfs4_op_write(nfs_argop4 *, nfs_resop4 *, struct svc_req *,
			struct compound_state *);
static void	rfs4_op_setclientid(nfs_argop4 *, nfs_resop4 *,
			struct svc_req *, struct compound_state *);
static void	rfs4_op_setclientid_confirm(nfs_argop4 *, nfs_resop4 *,
			struct svc_req *req, struct compound_state *);
static void	rfs4_op_secinfo(nfs_argop4 *, nfs_resop4 *, struct svc_req *,
			struct compound_state *);
static void	rfs4_op_secinfo_free(nfs_resop4 *);

static nfsstat4 check_open_access(uint32_t,
				struct compound_state *, struct svc_req *);
nfsstat4 rfs4_client_sysid(rfs4_client_t *, sysid_t *);
static int	vop_shrlock(vnode_t *, int, struct shrlock *, int);
static int 	rfs4_shrlock(rfs4_state_t *, int);
static int	rfs4_share(rfs4_state_t *);
void rfs4_ss_clid(rfs4_client_t *, struct svc_req *);

/*
 * translation table for attrs
 */
struct nfs4_ntov_table {
	union nfs4_attr_u *na;
	uint8_t amap[NFS4_MAXNUM_ATTRS];
	int attrcnt;
	bool_t vfsstat;
};

static void	nfs4_ntov_table_init(struct nfs4_ntov_table *ntovp);
static void	nfs4_ntov_table_free(struct nfs4_ntov_table *ntovp,
				    struct nfs4_svgetit_arg *sargp);

static nfsstat4	do_rfs4_set_attrs(bitmap4 *resp, fattr4 *fattrp,
		    struct compound_state *cs, struct nfs4_svgetit_arg *sargp,
		    struct nfs4_ntov_table *ntovp, nfs4_attr_cmd_t cmd);

fem_t	*deleg_rdops;
fem_t	*deleg_wrops;

rfs4_servinst_t	*rfs4_cur_servinst = NULL;	/* current server instance */
kmutex_t	rfs4_servinst_lock;		/* protects linked list */
int		rfs4_seen_first_compound;	/* set first time we see one */

#ifdef DEBUG
int	rfs4_servinst_debug = 0;
#endif

/*
 * NFS4 op dispatch table
 */

struct rfsv4disp {
	void	(*dis_proc)();		/* proc to call */
	void	(*dis_resfree)();	/* frees space allocated by proc */
	int	dis_flags;		/* RPC_IDEMPOTENT, etc... */
};

static struct rfsv4disp rfsv4disptab[] = {
	/*
	 * NFS VERSION 4
	 */

	/* RFS_NULL = 0 */
	{rfs4_op_illegal, nullfree, 0},

	/* UNUSED = 1 */
	{rfs4_op_illegal, nullfree, 0},

	/* UNUSED = 2 */
	{rfs4_op_illegal, nullfree, 0},

	/* OP_ACCESS = 3 */
	{rfs4_op_access, nullfree, RPC_IDEMPOTENT},

	/* OP_CLOSE = 4 */
	{rfs4_op_close, nullfree, 0},

	/* OP_COMMIT = 5 */
	{rfs4_op_commit, nullfree, RPC_IDEMPOTENT},

	/* OP_CREATE = 6 */
	{rfs4_op_create, nullfree, 0},

	/* OP_DELEGPURGE = 7 */
	{rfs4_op_inval, nullfree, 0},

	/* OP_DELEGRETURN = 8 */
	{rfs4_op_delegreturn, nullfree, 0},

	/* OP_GETATTR = 9 */
	{rfs4_op_getattr, rfs4_op_getattr_free, RPC_IDEMPOTENT},

	/* OP_GETFH = 10 */
	{rfs4_op_getfh, rfs4_op_getfh_free, RPC_ALL},

	/* OP_LINK = 11 */
	{rfs4_op_link, nullfree, 0},

	/* OP_LOCK = 12 */
	{rfs4_op_lock, lock_denied_free, 0},

	/* OP_LOCKT = 13 */
	{rfs4_op_lockt, lock_denied_free, 0},

	/* OP_LOCKU = 14 */
	{rfs4_op_locku, nullfree, 0},

	/* OP_LOOKUP = 15 */
	{rfs4_op_lookup, nullfree, (RPC_IDEMPOTENT|RPC_PUBLICFH_OK)},

	/* OP_LOOKUPP = 16 */
	{rfs4_op_lookupp, nullfree, (RPC_IDEMPOTENT|RPC_PUBLICFH_OK)},

	/* OP_NVERIFY = 17 */
	{rfs4_op_nverify, nullfree, RPC_IDEMPOTENT},

	/* OP_OPEN = 18 */
	{rfs4_op_open, rfs4_free_reply, 0},

	/* OP_OPENATTR = 19 */
	{rfs4_op_openattr, nullfree, 0},

	/* OP_OPEN_CONFIRM = 20 */
	{rfs4_op_open_confirm, nullfree, 0},

	/* OP_OPEN_DOWNGRADE = 21 */
	{rfs4_op_open_downgrade, nullfree, 0},

	/* OP_OPEN_PUTFH = 22 */
	{rfs4_op_putfh, nullfree, RPC_ALL},

	/* OP_PUTPUBFH = 23 */
	{rfs4_op_putpubfh, nullfree, RPC_ALL},

	/* OP_PUTROOTFH = 24 */
	{rfs4_op_putrootfh, nullfree, RPC_ALL},

	/* OP_READ = 25 */
	{rfs4_op_read, rfs4_op_read_free, RPC_IDEMPOTENT},

	/* OP_READDIR = 26 */
	{rfs4_op_readdir, rfs4_op_readdir_free, RPC_IDEMPOTENT},

	/* OP_READLINK = 27 */
	{rfs4_op_readlink, rfs4_op_readlink_free, RPC_IDEMPOTENT},

	/* OP_REMOVE = 28 */
	{rfs4_op_remove, nullfree, 0},

	/* OP_RENAME = 29 */
	{rfs4_op_rename, nullfree, 0},

	/* OP_RENEW = 30 */
	{rfs4_op_renew, nullfree, 0},

	/* OP_RESTOREFH = 31 */
	{rfs4_op_restorefh, nullfree, RPC_ALL},

	/* OP_SAVEFH = 32 */
	{rfs4_op_savefh, nullfree, RPC_ALL},

	/* OP_SECINFO = 33 */
	{rfs4_op_secinfo, rfs4_op_secinfo_free, 0},

	/* OP_SETATTR = 34 */
	{rfs4_op_setattr, nullfree, 0},

	/* OP_SETCLIENTID = 35 */
	{rfs4_op_setclientid, nullfree, 0},

	/* OP_SETCLIENTID_CONFIRM = 36 */
	{rfs4_op_setclientid_confirm, nullfree, 0},

	/* OP_VERIFY = 37 */
	{rfs4_op_verify, nullfree, RPC_IDEMPOTENT},

	/* OP_WRITE = 38 */
	{rfs4_op_write, nullfree, 0},

	/* OP_RELEASE_LOCKOWNER = 39 */
	{rfs4_op_release_lockowner, nullfree, 0},
};

static uint_t rfsv4disp_cnt = sizeof (rfsv4disptab) / sizeof (rfsv4disptab[0]);

#define	OP_ILLEGAL_IDX (rfsv4disp_cnt)

#ifdef DEBUG

int rfs4_fillone_debug = 0;
int rfs4_shrlock_debug = 0;
int rfs4_no_stub_access = 1;
int rfs4_rddir_debug = 0;

static char *rfs4_op_string[] = {
	"rfs4_op_null",
	"rfs4_op_1 unused",
	"rfs4_op_2 unused",
	"rfs4_op_access",
	"rfs4_op_close",
	"rfs4_op_commit",
	"rfs4_op_create",
	"rfs4_op_delegpurge",
	"rfs4_op_delegreturn",
	"rfs4_op_getattr",
	"rfs4_op_getfh",
	"rfs4_op_link",
	"rfs4_op_lock",
	"rfs4_op_lockt",
	"rfs4_op_locku",
	"rfs4_op_lookup",
	"rfs4_op_lookupp",
	"rfs4_op_nverify",
	"rfs4_op_open",
	"rfs4_op_openattr",
	"rfs4_op_open_confirm",
	"rfs4_op_open_downgrade",
	"rfs4_op_putfh",
	"rfs4_op_putpubfh",
	"rfs4_op_putrootfh",
	"rfs4_op_read",
	"rfs4_op_readdir",
	"rfs4_op_readlink",
	"rfs4_op_remove",
	"rfs4_op_rename",
	"rfs4_op_renew",
	"rfs4_op_restorefh",
	"rfs4_op_savefh",
	"rfs4_op_secinfo",
	"rfs4_op_setattr",
	"rfs4_op_setclientid",
	"rfs4_op_setclient_confirm",
	"rfs4_op_verify",
	"rfs4_op_write",
	"rfs4_op_release_lockowner",
	"rfs4_op_illegal"
};
#endif

void rfs4_ss_chkclid(rfs4_client_t *);

#ifdef	nextdp
#undef nextdp
#endif
#define	nextdp(dp)	((struct dirent64 *)((char *)(dp) + (dp)->d_reclen))

static const fs_operation_def_t nfs4_rd_deleg_tmpl[] = {
	VOPNAME_OPEN, deleg_rdopen,
	VOPNAME_WRITE, deleg_write,
	VOPNAME_SETATTR, deleg_setattr,
	VOPNAME_RWLOCK, deleg_rd_rwlock,
	VOPNAME_SPACE, deleg_space,
	VOPNAME_SETSECATTR, deleg_setsecattr,
	VOPNAME_VNEVENT, deleg_vnevent,
	NULL, NULL
};
static const fs_operation_def_t nfs4_wr_deleg_tmpl[] = {
	VOPNAME_OPEN, deleg_wropen,
	VOPNAME_READ, deleg_read,
	VOPNAME_WRITE, deleg_write,
	VOPNAME_SETATTR, deleg_setattr,
	VOPNAME_RWLOCK, deleg_wr_rwlock,
	VOPNAME_SPACE, deleg_space,
	VOPNAME_SETSECATTR, deleg_setsecattr,
	VOPNAME_VNEVENT, deleg_vnevent,
	NULL, NULL
};

int
rfs4_srvrinit(void)
{
	timespec32_t verf;
	int error;
	extern void rfs4_attr_init();
	extern krwlock_t rfs4_deleg_policy_lock;

	/*
	 * The following algorithm attempts to find a unique verifier
	 * to be used as the write verifier returned from the server
	 * to the client.  It is important that this verifier change
	 * whenever the server reboots.  Of secondary importance, it
	 * is important for the verifier to be unique between two
	 * different servers.
	 *
	 * Thus, an attempt is made to use the system hostid and the
	 * current time in seconds when the nfssrv kernel module is
	 * loaded.  It is assumed that an NFS server will not be able
	 * to boot and then to reboot in less than a second.  If the
	 * hostid has not been set, then the current high resolution
	 * time is used.  This will ensure different verifiers each
	 * time the server reboots and minimize the chances that two
	 * different servers will have the same verifier.
	 * XXX - this is broken on LP64 kernels.
	 */
	verf.tv_sec = (time_t)nfs_atoi(hw_serial);
	if (verf.tv_sec != 0) {
		verf.tv_nsec = gethrestime_sec();
	} else {
		timespec_t tverf;

		gethrestime(&tverf);
		verf.tv_sec = (time_t)tverf.tv_sec;
		verf.tv_nsec = tverf.tv_nsec;
	}

	Write4verf = *(uint64_t *)&verf;

	rfs4_attr_init();
	mutex_init(&rfs4_deleg_lock, NULL, MUTEX_DEFAULT, NULL);

	/* Used to manage create/destroy of server state */
	mutex_init(&rfs4_state_lock, NULL, MUTEX_DEFAULT, NULL);

	/* Used to manage access to server instance linked list */
	mutex_init(&rfs4_servinst_lock, NULL, MUTEX_DEFAULT, NULL);

	/* Used to manage access to rfs4_deleg_policy */
	rw_init(&rfs4_deleg_policy_lock, NULL, RW_DEFAULT, NULL);

	error = fem_create("deleg_rdops", nfs4_rd_deleg_tmpl, &deleg_rdops);
	if (error != 0) {
		rfs4_disable_delegation();
	} else {
		error = fem_create("deleg_wrops", nfs4_wr_deleg_tmpl,
				&deleg_wrops);
		if (error != 0) {
			rfs4_disable_delegation();
			fem_free(deleg_rdops);
		}
	}

	nfs4_srv_caller_id = fs_new_caller_id();

	lockt_sysid = lm_alloc_sysidt();

	return (0);
}

void
rfs4_srvrfini(void)
{
	extern krwlock_t rfs4_deleg_policy_lock;

	if (lockt_sysid != LM_NOSYSID) {
		lm_free_sysidt(lockt_sysid);
		lockt_sysid = LM_NOSYSID;
	}

	mutex_destroy(&rfs4_deleg_lock);
	mutex_destroy(&rfs4_state_lock);
	rw_destroy(&rfs4_deleg_policy_lock);

	fem_free(deleg_rdops);
	fem_free(deleg_wrops);
}

void
rfs4_init_compound_state(struct compound_state *cs)
{
	bzero(cs, sizeof (*cs));
	cs->cont = TRUE;
	cs->access = CS_ACCESS_DENIED;
	cs->deleg = FALSE;
	cs->mandlock = FALSE;
	cs->fh.nfs_fh4_val = cs->fhbuf;
}

void
rfs4_grace_start(rfs4_servinst_t *sip)
{
	time_t now = gethrestime_sec();

	NFS4_DEBUG(rfs4_servinst_debug, (CE_NOTE,
	    "rfs4_grace_start: inst %p: 0x%lx", (void *)sip, now));

	rw_enter(&sip->rwlock, RW_WRITER);
	sip->start_time = now;
	sip->grace_period = rfs4_grace_period;
	rw_exit(&sip->rwlock);
}

/*
 * returns true if the instance's grace period has never been started
 */
int
rfs4_servinst_grace_new(rfs4_servinst_t *sip)
{
	time_t start_time;

	rw_enter(&sip->rwlock, RW_READER);
	start_time = sip->start_time;
	rw_exit(&sip->rwlock);

	return (start_time == 0);
}

/*
 * Indicates if server instance is within the
 * grace period.
 */
int
rfs4_servinst_in_grace(rfs4_servinst_t *sip)
{
	time_t grace_expiry;

	rw_enter(&sip->rwlock, RW_READER);
	grace_expiry = sip->start_time + sip->grace_period;
	rw_exit(&sip->rwlock);

	return (gethrestime_sec() < grace_expiry);
}

int
rfs4_clnt_in_grace(rfs4_client_t *cp)
{
	ASSERT(rfs4_dbe_refcnt(cp->dbe) > 0);

	return (rfs4_servinst_in_grace(cp->server_instance));
}

/*
 * reset all currently active grace periods
 */
void
rfs4_grace_reset_all(void)
{
#ifdef DEBUG
	int n = 0;
#endif
	rfs4_servinst_t *sip;

	mutex_enter(&rfs4_servinst_lock);
	for (sip = rfs4_cur_servinst; sip != NULL; sip = sip->prev) {
		if (rfs4_servinst_in_grace(sip)) {
			rfs4_grace_start(sip);
#ifdef DEBUG
			n++;
#endif
		}
	}
	mutex_exit(&rfs4_servinst_lock);

	NFS4_DEBUG(rfs4_servinst_debug, (CE_NOTE,
	    "rfs4_grace_reset_all: reset %d instances", n));
}

/*
 * start any new instances' grace periods
 */
void
rfs4_grace_start_new(void)
{
#ifdef DEBUG
	int n = 0;
#endif
	rfs4_servinst_t *sip;

	mutex_enter(&rfs4_servinst_lock);
	for (sip = rfs4_cur_servinst; sip != NULL; sip = sip->prev) {
		if (rfs4_servinst_grace_new(sip))
			rfs4_grace_start(sip);
#ifdef DEBUG
		n++;
#endif
	}
	mutex_exit(&rfs4_servinst_lock);

	NFS4_DEBUG(rfs4_servinst_debug, (CE_NOTE,
	    "rfs4_grace_start_new: started %d new instances", n));
}

/*
 * Create a new server instance, and make it the currently active instance.
 * Note that starting the grace period too early will reduce the clients'
 * recovery window.
 */
void
rfs4_servinst_create(int start_grace)
{
	rfs4_servinst_t *sip;

	sip = kmem_alloc(sizeof (rfs4_servinst_t), KM_SLEEP);
	rw_init(&sip->rwlock, NULL, RW_DEFAULT, NULL);

	sip->start_time = (time_t)0;
	sip->grace_period = (time_t)0;
	sip->next = NULL;
	sip->prev = NULL;

	mutex_enter(&rfs4_servinst_lock);
	if (rfs4_cur_servinst == NULL) {
		NFS4_DEBUG(rfs4_servinst_debug, (CE_NOTE,
		    "rfs4_servinst_create: creating first instance"));
	} else {
		/* add to linked list */
		sip->prev = rfs4_cur_servinst;
		rfs4_cur_servinst->next = sip;
	}
	if (start_grace)
		rfs4_grace_start(sip);
	/* make the new instance "current" */
	rfs4_cur_servinst = sip;
	mutex_exit(&rfs4_servinst_lock);

	NFS4_DEBUG(rfs4_servinst_debug, (CE_NOTE,
	    "rfs4_servinst_create: new current instance: %p; start_grace: %d",
	    (void *)sip, start_grace));
}

/*
 * In future, we might add a rfs4_servinst_destroy(sip) but, for now, destroy
 * all instances directly.
 */
void
rfs4_servinst_destroy_all(void)
{
	rfs4_servinst_t *sip, *prev, *current;
#ifdef DEBUG
	int n = 0;
#endif

	mutex_enter(&rfs4_servinst_lock);
	ASSERT(rfs4_cur_servinst != NULL);
	current = rfs4_cur_servinst;
	rfs4_cur_servinst = NULL;
	for (sip = current; sip != NULL; sip = prev) {
		prev = sip->prev;
		rw_destroy(&sip->rwlock);
		kmem_free(sip, sizeof (rfs4_servinst_t));
#ifdef DEBUG
		n++;
#endif
	}
	mutex_exit(&rfs4_servinst_lock);

	NFS4_DEBUG(rfs4_servinst_debug, (CE_NOTE,
	    "rfs4_servinst_destroy_all: destroyed %d instances", n));
}

/*
 * Assign the current server instance to a client_t.
 * Should be called with cp->dbe held.
 */
void
rfs4_servinst_assign(rfs4_client_t *cp, rfs4_servinst_t *sip)
{
	ASSERT(rfs4_dbe_refcnt(cp->dbe) > 0);

	NFS4_DEBUG(rfs4_servinst_debug, (CE_NOTE,
	    "rfs4_servinst_assign: client: %p, old: %p, new: %p", (void *)cp,
	    (void *)cp->server_instance, (void *)sip));

	/*
	 * The lock ensures that if the current instance is in the process
	 * of changing, we will see the new one.
	 */
	mutex_enter(&rfs4_servinst_lock);
	cp->server_instance = sip;
	mutex_exit(&rfs4_servinst_lock);
}

rfs4_servinst_t *
rfs4_servinst(rfs4_client_t *cp)
{
	ASSERT(rfs4_dbe_refcnt(cp->dbe) > 0);

	return (cp->server_instance);
}

/* ARGSUSED */
static void
nullfree(caddr_t resop)
{
}

/*
 * This is a fall-through for invalid or not implemented (yet) ops
 */
/* ARGSUSED */
static void
rfs4_op_inval(nfs_argop4 *argop, nfs_resop4 *resop, struct svc_req *req,
	struct compound_state *cs)
{
	*cs->statusp = *((nfsstat4 *)&(resop)->nfs_resop4_u) = NFS4ERR_INVAL;
}

/*
 * Check if the security flavor, nfsnum, is in the flavor_list.
 */
bool_t
in_flavor_list(int nfsnum, int *flavor_list, int count)
{
	int i;

	for (i = 0; i < count; i++) {
		if (nfsnum == flavor_list[i])
			return (TRUE);
	}
	return (FALSE);
}

/*
 * Used by rfs4_op_secinfo to get the security information from the
 * export structure associated with the component.
 */
/* ARGSUSED */
static nfsstat4
do_rfs4_op_secinfo(struct compound_state *cs, char *nm, SECINFO4res *resp)
{
	int error, different_export = 0;
	vnode_t *dvp, *vp, *tvp;
	struct exportinfo *exi = NULL;
	fid_t fid;
	uint_t count, i;
	secinfo4 *resok_val;
	struct secinfo *secp;
	bool_t did_traverse;
	int dotdot, walk;

	dvp = cs->vp;
	dotdot = (nm[0] == '.' && nm[1] == '.' && nm[2] == '\0');

	/*
	 * If dotdotting, then need to check whether it's above the
	 * root of a filesystem, or above an export point.
	 */
	if (dotdot) {

		/*
		 * If dotdotting at the root of a filesystem, then
		 * need to traverse back to the mounted-on filesystem
		 * and do the dotdot lookup there.
		 */
		if (cs->vp->v_flag & VROOT) {

			/*
			 * If at the system root, then can
			 * go up no further.
			 */
			if (VN_CMP(dvp, rootdir))
				return (puterrno4(ENOENT));

			/*
			 * Traverse back to the mounted-on filesystem
			 */
			dvp = untraverse(cs->vp);

			/*
			 * Set the different_export flag so we remember
			 * to pick up a new exportinfo entry for
			 * this new filesystem.
			 */
			different_export = 1;
		} else {

			/*
			 * If dotdotting above an export point then set
			 * the different_export to get new export info.
			 */
			different_export = nfs_exported(cs->exi, cs->vp);
		}
	}

	/*
	 * Get the vnode for the component "nm".
	 */
	error = VOP_LOOKUP(dvp, nm, &vp, NULL, 0, NULL, cs->cr);
	if (error)
		return (puterrno4(error));

	/*
	 * If the vnode is in a pseudo filesystem, or if the security flavor
	 * used in the request is valid but not an explicitly shared flavor,
	 * or the access bit indicates that this is a limited access,
	 * check whether this vnode is visible.
	 */
	if (!different_export &&
	    (PSEUDO(cs->exi) || ! is_exported_sec(cs->nfsflavor, cs->exi) ||
	    cs->access & CS_ACCESS_LIMITED)) {
		if (! nfs_visible(cs->exi, vp, &different_export)) {
			VN_RELE(vp);
			return (puterrno4(ENOENT));
		}
	}

	/*
	 * If it's a mountpoint, then traverse it.
	 */
	if (vn_ismntpt(vp)) {
		tvp = vp;
		if ((error = traverse(&tvp)) != 0) {
			VN_RELE(vp);
			return (puterrno4(error));
		}
		/* remember that we had to traverse mountpoint */
		did_traverse = TRUE;
		vp = tvp;
		different_export = 1;
	} else if (vp->v_vfsp != dvp->v_vfsp) {
		/*
		 * If vp isn't a mountpoint and the vfs ptrs aren't the same,
		 * then vp is probably an LOFS object.  We don't need the
		 * realvp, we just need to know that we might have crossed
		 * a server fs boundary and need to call checkexport4.
		 * (LOFS lookup hides server fs mountpoints, and actually calls
		 * traverse)
		 */
		different_export = 1;
		did_traverse = FALSE;
	}

	/*
	 * Get the export information for it.
	 */
	if (different_export) {

		bzero(&fid, sizeof (fid));
		fid.fid_len = MAXFIDSZ;
		error = vop_fid_pseudo(vp, &fid);
		if (error) {
			VN_RELE(vp);
			return (puterrno4(error));
		}

		if (dotdot)
			exi = nfs_vptoexi(NULL, vp, cs->cr, &walk, NULL, TRUE);
		else
			exi = checkexport4(&vp->v_vfsp->vfs_fsid, &fid, vp);

		if (exi == NULL) {
			if (did_traverse == TRUE) {
				/*
				 * If this vnode is a mounted-on vnode,
				 * but the mounted-on file system is not
				 * exported, send back the secinfo for
				 * the exported node that the mounted-on
				 * vnode lives in.
				 */
				exi = cs->exi;
			} else {
				VN_RELE(vp);
				return (puterrno4(EACCES));
			}
		}
	} else {
		exi = cs->exi;
	}
	ASSERT(exi != NULL);


	/*
	 * Create the secinfo result based on the security information
	 * from the exportinfo structure (exi).
	 *
	 * Return all flavors for a pseudo node.
	 * For a real export node, return the flavor that the client
	 * has access with.
	 */
	ASSERT(RW_LOCK_HELD(&exported_lock));
	if (PSEUDO(exi)) {
		count = exi->exi_export.ex_seccnt; /* total sec count */
		resok_val = kmem_alloc(count * sizeof (secinfo4), KM_SLEEP);
		secp = exi->exi_export.ex_secinfo;

		for (i = 0; i < count; i++) {
		    resok_val[i].flavor = secp[i].s_secinfo.sc_rpcnum;
		    if (resok_val[i].flavor == RPCSEC_GSS) {
			rpcsec_gss_info *info;

			info = &resok_val[i].flavor_info;
			info->qop = secp[i].s_secinfo.sc_qop;
			info->service =
				(rpc_gss_svc_t)secp[i].s_secinfo.sc_service;

			/* get oid opaque data */
			info->oid.sec_oid4_len =
				secp[i].s_secinfo.sc_gss_mech_type->length;
			info->oid.sec_oid4_val =
				kmem_alloc(
				    secp[i].s_secinfo.sc_gss_mech_type->length,
				    KM_SLEEP);
			bcopy(secp[i].s_secinfo.sc_gss_mech_type->elements,
				info->oid.sec_oid4_val, info->oid.sec_oid4_len);
		    }
		}
		resp->SECINFO4resok_len = count;
		resp->SECINFO4resok_val = resok_val;
	} else {
		int ret_cnt = 0, k = 0;
		int *flavor_list;

		count = exi->exi_export.ex_seccnt; /* total sec count */
		secp = exi->exi_export.ex_secinfo;

		flavor_list = kmem_alloc(count * sizeof (int), KM_SLEEP);
		/* find out which flavors to return */
		for (i = 0; i < count; i ++) {
			int access, flavor, perm;

			flavor = secp[i].s_secinfo.sc_nfsnum;
			perm = secp[i].s_flags;

			access = nfsauth4_secinfo_access(exi, cs->req,
						flavor, perm);

			if (! (access & NFSAUTH_DENIED) &&
			    ! (access & NFSAUTH_WRONGSEC)) {
				flavor_list[ret_cnt] = flavor;
				ret_cnt++;
			}
		}

		/* Create the returning SECINFO value */
		resok_val = kmem_alloc(ret_cnt * sizeof (secinfo4), KM_SLEEP);

		for (i = 0; i < count; i++) {
		/* If the flavor is in the flavor list, fill in resok_val. */
		    if (in_flavor_list(secp[i].s_secinfo.sc_nfsnum,
						flavor_list, ret_cnt)) {
			resok_val[k].flavor = secp[i].s_secinfo.sc_rpcnum;
			if (resok_val[k].flavor == RPCSEC_GSS) {
			    rpcsec_gss_info *info;

			    info = &resok_val[k].flavor_info;
			    info->qop = secp[i].s_secinfo.sc_qop;
			    info->service =
				(rpc_gss_svc_t)secp[i].s_secinfo.sc_service;

			    /* get oid opaque data */
			    info->oid.sec_oid4_len =
				secp[i].s_secinfo.sc_gss_mech_type->length;
			    info->oid.sec_oid4_val =
				kmem_alloc(
				    secp[i].s_secinfo.sc_gss_mech_type->length,
				    KM_SLEEP);
			    bcopy(secp[i].s_secinfo.sc_gss_mech_type->elements,
				info->oid.sec_oid4_val, info->oid.sec_oid4_len);
			}
			k++;
		    }
		    if (k >= ret_cnt)
			break;
		}
		resp->SECINFO4resok_len = ret_cnt;
		resp->SECINFO4resok_val = resok_val;
		kmem_free(flavor_list, count * sizeof (int));
	}

	VN_RELE(vp);
	return (NFS4_OK);
}

/*
 * SECINFO (Operation 33): Obtain required security information on
 * the component name in the format of (security-mechanism-oid, qop, service)
 * triplets.
 */
/* ARGSUSED */
static void
rfs4_op_secinfo(nfs_argop4 *argop, nfs_resop4 *resop, struct svc_req *req,
	struct compound_state *cs)
{
	SECINFO4res *resp = &resop->nfs_resop4_u.opsecinfo;
	utf8string *utfnm = &argop->nfs_argop4_u.opsecinfo.name;
	uint_t len;
	char *nm;

	/*
	 * Current file handle (cfh) should have been set before getting
	 * into this function. If not, return error.
	 */
	if (cs->vp == NULL) {
		*cs->statusp = resp->status = NFS4ERR_NOFILEHANDLE;
		return;
	}

	if (cs->vp->v_type != VDIR) {
		*cs->statusp = resp->status = NFS4ERR_NOTDIR;
		return;
	}

	/*
	 * Verify the component name. If failed, error out, but
	 * do not error out if the component name is a "..".
	 * SECINFO will return its parents secinfo data for SECINFO "..".
	 */
	if (!utf8_dir_verify(utfnm)) {
		if (utfnm->utf8string_len != 2 ||
				utfnm->utf8string_val[0] != '.' ||
				utfnm->utf8string_val[1] != '.') {
			*cs->statusp = resp->status = NFS4ERR_INVAL;
			return;
		}
	}

	nm = utf8_to_str(utfnm, &len, NULL);
	if (nm == NULL) {
		*cs->statusp = resp->status = NFS4ERR_INVAL;
		return;
	}

	if (len > MAXNAMELEN) {
		*cs->statusp = resp->status = NFS4ERR_NAMETOOLONG;
		kmem_free(nm, len);
		return;
	}

	*cs->statusp = resp->status = do_rfs4_op_secinfo(cs, nm, resp);

	kmem_free(nm, len);
}

/*
 * Free SECINFO result.
 */
/* ARGSUSED */
static void
rfs4_op_secinfo_free(nfs_resop4 *resop)
{
	SECINFO4res *resp = &resop->nfs_resop4_u.opsecinfo;
	int count, i;
	secinfo4 *resok_val;

	/* If this is not an Ok result, nothing to free. */
	if (resp->status != NFS4_OK) {
		return;
	}

	count = resp->SECINFO4resok_len;
	resok_val = resp->SECINFO4resok_val;

	for (i = 0; i < count; i++) {
	    if (resok_val[i].flavor == RPCSEC_GSS) {
		rpcsec_gss_info *info;

		info = &resok_val[i].flavor_info;
		kmem_free(info->oid.sec_oid4_val, info->oid.sec_oid4_len);
	    }
	}
	kmem_free(resok_val, count * sizeof (secinfo4));
	resp->SECINFO4resok_len = 0;
	resp->SECINFO4resok_val = NULL;
}

/*
 * do label check on client label and server's file lable.
 */
static boolean_t
do_rfs4_label_check(bslabel_t *clabel, vnode_t *vp, int flag)
{
	bslabel_t *slabel;
	ts_label_t *tslabel;
	boolean_t result;

	if ((tslabel = nfs4_getflabel(vp)) == NULL) {
		return (B_FALSE);
	}
	slabel = label2bslabel(tslabel);
	DTRACE_PROBE4(tx__rfs4__log__info__labelcheck, char *,
	    "comparing server's file label(1) with client label(2) (vp(3))",
	    bslabel_t *, slabel, bslabel_t *, clabel, vnode_t *, vp);

	if (flag == EQUALITY_CHECK)
		result = blequal(clabel, slabel);
	else
		result = bldominates(clabel, slabel);
	label_rele(tslabel);
	return (result);
}

/* ARGSUSED */
static void
rfs4_op_access(nfs_argop4 *argop, nfs_resop4 *resop, struct svc_req *req,
	struct compound_state *cs)
{
	ACCESS4args *args = &argop->nfs_argop4_u.opaccess;
	ACCESS4res *resp = &resop->nfs_resop4_u.opaccess;
	int error;
	vnode_t *vp;
	struct vattr va;
	int checkwriteperm;
	cred_t *cr = cs->cr;
	bslabel_t *clabel, *slabel;
	ts_label_t *tslabel;
	boolean_t admin_low_client;

#if 0	/* XXX allow access even if !cs->access. Eventually only pseudo fs */
	if (cs->access == CS_ACCESS_DENIED) {
		*cs->statusp = resp->status = NFS4ERR_ACCESS;
		return;
	}
#endif
	if (cs->vp == NULL) {
		*cs->statusp = resp->status = NFS4ERR_NOFILEHANDLE;
		return;
	}

	ASSERT(cr != NULL);

	vp = cs->vp;

	/*
	 * If the file system is exported read only, it is not appropriate
	 * to check write permissions for regular files and directories.
	 * Special files are interpreted by the client, so the underlying
	 * permissions are sent back to the client for interpretation.
	 */
	if (rdonly4(cs->exi, cs->vp, req) &&
		(vp->v_type == VREG || vp->v_type == VDIR))
		checkwriteperm = 0;
	else
		checkwriteperm = 1;

	/*
	 * XXX
	 * We need the mode so that we can correctly determine access
	 * permissions relative to a mandatory lock file.  Access to
	 * mandatory lock files is denied on the server, so it might
	 * as well be reflected to the server during the open.
	 */
	va.va_mask = AT_MODE;
	error = VOP_GETATTR(vp, &va, 0, cr);
	if (error) {
		*cs->statusp = resp->status = puterrno4(error);
		return;
	}
	resp->access = 0;
	resp->supported = 0;

	if (is_system_labeled()) {
		ASSERT(req->rq_label != NULL);
		clabel = req->rq_label;
		DTRACE_PROBE2(tx__rfs4__log__info__opaccess__clabel, char *,
		    "got client label from request(1)",
		    struct svc_req *, req);
		if (!blequal(&l_admin_low->tsl_label, clabel)) {
			if ((tslabel = nfs4_getflabel(vp)) == NULL) {
				*cs->statusp = resp->status = puterrno4(EACCES);
				return;
			}
			slabel = label2bslabel(tslabel);
			DTRACE_PROBE3(tx__rfs4__log__info__opaccess__slabel,
			    char *, "got server label(1) for vp(2)",
			    bslabel_t *, slabel, vnode_t *, vp);

			admin_low_client = B_FALSE;
		} else
			admin_low_client = B_TRUE;
	}

	if (args->access & ACCESS4_READ) {
		error = VOP_ACCESS(vp, VREAD, 0, cr);
		if (!error && !MANDLOCK(vp, va.va_mode) &&
		    (!is_system_labeled() || admin_low_client ||
		    bldominates(clabel, slabel)))
			resp->access |= ACCESS4_READ;
		resp->supported |= ACCESS4_READ;
	}
	if ((args->access & ACCESS4_LOOKUP) && vp->v_type == VDIR) {
		error = VOP_ACCESS(vp, VEXEC, 0, cr);
		if (!error && (!is_system_labeled() || admin_low_client ||
		    bldominates(clabel, slabel)))
			resp->access |= ACCESS4_LOOKUP;
		resp->supported |= ACCESS4_LOOKUP;
	}
	if (checkwriteperm &&
	    (args->access & (ACCESS4_MODIFY|ACCESS4_EXTEND))) {
		error = VOP_ACCESS(vp, VWRITE, 0, cr);
		if (!error && !MANDLOCK(vp, va.va_mode) &&
		    (!is_system_labeled() || admin_low_client ||
		    blequal(clabel, slabel)))
			resp->access |=
			    (args->access & (ACCESS4_MODIFY|ACCESS4_EXTEND));
		resp->supported |= (ACCESS4_MODIFY|ACCESS4_EXTEND);
	}

	if (checkwriteperm &&
	    (args->access & ACCESS4_DELETE) && vp->v_type == VDIR) {
		error = VOP_ACCESS(vp, VWRITE, 0, cr);
		if (!error && (!is_system_labeled() || admin_low_client ||
		    blequal(clabel, slabel)))
			resp->access |= ACCESS4_DELETE;
		resp->supported |= ACCESS4_DELETE;
	}
	if (args->access & ACCESS4_EXECUTE && vp->v_type != VDIR) {
		error = VOP_ACCESS(vp, VEXEC, 0, cr);
		if (!error && !MANDLOCK(vp, va.va_mode) &&
		    (!is_system_labeled() || admin_low_client ||
		    bldominates(clabel, slabel)))
			resp->access |= ACCESS4_EXECUTE;
		resp->supported |= ACCESS4_EXECUTE;
	}

	if (is_system_labeled() && !admin_low_client)
		label_rele(tslabel);

	*cs->statusp = resp->status = NFS4_OK;
}

/* ARGSUSED */
static void
rfs4_op_commit(nfs_argop4 *argop, nfs_resop4 *resop, struct svc_req *req,
	struct compound_state *cs)
{
	COMMIT4args *args = &argop->nfs_argop4_u.opcommit;
	COMMIT4res *resp = &resop->nfs_resop4_u.opcommit;
	int error;
	vnode_t *vp = cs->vp;
	cred_t *cr = cs->cr;
	vattr_t va;

	if (vp == NULL) {
		*cs->statusp = resp->status = NFS4ERR_NOFILEHANDLE;
		return;
	}
	if (cs->access == CS_ACCESS_DENIED) {
		*cs->statusp = resp->status = NFS4ERR_ACCESS;
		return;
	}

	if (args->offset + args->count < args->offset) {
		*cs->statusp = resp->status = NFS4ERR_INVAL;
		return;
	}

	va.va_mask = AT_UID;
	error = VOP_GETATTR(vp, &va, 0, cr);

	/*
	 * If we can't get the attributes, then we can't do the
	 * right access checking.  So, we'll fail the request.
	 */
	if (error) {
		*cs->statusp = resp->status = puterrno4(error);
		return;
	}
	if (rdonly4(cs->exi, cs->vp, req)) {
		*cs->statusp = resp->status = NFS4ERR_ROFS;
		return;
	}

	if (vp->v_type != VREG) {
		if (vp->v_type == VDIR)
			resp->status = NFS4ERR_ISDIR;
		else
			resp->status = NFS4ERR_INVAL;
		*cs->statusp = resp->status;
		return;
	}

	if (crgetuid(cr) != va.va_uid &&
	    (error = VOP_ACCESS(vp, VWRITE, 0, cs->cr))) {
		*cs->statusp = resp->status = puterrno4(error);
		return;
	}

	error = VOP_PUTPAGE(vp, args->offset, args->count, 0, cr);
	if (!error)
		error = VOP_FSYNC(vp, FNODSYNC, cr);

	if (error) {
		*cs->statusp = resp->status = puterrno4(error);
		return;
	}

	*cs->statusp = resp->status = NFS4_OK;
	resp->writeverf = Write4verf;
}

/*
 * rfs4_op_mknod is called from rfs4_op_create after all initial verification
 * was completed. It does the nfsv4 create for special files.
 */
/* ARGSUSED */
static vnode_t *
do_rfs4_op_mknod(CREATE4args *args, CREATE4res *resp, struct svc_req *req,
	struct compound_state *cs, vattr_t *vap, char *nm)
{
	int error;
	cred_t *cr = cs->cr;
	vnode_t *dvp = cs->vp;
	vnode_t *vp = NULL;
	int mode;
	enum vcexcl excl;

	switch (args->type) {
	case NF4CHR:
	case NF4BLK:
		if (secpolicy_sys_devices(cr) != 0) {
			*cs->statusp = resp->status = NFS4ERR_PERM;
			return (NULL);
		}
		if (args->type == NF4CHR)
			vap->va_type = VCHR;
		else
			vap->va_type = VBLK;
		vap->va_rdev = makedevice(args->ftype4_u.devdata.specdata1,
					args->ftype4_u.devdata.specdata2);
		vap->va_mask |= AT_RDEV;
		break;
	case NF4SOCK:
		vap->va_type = VSOCK;
		break;
	case NF4FIFO:
		vap->va_type = VFIFO;
		break;
	default:
		*cs->statusp = resp->status = NFS4ERR_BADTYPE;
		return (NULL);
	}

	/*
	 * Must specify the mode.
	 */
	if (!(vap->va_mask & AT_MODE)) {
		*cs->statusp = resp->status = NFS4ERR_INVAL;
		return (NULL);
	}

	excl = EXCL;

	mode = 0;

	error = VOP_CREATE(dvp, nm, vap, excl, mode, &vp, cr, 0);
	if (error) {
		*cs->statusp = resp->status = puterrno4(error);
		return (NULL);
	}
	return (vp);
}

/*
 * nfsv4 create is used to create non-regular files. For regular files,
 * use nfsv4 open.
 */
/* ARGSUSED */
static void
rfs4_op_create(nfs_argop4 *argop, nfs_resop4 *resop, struct svc_req *req,
	struct compound_state *cs)
{
	CREATE4args *args = &argop->nfs_argop4_u.opcreate;
	CREATE4res *resp = &resop->nfs_resop4_u.opcreate;
	int error;
	struct vattr bva, iva, iva2, ava, *vap;
	cred_t *cr = cs->cr;
	vnode_t *dvp = cs->vp;
	vnode_t *vp = NULL;
	char *nm, *lnm;
	uint_t len, llen;
	int syncval = 0;
	struct nfs4_svgetit_arg sarg;
	struct nfs4_ntov_table ntov;
	struct statvfs64 sb;
	nfsstat4 status;

	resp->attrset = 0;

	if (dvp == NULL) {
		*cs->statusp = resp->status = NFS4ERR_NOFILEHANDLE;
		return;
	}

	/*
	 * If there is an unshared filesystem mounted on this vnode,
	 * do not allow to create an object in this directory.
	 */
	if (vn_ismntpt(dvp)) {
		*cs->statusp = resp->status = NFS4ERR_ACCESS;
		return;
	}

	/* Verify that type is correct */
	switch (args->type) {
	case NF4LNK:
	case NF4BLK:
	case NF4CHR:
	case NF4SOCK:
	case NF4FIFO:
	case NF4DIR:
		break;
	default:
		*cs->statusp = resp->status = NFS4ERR_BADTYPE;
		return;
	};

	if (cs->access == CS_ACCESS_DENIED) {
		*cs->statusp = resp->status = NFS4ERR_ACCESS;
		return;
	}
	if (dvp->v_type != VDIR) {
		*cs->statusp = resp->status = NFS4ERR_NOTDIR;
		return;
	}
	if (!utf8_dir_verify(&args->objname)) {
		*cs->statusp = resp->status = NFS4ERR_INVAL;
		return;
	}

	if (rdonly4(cs->exi, cs->vp, req)) {
		*cs->statusp = resp->status = NFS4ERR_ROFS;
		return;
	}

	/*
	 * Name of newly created object
	 */
	nm = utf8_to_fn(&args->objname, &len, NULL);
	if (nm == NULL) {
		*cs->statusp = resp->status = NFS4ERR_INVAL;
		return;
	}

	if (len > MAXNAMELEN) {
		*cs->statusp = resp->status = NFS4ERR_NAMETOOLONG;
		kmem_free(nm, len);
		return;
	}

	resp->attrset = 0;

	sarg.sbp = &sb;
	nfs4_ntov_table_init(&ntov);

	status = do_rfs4_set_attrs(&resp->attrset,
					&args->createattrs, cs, &sarg,
					&ntov, NFS4ATTR_SETIT);

	if (sarg.vap->va_mask == 0 && status == NFS4_OK)
		status = NFS4ERR_INVAL;

	if (status != NFS4_OK) {
		*cs->statusp = resp->status = status;
		kmem_free(nm, len);
		nfs4_ntov_table_free(&ntov, &sarg);
		resp->attrset = 0;
		return;
	}

	/* Get "before" change value */
	bva.va_mask = AT_CTIME|AT_SEQ;
	error = VOP_GETATTR(dvp, &bva, 0, cr);
	if (error) {
		*cs->statusp = resp->status = puterrno4(error);
		kmem_free(nm, len);
		nfs4_ntov_table_free(&ntov, &sarg);
		resp->attrset = 0;
		return;
	}
	NFS4_SET_FATTR4_CHANGE(resp->cinfo.before, bva.va_ctime)

	vap = sarg.vap;

	/*
	 * Set default initial values for attributes when not specified
	 * in createattrs.
	 */
	if ((vap->va_mask & AT_UID) == 0) {
		vap->va_uid = crgetuid(cr);
		vap->va_mask |= AT_UID;
	}
	if ((vap->va_mask & AT_GID) == 0) {
		vap->va_gid = crgetgid(cr);
		vap->va_mask |= AT_GID;
	}

	vap->va_mask |= AT_TYPE;
	switch (args->type) {
	case NF4DIR:
		vap->va_type = VDIR;
		if ((vap->va_mask & AT_MODE) == 0) {
			vap->va_mode = 0700;	/* default: owner rwx only */
			vap->va_mask |= AT_MODE;
		}
		error = VOP_MKDIR(dvp, nm, vap, &vp, cr);
		if (error)
			break;

		/*
		 * Get the initial "after" sequence number, if it fails,
		 * set to zero
		 */
		iva.va_mask = AT_SEQ;
		if (VOP_GETATTR(dvp, &iva, 0, cs->cr))
			iva.va_seq = 0;
		break;
	case NF4LNK:
		vap->va_type = VLNK;
		if ((vap->va_mask & AT_MODE) == 0) {
			vap->va_mode = 0700;	/* default: owner rwx only */
			vap->va_mask |= AT_MODE;
		}

		/*
		 * symlink names must be treated as data
		 */
		lnm = utf8_to_str(&args->ftype4_u.linkdata, &llen, NULL);

		if (lnm == NULL) {
			*cs->statusp = resp->status = NFS4ERR_INVAL;
			kmem_free(nm, len);
			nfs4_ntov_table_free(&ntov, &sarg);
			resp->attrset = 0;
			return;
		}

		if (llen > MAXPATHLEN) {
			*cs->statusp = resp->status = NFS4ERR_NAMETOOLONG;
			kmem_free(nm, len);
			kmem_free(lnm, llen);
			nfs4_ntov_table_free(&ntov, &sarg);
			resp->attrset = 0;
			return;
		}

		error = VOP_SYMLINK(dvp, nm, vap, lnm, cr);
		if (lnm != NULL)
			kmem_free(lnm, llen);
		if (error)
			break;

		/*
		 * Get the initial "after" sequence number, if it fails,
		 * set to zero
		 */
		iva.va_mask = AT_SEQ;
		if (VOP_GETATTR(dvp, &iva, 0, cs->cr))
			iva.va_seq = 0;

		error = VOP_LOOKUP(dvp, nm, &vp, NULL, 0, NULL, cr);
		if (error)
			break;

		/*
		 * va_seq is not safe over VOP calls, check it again
		 * if it has changed zero out iva to force atomic = FALSE.
		 */
		iva2.va_mask = AT_SEQ;
		if (VOP_GETATTR(dvp, &iva2, 0, cs->cr) ||
						iva2.va_seq != iva.va_seq)
			iva.va_seq = 0;
		break;
	default:
		/*
		 * probably a special file.
		 */
		if ((vap->va_mask & AT_MODE) == 0) {
			vap->va_mode = 0600;	/* default: owner rw only */
			vap->va_mask |= AT_MODE;
		}
		syncval = FNODSYNC;
		/*
		 * We know this will only generate one VOP call
		 */
		vp = do_rfs4_op_mknod(args, resp, req, cs, vap, nm);

		if (vp == NULL) {
			kmem_free(nm, len);
			nfs4_ntov_table_free(&ntov, &sarg);
			resp->attrset = 0;
			return;
		}

		/*
		 * Get the initial "after" sequence number, if it fails,
		 * set to zero
		 */
		iva.va_mask = AT_SEQ;
		if (VOP_GETATTR(dvp, &iva, 0, cs->cr))
			iva.va_seq = 0;

		break;
	}
	kmem_free(nm, len);

	if (error) {
		*cs->statusp = resp->status = puterrno4(error);
	}

	/*
	 * Force modified data and metadata out to stable storage.
	 */
	(void) VOP_FSYNC(dvp, 0, cr);

	if (resp->status != NFS4_OK) {
		if (vp != NULL)
			VN_RELE(vp);
		nfs4_ntov_table_free(&ntov, &sarg);
		resp->attrset = 0;
		return;
	}

	/*
	 * Finish setup of cinfo response, "before" value already set.
	 * Get "after" change value, if it fails, simply return the
	 * before value.
	 */
	ava.va_mask = AT_CTIME|AT_SEQ;
	if (VOP_GETATTR(dvp, &ava, 0, cr)) {
		ava.va_ctime = bva.va_ctime;
		ava.va_seq = 0;
	}
	NFS4_SET_FATTR4_CHANGE(resp->cinfo.after, ava.va_ctime);

	/*
	 * True verification that object was created with correct
	 * attrs is impossible.  The attrs could have been changed
	 * immediately after object creation.  If attributes did
	 * not verify, the only recourse for the server is to
	 * destroy the object.  Maybe if some attrs (like gid)
	 * are set incorrectly, the object should be destroyed;
	 * however, seems bad as a default policy.  Do we really
	 * want to destroy an object over one of the times not
	 * verifying correctly?  For these reasons, the server
	 * currently sets bits in attrset for createattrs
	 * that were set; however, no verification is done.
	 *
	 * vmask_to_nmask accounts for vattr bits set on create
	 *	[do_rfs4_set_attrs() only sets resp bits for
	 *	 non-vattr/vfs bits.]
	 * Mask off any bits set by default so as not to return
	 * more attrset bits than were requested in createattrs
	 */
	nfs4_vmask_to_nmask(sarg.vap->va_mask, &resp->attrset);
	resp->attrset &= args->createattrs.attrmask;
	nfs4_ntov_table_free(&ntov, &sarg);

	error = makefh4(&cs->fh, vp, cs->exi);
	if (error) {
		*cs->statusp = resp->status = puterrno4(error);
	}

	/*
	 * The cinfo.atomic = TRUE only if we got no errors, we have
	 * non-zero va_seq's, and it has incremented by exactly one
	 * during the creation and it didn't change during the VOP_LOOKUP
	 * or VOP_FSYNC.
	 */
	if (!error && bva.va_seq && iva.va_seq && ava.va_seq &&
			iva.va_seq == (bva.va_seq + 1) &&
			iva.va_seq == ava.va_seq)
		resp->cinfo.atomic = TRUE;
	else
		resp->cinfo.atomic = FALSE;

	(void) VOP_FSYNC(vp, syncval, cr);

	if (resp->status != NFS4_OK) {
		VN_RELE(vp);
		return;
	}
	if (cs->vp)
		VN_RELE(cs->vp);

	cs->vp = vp;
	*cs->statusp = resp->status = NFS4_OK;
}


/*ARGSUSED*/
static void
rfs4_op_delegreturn(nfs_argop4 *argop, nfs_resop4 *resop, struct svc_req *req,
	struct compound_state *cs)
{
	DELEGRETURN4args *args = &argop->nfs_argop4_u.opdelegreturn;
	DELEGRETURN4res *resp = &resop->nfs_resop4_u.opdelegreturn;
	rfs4_deleg_state_t *dsp;
	nfsstat4 status;

	status = rfs4_get_deleg_state(&args->deleg_stateid, &dsp);
	resp->status = *cs->statusp = status;
	if (status != NFS4_OK)
		return;

	/* Ensure specified filehandle matches */
	if (cs->vp != dsp->finfo->vp) {
		resp->status = *cs->statusp = NFS4ERR_BAD_STATEID;
	} else
		rfs4_return_deleg(dsp, FALSE);

	rfs4_update_lease(dsp->client);

	rfs4_deleg_state_rele(dsp);
}

/*
 * Check to see if a given "flavor" is an explicitly shared flavor.
 * The assumption of this routine is the "flavor" is already a valid
 * flavor in the secinfo list of "exi".
 *
 *	e.g.
 *		# share -o sec=flavor1 /export
 *		# share -o sec=flavor2 /export/home
 *
 *		flavor2 is not an explicitly shared flavor for /export,
 *		however it is in the secinfo list for /export thru the
 *		server namespace setup.
 */
int
is_exported_sec(int flavor, struct exportinfo *exi)
{
	int	i;
	struct secinfo *sp;

	sp = exi->exi_export.ex_secinfo;
	for (i = 0; i < exi->exi_export.ex_seccnt; i++) {
		if (flavor == sp[i].s_secinfo.sc_nfsnum ||
		    sp[i].s_secinfo.sc_nfsnum == AUTH_NONE) {
			return (SEC_REF_EXPORTED(&sp[i]));
		}
	}

	/* Should not reach this point based on the assumption */
	return (0);
}

/*
 * Check if the security flavor used in the request matches what is
 * required at the export point or at the root pseudo node (exi_root).
 *
 * returns 1 if there's a match or if exported with AUTH_NONE; 0 otherwise.
 *
 */
static int
secinfo_match_or_authnone(struct compound_state *cs)
{
	int	i;
	struct secinfo *sp;

	/*
	 * Check cs->nfsflavor (from the request) against
	 * the current export data in cs->exi.
	 */
	sp = cs->exi->exi_export.ex_secinfo;
	for (i = 0; i < cs->exi->exi_export.ex_seccnt; i++) {
		if (cs->nfsflavor == sp[i].s_secinfo.sc_nfsnum ||
		    sp[i].s_secinfo.sc_nfsnum == AUTH_NONE)
			return (1);
	}

	return (0);
}

/*
 * Check the access authority for the client and return the correct error.
 */
nfsstat4
call_checkauth4(struct compound_state *cs, struct svc_req *req)
{
	int	authres;

	/*
	 * First, check if the security flavor used in the request
	 * are among the flavors set in the server namespace.
	 */
	if (!secinfo_match_or_authnone(cs)) {
		*cs->statusp = NFS4ERR_WRONGSEC;
		return (*cs->statusp);
	}

	authres = checkauth4(cs, req);

	if (authres > 0) {
		*cs->statusp = NFS4_OK;
		if (! (cs->access & CS_ACCESS_LIMITED))
			cs->access = CS_ACCESS_OK;
	} else if (authres == 0) {
		*cs->statusp = NFS4ERR_ACCESS;
	} else if (authres == -2) {
		*cs->statusp = NFS4ERR_WRONGSEC;
	} else {
		*cs->statusp = NFS4ERR_DELAY;
	}
	return (*cs->statusp);
}

/*
 * bitmap4_to_attrmask is called by getattr and readdir.
 * It sets up the vattr mask and determines whether vfsstat call is needed
 * based on the input bitmap.
 * Returns nfsv4 status.
 */
static nfsstat4
bitmap4_to_attrmask(bitmap4 breq, struct nfs4_svgetit_arg *sargp)
{
	int i;
	uint_t	va_mask;
	struct statvfs64 *sbp = sargp->sbp;

	sargp->sbp = NULL;
	sargp->flag = 0;
	sargp->rdattr_error = NFS4_OK;
	sargp->mntdfid_set = FALSE;
	if (sargp->cs->vp)
		sargp->xattr = get_fh4_flag(&sargp->cs->fh,
					    FH4_ATTRDIR | FH4_NAMEDATTR);
	else
		sargp->xattr = 0;

	/*
	 * Set rdattr_error_req to true if return error per
	 * failed entry rather than fail the readdir.
	 */
	if (breq & FATTR4_RDATTR_ERROR_MASK)
		sargp->rdattr_error_req = 1;
	else
		sargp->rdattr_error_req = 0;

	/*
	 * generate the va_mask
	 * Handle the easy cases first
	 */
	switch (breq) {
	case NFS4_NTOV_ATTR_MASK:
		sargp->vap->va_mask = NFS4_NTOV_ATTR_AT_MASK;
		return (NFS4_OK);

	case NFS4_FS_ATTR_MASK:
		sargp->vap->va_mask = NFS4_FS_ATTR_AT_MASK;
		sargp->sbp = sbp;
		return (NFS4_OK);

	case NFS4_NTOV_ATTR_CACHE_MASK:
		sargp->vap->va_mask = NFS4_NTOV_ATTR_CACHE_AT_MASK;
		return (NFS4_OK);

	case FATTR4_LEASE_TIME_MASK:
		sargp->vap->va_mask = 0;
		return (NFS4_OK);

	default:
		va_mask = 0;
		for (i = 0; i < nfs4_ntov_map_size; i++) {
			if ((breq & nfs4_ntov_map[i].fbit) &&
							nfs4_ntov_map[i].vbit)
				va_mask |= nfs4_ntov_map[i].vbit;
		}

		/*
		 * Check is vfsstat is needed
		 */
		if (breq & NFS4_FS_ATTR_MASK)
			sargp->sbp = sbp;

		sargp->vap->va_mask = va_mask;
		return (NFS4_OK);
	}
	/* NOTREACHED */
}

/*
 * bitmap4_get_sysattrs is called by getattr and readdir.
 * It calls both VOP_GETATTR and VFS_STATVFS calls to get the attrs.
 * Returns nfsv4 status.
 */
static nfsstat4
bitmap4_get_sysattrs(struct nfs4_svgetit_arg *sargp)
{
	int error;
	struct compound_state *cs = sargp->cs;
	vnode_t *vp = cs->vp;

	if (sargp->sbp != NULL) {
		if (error = VFS_STATVFS(vp->v_vfsp, sargp->sbp)) {
			sargp->sbp = NULL;	/* to identify error */
			return (puterrno4(error));
		}
	}

	return (rfs4_vop_getattr(vp, sargp->vap, 0, cs->cr));
}

static void
nfs4_ntov_table_init(struct nfs4_ntov_table *ntovp)
{
	ntovp->na = kmem_zalloc(sizeof (union nfs4_attr_u) * nfs4_ntov_map_size,
			KM_SLEEP);
	ntovp->attrcnt = 0;
	ntovp->vfsstat = FALSE;
}

static void
nfs4_ntov_table_free(struct nfs4_ntov_table *ntovp,
	struct nfs4_svgetit_arg *sargp)
{
	int i;
	union nfs4_attr_u *na;
	uint8_t *amap;

	/*
	 * XXX Should do the same checks for whether the bit is set
	 */
	for (i = 0, na = ntovp->na, amap = ntovp->amap;
		i < ntovp->attrcnt; i++, na++, amap++) {
		(void) (*nfs4_ntov_map[*amap].sv_getit)(
			NFS4ATTR_FREEIT, sargp, na);
	}
	if ((sargp->op == NFS4ATTR_SETIT) || (sargp->op == NFS4ATTR_VERIT)) {
		/*
		 * xdr_free for getattr will be done later
		 */
		for (i = 0, na = ntovp->na, amap = ntovp->amap;
			i < ntovp->attrcnt; i++, na++, amap++) {
			xdr_free(nfs4_ntov_map[*amap].xfunc, (caddr_t)na);
		}
	}
	kmem_free(ntovp->na, sizeof (union nfs4_attr_u) * nfs4_ntov_map_size);
}

/*
 * do_rfs4_op_getattr gets the system attrs and converts into fattr4.
 */
static nfsstat4
do_rfs4_op_getattr(bitmap4 breq, fattr4 *fattrp,
	struct nfs4_svgetit_arg *sargp)
{
	int error = 0;
	int i, k;
	struct nfs4_ntov_table ntov;
	XDR xdr;
	ulong_t xdr_size;
	char *xdr_attrs;
	nfsstat4 status = NFS4_OK;
	nfsstat4 prev_rdattr_error = sargp->rdattr_error;
	union nfs4_attr_u *na;
	uint8_t *amap;

	sargp->op = NFS4ATTR_GETIT;
	sargp->flag = 0;

	fattrp->attrmask = 0;
	/* if no bits requested, then return empty fattr4 */
	if (breq == 0) {
		fattrp->attrlist4_len = 0;
		fattrp->attrlist4 = NULL;
		return (NFS4_OK);
	}

	/*
	 * return NFS4ERR_INVAL when client requests write-only attrs
	 */
	if (breq & (FATTR4_TIME_ACCESS_SET_MASK | FATTR4_TIME_MODIFY_SET_MASK))
		return (NFS4ERR_INVAL);

	nfs4_ntov_table_init(&ntov);
	na = ntov.na;
	amap = ntov.amap;

	/*
	 * Now loop to get or verify the attrs
	 */
	for (i = 0; i < nfs4_ntov_map_size; i++) {
		if (breq & nfs4_ntov_map[i].fbit) {
			if ((*nfs4_ntov_map[i].sv_getit)(
				    NFS4ATTR_SUPPORTED, sargp, NULL) == 0) {

				error = (*nfs4_ntov_map[i].sv_getit)(
						NFS4ATTR_GETIT, sargp, na);

				/*
				 * Possible error values:
				 * >0 if sv_getit failed to
				 * get the attr; 0 if succeeded;
				 * <0 if rdattr_error and the
				 * attribute cannot be returned.
				 */
				if (error && !(sargp->rdattr_error_req))
					goto done;
				/*
				 * If error then just for entry
				 */
				if (error == 0) {
					fattrp->attrmask |=
						nfs4_ntov_map[i].fbit;
					*amap++ =
						(uint8_t)nfs4_ntov_map[i].nval;
					na++;
					(ntov.attrcnt)++;
				} else if ((error > 0) &&
					(sargp->rdattr_error == NFS4_OK)) {
					sargp->rdattr_error = puterrno4(error);
				}
				error = 0;
			}
		}
	}

	/*
	 * If rdattr_error was set after the return value for it was assigned,
	 * update it.
	 */
	if (prev_rdattr_error != sargp->rdattr_error) {
		na = ntov.na;
		amap = ntov.amap;
		for (i = 0; i < ntov.attrcnt; i++, na++, amap++) {
			k = *amap;
			if (k < FATTR4_RDATTR_ERROR) {
				continue;
			}
			if ((k == FATTR4_RDATTR_ERROR) &&
			    ((*nfs4_ntov_map[k].sv_getit)(
				NFS4ATTR_SUPPORTED, sargp, NULL) == 0)) {

				(void) (*nfs4_ntov_map[k].sv_getit)(
						NFS4ATTR_GETIT, sargp, na);
			}
			break;
		}
	}

	xdr_size = 0;
	na = ntov.na;
	amap = ntov.amap;
	for (i = 0; i < ntov.attrcnt; i++, na++, amap++) {
		xdr_size += xdr_sizeof(nfs4_ntov_map[*amap].xfunc, na);
	}

	fattrp->attrlist4_len = xdr_size;
	if (xdr_size) {
		/* freed by rfs4_op_getattr_free() */
		fattrp->attrlist4 = xdr_attrs = kmem_zalloc(xdr_size, KM_SLEEP);

		xdrmem_create(&xdr, xdr_attrs, xdr_size, XDR_ENCODE);

		na = ntov.na;
		amap = ntov.amap;
		for (i = 0; i < ntov.attrcnt; i++, na++, amap++) {
			if (!(*nfs4_ntov_map[*amap].xfunc)(&xdr, na)) {
				cmn_err(CE_WARN, "do_rfs4_op_getattr: xdr "
					"encode of attribute %d failed\n",
					*amap);
				status = NFS4ERR_SERVERFAULT;
				break;
			}
		}
		/* xdrmem_destroy(&xdrs); */	/* NO-OP */
	} else {
		fattrp->attrlist4 = NULL;
	}
done:

	nfs4_ntov_table_free(&ntov, sargp);

	if (error != 0)
		status = puterrno4(error);

	return (status);
}

/* ARGSUSED */
static void
rfs4_op_getattr(nfs_argop4 *argop, nfs_resop4 *resop, struct svc_req *req,
	struct compound_state *cs)
{
	GETATTR4args *args = &argop->nfs_argop4_u.opgetattr;
	GETATTR4res *resp = &resop->nfs_resop4_u.opgetattr;
	struct nfs4_svgetit_arg sarg;
	struct statvfs64 sb;
	nfsstat4 status;

	if (cs->vp == NULL) {
		*cs->statusp = resp->status = NFS4ERR_NOFILEHANDLE;
		return;
	}

	if (cs->access == CS_ACCESS_DENIED) {
		*cs->statusp = resp->status = NFS4ERR_ACCESS;
		return;
	}

	sarg.sbp = &sb;
	sarg.cs = cs;

	status = bitmap4_to_attrmask(args->attr_request, &sarg);
	if (status == NFS4_OK) {
		status = bitmap4_get_sysattrs(&sarg);
		if (status == NFS4_OK)
			status = do_rfs4_op_getattr(args->attr_request,
				&resp->obj_attributes, &sarg);
	}
	*cs->statusp = resp->status = status;
}

static void
rfs4_op_getattr_free(nfs_resop4 *resop)
{
	GETATTR4res *resp = &resop->nfs_resop4_u.opgetattr;

	nfs4_fattr4_free(&resp->obj_attributes);
}

/* ARGSUSED */
static void
rfs4_op_getfh(nfs_argop4 *argop, nfs_resop4 *resop, struct svc_req *req,
	struct compound_state *cs)
{
	GETFH4res *resp = &resop->nfs_resop4_u.opgetfh;

	if (cs->vp == NULL) {
		*cs->statusp = resp->status = NFS4ERR_NOFILEHANDLE;
		return;
	}
	if (cs->access == CS_ACCESS_DENIED) {
		*cs->statusp = resp->status = NFS4ERR_ACCESS;
		return;
	}

	resp->object.nfs_fh4_val =
		kmem_alloc(cs->fh.nfs_fh4_len, KM_SLEEP);
	nfs_fh4_copy(&cs->fh, &resp->object);
	*cs->statusp = resp->status = NFS4_OK;
}

static void
rfs4_op_getfh_free(nfs_resop4 *resop)
{
	GETFH4res *resp = &resop->nfs_resop4_u.opgetfh;

	if (resp->status == NFS4_OK &&
	    resp->object.nfs_fh4_val != NULL) {
		kmem_free(resp->object.nfs_fh4_val, resp->object.nfs_fh4_len);
		resp->object.nfs_fh4_val = NULL;
		resp->object.nfs_fh4_len = 0;
	}
}

/*
 * illegal: args: void
 *	    res : status (NFS4ERR_OP_ILLEGAL)
 */
/* ARGSUSED */
static void
rfs4_op_illegal(nfs_argop4 *argop, nfs_resop4 *resop,
	struct svc_req *req, struct compound_state *cs)
{
	ILLEGAL4res *resp = &resop->nfs_resop4_u.opillegal;

	resop->resop = OP_ILLEGAL;
	*cs->statusp = resp->status = NFS4ERR_OP_ILLEGAL;
}

/*
 * link: args: SAVED_FH: file, CURRENT_FH: target directory
 *	 res: status. If success - CURRENT_FH unchanged, return change_info
 */
/* ARGSUSED */
static void
rfs4_op_link(nfs_argop4 *argop, nfs_resop4 *resop, struct svc_req *req,
	struct compound_state *cs)
{
	LINK4args *args = &argop->nfs_argop4_u.oplink;
	LINK4res *resp = &resop->nfs_resop4_u.oplink;
	int error;
	vnode_t *vp;
	vnode_t *dvp;
	struct vattr bdva, idva, adva;
	char *nm;
	uint_t  len;

	/* SAVED_FH: source object */
	vp = cs->saved_vp;
	if (vp == NULL) {
		*cs->statusp = resp->status = NFS4ERR_NOFILEHANDLE;
		return;
	}

	/* CURRENT_FH: target directory */
	dvp = cs->vp;
	if (dvp == NULL) {
		*cs->statusp = resp->status = NFS4ERR_NOFILEHANDLE;
		return;
	}

	/*
	 * If there is a non-shared filesystem mounted on this vnode,
	 * do not allow to link any file in this directory.
	 */
	if (vn_ismntpt(dvp)) {
		*cs->statusp = resp->status = NFS4ERR_ACCESS;
		return;
	}

	if (cs->access == CS_ACCESS_DENIED) {
		*cs->statusp = resp->status = NFS4ERR_ACCESS;
		return;
	}

	/* Check source object's type validity */
	if (vp->v_type == VDIR) {
		*cs->statusp = resp->status = NFS4ERR_ISDIR;
		return;
	}

	/* Check target directory's type */
	if (dvp->v_type != VDIR) {
		*cs->statusp = resp->status = NFS4ERR_NOTDIR;
		return;
	}

	if (cs->saved_exi != cs->exi) {
		*cs->statusp = resp->status = NFS4ERR_XDEV;
		return;
	}

	if (!utf8_dir_verify(&args->newname)) {
		*cs->statusp = resp->status = NFS4ERR_INVAL;
		return;
	}

	nm = utf8_to_fn(&args->newname, &len, NULL);
	if (nm == NULL) {
		*cs->statusp = resp->status = NFS4ERR_INVAL;
		return;
	}

	if (len > MAXNAMELEN) {
		*cs->statusp = resp->status = NFS4ERR_NAMETOOLONG;
		kmem_free(nm, len);
		return;
	}

	if (rdonly4(cs->exi, cs->vp, req)) {
		*cs->statusp = resp->status = NFS4ERR_ROFS;
		kmem_free(nm, len);
		return;
	}

	/* Get "before" change value */
	bdva.va_mask = AT_CTIME|AT_SEQ;
	error = VOP_GETATTR(dvp, &bdva, 0, cs->cr);
	if (error) {
		*cs->statusp = resp->status = puterrno4(error);
		kmem_free(nm, len);
		return;
	}

	NFS4_SET_FATTR4_CHANGE(resp->cinfo.before, bdva.va_ctime)

	error = VOP_LINK(dvp, vp, nm, cs->cr);

	kmem_free(nm, len);

	/*
	 * Get the initial "after" sequence number, if it fails, set to zero
	 */
	idva.va_mask = AT_SEQ;
	if (VOP_GETATTR(dvp, &idva, 0, cs->cr))
		idva.va_seq = 0;

	/*
	 * Force modified data and metadata out to stable storage.
	 */
	(void) VOP_FSYNC(vp, FNODSYNC, cs->cr);
	(void) VOP_FSYNC(dvp, 0, cs->cr);

	if (error) {
		*cs->statusp = resp->status = puterrno4(error);
		return;
	}

	/*
	 * Get "after" change value, if it fails, simply return the
	 * before value.
	 */
	adva.va_mask = AT_CTIME|AT_SEQ;
	if (VOP_GETATTR(dvp, &adva, 0, cs->cr)) {
		adva.va_ctime = bdva.va_ctime;
		adva.va_seq = 0;
	}

	NFS4_SET_FATTR4_CHANGE(resp->cinfo.after, adva.va_ctime)

	/*
	 * The cinfo.atomic = TRUE only if we have
	 * non-zero va_seq's, and it has incremented by exactly one
	 * during the VOP_LINK and it didn't change during the VOP_FSYNC.
	 */
	if (bdva.va_seq && idva.va_seq && adva.va_seq &&
			idva.va_seq == (bdva.va_seq + 1) &&
			idva.va_seq == adva.va_seq)
		resp->cinfo.atomic = TRUE;
	else
		resp->cinfo.atomic = FALSE;

	*cs->statusp = resp->status = NFS4_OK;
}

/*
 * Used by rfs4_op_lookup and rfs4_op_lookupp to do the actual work.
 */

/* ARGSUSED */
static nfsstat4
do_rfs4_op_lookup(char *nm, uint_t buflen, struct svc_req *req,
	struct compound_state *cs)
{
	int error;
	int different_export = 0;
	vnode_t *vp, *tvp, *pre_tvp = NULL, *oldvp = NULL;
	struct exportinfo *exi = NULL, *pre_exi = NULL;
	nfsstat4 stat;
	fid_t fid;
	int attrdir, dotdot, walk;
	bool_t is_newvp = FALSE;

	if (cs->vp->v_flag & V_XATTRDIR) {
		attrdir = 1;
		ASSERT(get_fh4_flag(&cs->fh, FH4_ATTRDIR));
	} else {
		attrdir = 0;
		ASSERT(! get_fh4_flag(&cs->fh, FH4_ATTRDIR));
	}

	dotdot = (nm[0] == '.' && nm[1] == '.' && nm[2] == '\0');

	/*
	 * If dotdotting, then need to check whether it's
	 * above the root of a filesystem, or above an
	 * export point.
	 */
	if (dotdot) {

		/*
		 * If dotdotting at the root of a filesystem, then
		 * need to traverse back to the mounted-on filesystem
		 * and do the dotdot lookup there.
		 */
		if (cs->vp->v_flag & VROOT) {

			/*
			 * If at the system root, then can
			 * go up no further.
			 */
			if (VN_CMP(cs->vp, rootdir))
				return (puterrno4(ENOENT));

			/*
			 * Traverse back to the mounted-on filesystem
			 */
			cs->vp = untraverse(cs->vp);

			/*
			 * Set the different_export flag so we remember
			 * to pick up a new exportinfo entry for
			 * this new filesystem.
			 */
			different_export = 1;
		} else {

			/*
			 * If dotdotting above an export point then set
			 * the different_export to get new export info.
			 */
			different_export = nfs_exported(cs->exi, cs->vp);
		}
	}

	error = VOP_LOOKUP(cs->vp, nm, &vp, NULL, 0, NULL, cs->cr);
	if (error)
		return (puterrno4(error));

	/*
	 * If the vnode is in a pseudo filesystem, check whether it is visible.
	 *
	 * XXX if the vnode is a symlink and it is not visible in
	 * a pseudo filesystem, return ENOENT (not following symlink).
	 * V4 client can not mount such symlink. This is a regression
	 * from V2/V3.
	 *
	 * In the same exported filesystem, if the security flavor used
	 * is not an explicitly shared flavor, limit the view to the visible
	 * list entries only. This is not a WRONGSEC case because it's already
	 * checked via PUTROOTFH/PUTPUBFH or PUTFH.
	 */
	if (!different_export &&
	    (PSEUDO(cs->exi) || ! is_exported_sec(cs->nfsflavor, cs->exi) ||
	    cs->access & CS_ACCESS_LIMITED)) {
		if (! nfs_visible(cs->exi, vp, &different_export)) {
			VN_RELE(vp);
			return (puterrno4(ENOENT));
		}
	}

	/*
	 * If it's a mountpoint, then traverse it.
	 */
	if (vn_ismntpt(vp)) {
		pre_exi = cs->exi;	/* save pre-traversed exportinfo */
		pre_tvp = vp;		/* save pre-traversed vnode	*/

		/*
		 * hold pre_tvp to counteract rele by traverse.  We will
		 * need pre_tvp below if checkexport4 fails
		 */
		VN_HOLD(pre_tvp);
		tvp = vp;
		if ((error = traverse(&tvp)) != 0) {
			VN_RELE(vp);
			VN_RELE(pre_tvp);
			return (puterrno4(error));
		}
		vp = tvp;
		different_export = 1;
	} else if (vp->v_vfsp != cs->vp->v_vfsp) {
		/*
		 * The vfsp comparison is to handle the case where
		 * a LOFS mount is shared.  lo_lookup traverses mount points,
		 * and NFS is unaware of local fs transistions because
		 * v_vfsmountedhere isn't set.  For this special LOFS case,
		 * the dir and the obj returned by lookup will have different
		 * vfs ptrs.
		 */
		different_export = 1;
	}

	if (different_export) {

		bzero(&fid, sizeof (fid));
		fid.fid_len = MAXFIDSZ;
		error = vop_fid_pseudo(vp, &fid);
		if (error) {
			VN_RELE(vp);
			if (pre_tvp)
				VN_RELE(pre_tvp);
			return (puterrno4(error));
		}

		if (dotdot)
			exi = nfs_vptoexi(NULL, vp, cs->cr, &walk, NULL, TRUE);
		else
			exi = checkexport4(&vp->v_vfsp->vfs_fsid, &fid, vp);

		if (exi == NULL) {
			if (pre_tvp) {
				/*
				 * If this vnode is a mounted-on vnode,
				 * but the mounted-on file system is not
				 * exported, send back the filehandle for
				 * the mounted-on vnode, not the root of
				 * the mounted-on file system.
				 */
				VN_RELE(vp);
				vp = pre_tvp;
				exi = pre_exi;
			} else {
				VN_RELE(vp);
				return (puterrno4(EACCES));
			}
		} else if (pre_tvp) {
			/* we're done with pre_tvp now. release extra hold */
			VN_RELE(pre_tvp);
		}

		cs->exi = exi;

		/*
		 * Now we do a checkauth4. The reason is that
		 * this client/user may not have access to the new
		 * exported file system, and if he does,
		 * the client/user may be mapped to a different uid.
		 *
		 * We start with a new cr, because the checkauth4 done
		 * in the PUT*FH operation over wrote the cred's uid,
		 * gid, etc, and we want the real thing before calling
		 * checkauth4()
		 */
		crfree(cs->cr);
		cs->cr = crdup(cs->basecr);

		if (cs->vp)
			oldvp = cs->vp;
		cs->vp = vp;
		is_newvp = TRUE;

		stat = call_checkauth4(cs, req);
		if (stat != NFS4_OK) {
			VN_RELE(cs->vp);
			cs->vp = oldvp;
			return (stat);
		}
	}

	/*
	 * After various NFS checks, do a label check on the path
	 * component. The label on this path should either be the
	 * global zone's label or a zone's label. We are only
	 * interested in the zone's label because exported files
	 * in global zone is accessible (though read-only) to
	 * clients. The exportability/visibility check is already
	 * done before reaching this code.
	 */
	if (is_system_labeled()) {
		bslabel_t *clabel;

		ASSERT(req->rq_label != NULL);
		clabel = req->rq_label;
		DTRACE_PROBE2(tx__rfs4__log__info__oplookup__clabel, char *,
		    "got client label from request(1)", struct svc_req *, req);

		if (!blequal(&l_admin_low->tsl_label, clabel)) {
			if (!do_rfs4_label_check(clabel, vp, DOMINANCE_CHECK)) {
				error = EACCES;
				goto err_out;
			}
		} else {
			/*
			 * We grant access to admin_low label clients
			 * only if the client is trusted, i.e. also
			 * running Solaris Trusted Extension.
			 */
			struct sockaddr	*ca;
			int		addr_type;
			void		*ipaddr;
			tsol_tpc_t	*tp;

			ca = (struct sockaddr *)svc_getrpccaller(
			    req->rq_xprt)->buf;
			if (ca->sa_family == AF_INET) {
				addr_type = IPV4_VERSION;
				ipaddr = &((struct sockaddr_in *)ca)->sin_addr;
			} else if (ca->sa_family == AF_INET6) {
				addr_type = IPV6_VERSION;
				ipaddr = &((struct sockaddr_in6 *)
				    ca)->sin6_addr;
			}
			tp = find_tpc(ipaddr, addr_type, B_FALSE);
			if (tp == NULL || tp->tpc_tp.tp_doi !=
			    l_admin_low->tsl_doi || tp->tpc_tp.host_type !=
			    SUN_CIPSO) {
				error = EACCES;
				goto err_out;
			}
		}
	}

	error = makefh4(&cs->fh, vp, cs->exi);

err_out:
	if (error) {
		if (is_newvp) {
			VN_RELE(cs->vp);
			cs->vp = oldvp;
		} else
			VN_RELE(vp);
		return (puterrno4(error));
	}

	if (!is_newvp) {
		if (cs->vp)
			VN_RELE(cs->vp);
		cs->vp = vp;
	} else if (oldvp)
		VN_RELE(oldvp);

	/*
	 * if did lookup on attrdir and didn't lookup .., set named
	 * attr fh flag
	 */
	if (attrdir && ! dotdot)
		set_fh4_flag(&cs->fh, FH4_NAMEDATTR);

	/* Assume false for now, open proc will set this */
	cs->mandlock = FALSE;

	return (NFS4_OK);
}

/* ARGSUSED */
static void
rfs4_op_lookup(nfs_argop4 *argop, nfs_resop4 *resop, struct svc_req *req,
	struct compound_state *cs)
{
	LOOKUP4args *args = &argop->nfs_argop4_u.oplookup;
	LOOKUP4res *resp = &resop->nfs_resop4_u.oplookup;
	char *nm;
	uint_t len;

	if (cs->vp == NULL) {
		*cs->statusp = resp->status = NFS4ERR_NOFILEHANDLE;
		return;
	}

	if (cs->vp->v_type == VLNK) {
		*cs->statusp = resp->status = NFS4ERR_SYMLINK;
		return;
	}

	if (cs->vp->v_type != VDIR) {
		*cs->statusp = resp->status = NFS4ERR_NOTDIR;
		return;
	}

	if (!utf8_dir_verify(&args->objname)) {
		*cs->statusp = resp->status = NFS4ERR_INVAL;
		return;
	}

	nm = utf8_to_str(&args->objname, &len, NULL);
	if (nm == NULL) {
		*cs->statusp = resp->status = NFS4ERR_INVAL;
		return;
	}

	if (len > MAXNAMELEN) {
		*cs->statusp = resp->status = NFS4ERR_NAMETOOLONG;
		kmem_free(nm, len);
		return;
	}

	*cs->statusp = resp->status = do_rfs4_op_lookup(nm, len, req, cs);

	kmem_free(nm, len);
}

/* ARGSUSED */
static void
rfs4_op_lookupp(nfs_argop4 *args, nfs_resop4 *resop, struct svc_req *req,
	struct compound_state *cs)
{
	LOOKUPP4res *resp = &resop->nfs_resop4_u.oplookupp;

	if (cs->vp == NULL) {
		*cs->statusp = resp->status = NFS4ERR_NOFILEHANDLE;
		return;
	}

	if (cs->vp->v_type != VDIR) {
		*cs->statusp = resp->status = NFS4ERR_NOTDIR;
		return;
	}

	*cs->statusp = resp->status = do_rfs4_op_lookup("..", 3, req, cs);

	/*
	 * From NFSV4 Specification, LOOKUPP should not check for
	 * NFS4ERR_WRONGSEC. Retrun NFS4_OK instead.
	 */
	if (resp->status == NFS4ERR_WRONGSEC) {
		*cs->statusp = resp->status = NFS4_OK;
	}
}


/*ARGSUSED2*/
static void
rfs4_op_openattr(nfs_argop4 *argop, nfs_resop4 *resop, struct svc_req *req,
	struct compound_state *cs)
{
	OPENATTR4args	*args = &argop->nfs_argop4_u.opopenattr;
	OPENATTR4res	*resp = &resop->nfs_resop4_u.opopenattr;
	vnode_t		*avp = NULL;
	int		lookup_flags = LOOKUP_XATTR, error;
	int		exp_ro = 0;

	if (cs->vp == NULL) {
		*cs->statusp = resp->status = NFS4ERR_NOFILEHANDLE;
		return;
	}

	/*
	 * Make a couple of checks made by copen()
	 *
	 * Check to make sure underlying fs supports xattrs.  This
	 * is required because solaris filesystem implementations
	 * (UFS/TMPFS) don't enforce the noxattr mount option
	 * in VOP_LOOKUP(LOOKUP_XATTR).  If fs doesn't support this
	 * pathconf cmd or if fs supports cmd but doesn't claim
	 * support for xattr, return NOTSUPP.  It would be better
	 * to use VOP_PATHCONF( _PC_XATTR_ENABLED) for this; however,
	 * that cmd is not available to VOP_PATHCONF interface
	 * (it's only implemented inside pathconf syscall)...
	 *
	 * Verify permission to put attributes on files (access
	 * checks from copen).
	 */

	if ((cs->vp->v_vfsp->vfs_flag & VFS_XATTR) == 0) {
		error = ENOTSUP;
		goto error_out;
	}

	if ((VOP_ACCESS(cs->vp, VREAD, 0, cs->cr) != 0) &&<