]> git.saurik.com Git - apple/xnu.git/blobdiff - bsd/nfs/nfs_node.c
xnu-6153.61.1.tar.gz
[apple/xnu.git] / bsd / nfs / nfs_node.c
index f12696b3616c54819976f82fa8a3cdb3206b5590..8f7da7ea03bff19f3b246f78a8c4c2abc55d01b7 100644 (file)
@@ -1,23 +1,29 @@
 /*
 /*
- * Copyright (c) 2000 Apple Computer, Inc. All rights reserved.
+ * Copyright (c) 2000-2019 Apple Inc. All rights reserved.
  *
  *
- * @APPLE_LICENSE_HEADER_START@
- * 
- * The contents of this file constitute Original Code as defined in and
- * are subject to the Apple Public Source License Version 1.1 (the
- * "License").  You may not use this file except in compliance with the
- * License.  Please obtain a copy of the License at
- * http://www.apple.com/publicsource and read it before using this file.
- * 
- * This Original Code and all software distributed under the License are
- * distributed on an "AS IS" basis, WITHOUT WARRANTY OF ANY KIND, EITHER
+ * @APPLE_OSREFERENCE_LICENSE_HEADER_START@
+ *
+ * This file contains Original Code and/or Modifications of Original Code
+ * as defined in and that are subject to the Apple Public Source License
+ * Version 2.0 (the 'License'). You may not use this file except in
+ * compliance with the License. The rights granted to you under the License
+ * may not be used to create, or enable the creation or redistribution of,
+ * unlawful or unlicensed copies of an Apple operating system, or to
+ * circumvent, violate, or enable the circumvention or violation of, any
+ * terms of an Apple operating system software license agreement.
+ *
+ * Please obtain a copy of the License at
+ * http://www.opensource.apple.com/apsl/ and read it before using this file.
+ *
+ * The Original Code and all software distributed under the License are
+ * distributed on an 'AS IS' basis, WITHOUT WARRANTY OF ANY KIND, EITHER
  * EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES,
  * INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY,
  * EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES,
  * INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE OR NON-INFRINGEMENT.  Please see the
- * License for the specific language governing rights and limitations
- * under the License.
- * 
- * @APPLE_LICENSE_HEADER_END@
+ * FITNESS FOR A PARTICULAR PURPOSE, QUIET ENJOYMENT OR NON-INFRINGEMENT.
+ * Please see the License for the specific language governing rights and
+ * limitations under the License.
+ *
+ * @APPLE_OSREFERENCE_LICENSE_HEADER_END@
  */
 /* Copyright (c) 1995 NeXT Computer, Inc. All Rights Reserved */
 /*
  */
 /* Copyright (c) 1995 NeXT Computer, Inc. All Rights Reserved */
 /*
 
 
 #include <sys/param.h>
 
 
 #include <sys/param.h>
+#include <sys/kernel.h>
 #include <sys/systm.h>
 #include <sys/proc.h>
 #include <sys/systm.h>
 #include <sys/proc.h>
-#include <sys/mount.h>
-#include <sys/namei.h>
+#include <sys/kauth.h>
+#include <sys/mount_internal.h>
+#include <sys/vnode_internal.h>
 #include <sys/vnode.h>
 #include <sys/vnode.h>
+#include <sys/ubc.h>
 #include <sys/malloc.h>
 #include <sys/malloc.h>
+#include <sys/fcntl.h>
+#include <sys/time.h>
 
 #include <nfs/rpcv2.h>
 #include <nfs/nfsproto.h>
 #include <nfs/nfs.h>
 #include <nfs/nfsnode.h>
 
 #include <nfs/rpcv2.h>
 #include <nfs/nfsproto.h>
 #include <nfs/nfs.h>
 #include <nfs/nfsnode.h>
+#include <nfs/nfs_gss.h>
 #include <nfs/nfsmount.h>
 
 #include <nfs/nfsmount.h>
 
-#ifdef MALLOC_DEFINE
-static MALLOC_DEFINE(M_NFSNODE, "NFS node", "NFS vnode private part");
-#endif
+#define NFSNOHASH(fhsum) \
+       (&nfsnodehashtbl[(fhsum) & nfsnodehash])
+static LIST_HEAD(nfsnodehashhead, nfsnode) * nfsnodehashtbl;
+static u_long nfsnodehash;
 
 
-LIST_HEAD(nfsnodehashhead, nfsnode) *nfsnodehashtbl;
-u_long nfsnodehash;
+static lck_grp_t *nfs_node_hash_lck_grp;
+static lck_grp_t *nfs_node_lck_grp;
+static lck_grp_t *nfs_data_lck_grp;
+lck_mtx_t *nfs_node_hash_mutex;
 
 
-#define TRUE   1
-#define        FALSE   0
+#define NFS_NODE_DBG(...) NFS_DBG(NFS_FAC_NODE, 7, ## __VA_ARGS__)
 
 /*
  * Initialize hash links for nfsnodes
  * and build nfsnode free list.
  */
 void
 
 /*
  * Initialize hash links for nfsnodes
  * and build nfsnode free list.
  */
 void
-nfs_nhinit()
+nfs_nhinit(void)
 {
 {
-       nfsnodehashtbl = hashinit(desiredvnodes, M_NFSNODE, &nfsnodehash);
+       nfs_node_hash_lck_grp = lck_grp_alloc_init("nfs_node_hash", LCK_GRP_ATTR_NULL);
+       nfs_node_hash_mutex = lck_mtx_alloc_init(nfs_node_hash_lck_grp, LCK_ATTR_NULL);
+       nfs_node_lck_grp = lck_grp_alloc_init("nfs_node", LCK_GRP_ATTR_NULL);
+       nfs_data_lck_grp = lck_grp_alloc_init("nfs_data", LCK_GRP_ATTR_NULL);
+}
+
+void
+nfs_nhinit_finish(void)
+{
+       lck_mtx_lock(nfs_node_hash_mutex);
+       if (!nfsnodehashtbl) {
+               nfsnodehashtbl = hashinit(desiredvnodes, M_NFSNODE, &nfsnodehash);
+       }
+       lck_mtx_unlock(nfs_node_hash_mutex);
 }
 
 /*
  * Compute an entry in the NFS hash table structure
  */
 u_long
 }
 
 /*
  * Compute an entry in the NFS hash table structure
  */
 u_long
-nfs_hash(fhp, fhsize)
-       register nfsfh_t *fhp;
-       int fhsize;
+nfs_hash(u_char *fhp, int fhsize)
 {
 {
-       register u_char *fhpp;
-       register u_long fhsum;
-       register int i;
+       u_long fhsum;
+       int i;
 
 
-       fhpp = &fhp->fh_bytes[0];
        fhsum = 0;
        fhsum = 0;
-       for (i = 0; i < fhsize; i++)
-               fhsum += *fhpp++;
-       return (fhsum);
+       for (i = 0; i < fhsize; i++) {
+               fhsum += *fhp++;
+       }
+       return fhsum;
 }
 
 }
 
+
+int nfs_case_insensitive(mount_t);
+
+int
+nfs_case_insensitive(mount_t mp)
+{
+       struct nfsmount *nmp = VFSTONFS(mp);
+       int answer = 0;
+       int skip = 0;
+
+       if (nfs_mount_gone(nmp)) {
+               return 0;
+       }
+
+       if (nmp->nm_vers == NFS_VER2) {
+               /* V2 has no way to know */
+               return 0;
+       }
+
+       lck_mtx_lock(&nmp->nm_lock);
+       if (nmp->nm_vers == NFS_VER3) {
+               if (!(nmp->nm_state & NFSSTA_GOTPATHCONF)) {
+                       /* We're holding the node lock so we just return
+                        * with answer as case sensitive. Is very rare
+                        * for file systems not to be homogenous w.r.t. pathconf
+                        */
+                       skip = 1;
+               }
+       } else if (!(nmp->nm_fsattr.nfsa_flags & NFS_FSFLAG_HOMOGENEOUS)) {
+               /* no pathconf info cached */
+               skip = 1;
+       }
+
+       if (!skip && (nmp->nm_fsattr.nfsa_flags & NFS_FSFLAG_CASE_INSENSITIVE)) {
+               answer = 1;
+       }
+
+       lck_mtx_unlock(&nmp->nm_lock);
+
+       return answer;
+}
+
+
 /*
  * Look up a vnode/nfsnode by file handle.
  * Callers must check for mount points!!
  * In all cases, a pointer to a
  * nfsnode structure is returned.
  */
 /*
  * Look up a vnode/nfsnode by file handle.
  * Callers must check for mount points!!
  * In all cases, a pointer to a
  * nfsnode structure is returned.
  */
-int nfs_node_hash_lock;
-
 int
 int
-nfs_nget(mntp, fhp, fhsize, npp)
-       struct mount *mntp;
-       register nfsfh_t *fhp;
-       int fhsize;
-       struct nfsnode **npp;
+nfs_nget(
+       mount_t mp,
+       nfsnode_t dnp,
+       struct componentname *cnp,
+       u_char *fhp,
+       int fhsize,
+       struct nfs_vattr *nvap,
+       u_int64_t *xidp,
+       uint32_t auth,
+       int flags,
+       nfsnode_t *npp)
 {
 {
-       struct proc *p = current_proc();        /* XXX */
-       struct nfsnode *np;
+       nfsnode_t np;
        struct nfsnodehashhead *nhpp;
        struct nfsnodehashhead *nhpp;
-       register struct vnode *vp;
-       struct vnode *nvp;
-       int error;
+       vnode_t vp;
+       int error, nfsvers;
+       mount_t mp2;
+       struct vnode_fsparam vfsp;
+       uint32_t vid;
+
+       FSDBG_TOP(263, mp, dnp, flags, npp);
 
        /* Check for unmount in progress */
 
        /* Check for unmount in progress */
-       if (mntp->mnt_kern_flag & MNTK_UNMOUNT) {
-               *npp = 0;
-               return (EPERM);
+       if (!mp || vfs_isforce(mp)) {
+               *npp = NULL;
+               error = ENXIO;
+               FSDBG_BOT(263, mp, dnp, 0xd1e, error);
+               return error;
        }
        }
+       nfsvers = VFSTONFS(mp)->nm_vers;
 
        nhpp = NFSNOHASH(nfs_hash(fhp, fhsize));
 loop:
 
        nhpp = NFSNOHASH(nfs_hash(fhp, fhsize));
 loop:
+       lck_mtx_lock(nfs_node_hash_mutex);
        for (np = nhpp->lh_first; np != 0; np = np->n_hash.le_next) {
        for (np = nhpp->lh_first; np != 0; np = np->n_hash.le_next) {
-               if (mntp != NFSTOV(np)->v_mount || np->n_fhsize != fhsize ||
-                   bcmp((caddr_t)fhp, (caddr_t)np->n_fhp, fhsize))
+               mp2 = (np->n_hflag & NHINIT) ? np->n_mount : NFSTOMP(np);
+               if (mp != mp2 || np->n_fhsize != fhsize ||
+                   bcmp(fhp, np->n_fhp, fhsize)) {
                        continue;
                        continue;
+               }
+               if (nvap && (nvap->nva_flags & NFS_FFLAG_TRIGGER_REFERRAL) &&
+                   cnp && (cnp->cn_namelen > (fhsize - (int)sizeof(dnp)))) {
+                       /* The name was too long to fit in the file handle.  Check it against the node's name. */
+                       int namecmp = 0;
+                       const char *vname = vnode_getname(NFSTOV(np));
+                       if (vname) {
+                               if (cnp->cn_namelen != (int)strlen(vname)) {
+                                       namecmp = 1;
+                               } else {
+                                       namecmp = strncmp(vname, cnp->cn_nameptr, cnp->cn_namelen);
+                               }
+                               vnode_putname(vname);
+                       }
+                       if (namecmp) { /* full name didn't match */
+                               continue;
+                       }
+               }
+               FSDBG(263, dnp, np, np->n_flag, 0xcace0000);
+               /* if the node is locked, sleep on it */
+               if ((np->n_hflag & NHLOCKED) && !(flags & NG_NOCREATE)) {
+                       np->n_hflag |= NHLOCKWANT;
+                       FSDBG(263, dnp, np, np->n_flag, 0xcace2222);
+                       msleep(np, nfs_node_hash_mutex, PDROP | PINOD, "nfs_nget", NULL);
+                       FSDBG(263, dnp, np, np->n_flag, 0xcace3333);
+                       goto loop;
+               }
                vp = NFSTOV(np);
                vp = NFSTOV(np);
-               if (vget(vp, LK_EXCLUSIVE, p))
+               vid = vnode_vid(vp);
+               lck_mtx_unlock(nfs_node_hash_mutex);
+               if ((error = vnode_getwithvid(vp, vid))) {
+                       /*
+                        * If vnode is being reclaimed or has already
+                        * changed identity, no need to wait.
+                        */
+                       FSDBG_BOT(263, dnp, *npp, 0xcace0d1e, error);
+                       return error;
+               }
+               if ((error = nfs_node_lock(np))) {
+                       /* this only fails if the node is now unhashed */
+                       /* so let's see if we can find/create it again */
+                       FSDBG(263, dnp, *npp, 0xcaced1e2, error);
+                       vnode_put(vp);
+                       if (flags & NG_NOCREATE) {
+                               *npp = 0;
+                               FSDBG_BOT(263, dnp, *npp, 0xcaced1e0, ENOENT);
+                               return ENOENT;
+                       }
                        goto loop;
                        goto loop;
-               *npp = np;
-               return(0);
+               }
+               /* update attributes */
+               if (nvap) {
+                       error = nfs_loadattrcache(np, nvap, xidp, 0);
+               }
+               if (error) {
+                       nfs_node_unlock(np);
+                       vnode_put(vp);
+               } else {
+                       if (dnp && cnp && (flags & NG_MAKEENTRY)) {
+                               cache_enter(NFSTOV(dnp), vp, cnp);
+                       }
+                       /*
+                        * Update the vnode if the name/and or the parent has
+                        * changed. We need to do this so that if getattrlist is
+                        * called asking for ATTR_CMN_NAME, that the "most"
+                        * correct name is being returned. In addition for
+                        * monitored vnodes we need to kick the vnode out of the
+                        * name cache. We do this so that if there are hard
+                        * links in the same directory the link will not be
+                        * found and a lookup will get us here to return the
+                        * name of the current link. In addition by removing the
+                        * name from the name cache the old name will not be
+                        * found after a rename done on another client or the
+                        * server.  The principle reason to do this is because
+                        * Finder is asking for notifications on a directory.
+                        * The directory changes, Finder gets notified, reads
+                        * the directory (which we have purged) and for each
+                        * entry returned calls getattrlist with the name
+                        * returned from readdir. gettattrlist has to call
+                        * namei/lookup to resolve the name, because its not in
+                        * the cache we end up here. We need to update the name
+                        * so Finder will get the name it called us with.
+                        *
+                        * We had an imperfect solution with respect to case
+                        * sensitivity.  There is a test that is run in
+                        * FileBuster that does renames from some name to
+                        * another name differing only in case. It then reads
+                        * the directory looking for the new name, after it
+                        * finds that new name, it ask gettattrlist to verify
+                        * that the name is the new name.  Usually that works,
+                        * but renames generate fsevents and fseventsd will do a
+                        * lookup on the name via lstat. Since that test renames
+                        * old name to new name back and forth there is a race
+                        * that an fsevent will be behind and will access the
+                        * file by the old name, on a case insensitive file
+                        * system that will work. Problem is if we do a case
+                        * sensitive compare, we're going to change the name,
+                        * which the test's getattrlist verification step is
+                        * going to fail. So we will check the case sensitivity
+                        * of the file system and do the appropriate compare. In
+                        * a rare instance for non homogeneous file systems
+                        * w.r.t. pathconf we will use case sensitive compares.
+                        * That could break if the file system is actually case
+                        * insensitive.
+                        *
+                        * Note that V2 does not know the case, so we just
+                        * assume case sensitivity.
+                        *
+                        * This is clearly not perfect due to races, but this is
+                        * as good as its going to get. You can defeat the
+                        * handling of hard links simply by doing:
+                        *
+                        *      while :; do ls -l > /dev/null; done
+                        *
+                        * in a terminal window. Even a single ls -l can cause a
+                        * race.
+                        *
+                        * <rant>What we really need is for the caller, that
+                        * knows the name being used is valid since it got it
+                        * from a readdir to use that name and not ask for the
+                        * ATTR_CMN_NAME</rant>
+                        */
+                       if (dnp && cnp && (vp != NFSTOV(dnp))) {
+                               int update_flags = (vnode_ismonitored((NFSTOV(dnp)))) ? VNODE_UPDATE_CACHE : 0;
+                               int (*cmp)(const char *s1, const char *s2, size_t n);
+
+                               cmp = nfs_case_insensitive(mp) ? strncasecmp : strncmp;
+
+                               if (vp->v_name && (size_t)cnp->cn_namelen != strnlen(vp->v_name, MAXPATHLEN)) {
+                                       update_flags |= VNODE_UPDATE_NAME;
+                               }
+                               if (vp->v_name && cnp->cn_namelen && (*cmp)(cnp->cn_nameptr, vp->v_name, cnp->cn_namelen)) {
+                                       update_flags |= VNODE_UPDATE_NAME;
+                               }
+                               if ((vp->v_name == NULL && cnp->cn_namelen != 0) || (vp->v_name != NULL && cnp->cn_namelen == 0)) {
+                                       update_flags |= VNODE_UPDATE_NAME;
+                               }
+                               if (vnode_parent(vp) != NFSTOV(dnp)) {
+                                       update_flags |= VNODE_UPDATE_PARENT;
+                               }
+                               if (update_flags) {
+                                       NFS_NODE_DBG("vnode_update_identity old name %s new name %.*s update flags = %x\n",
+                                           vp->v_name, cnp->cn_namelen, cnp->cn_nameptr ? cnp->cn_nameptr : "", update_flags);
+                                       vnode_update_identity(vp, NFSTOV(dnp), cnp->cn_nameptr, cnp->cn_namelen, 0, update_flags);
+                               }
+                       }
+
+                       *npp = np;
+               }
+               FSDBG_BOT(263, dnp, *npp, 0xcace0000, error);
+               return error;
        }
        }
+
+       FSDBG(263, mp, dnp, npp, 0xaaaaaaaa);
+
+       if (flags & NG_NOCREATE) {
+               lck_mtx_unlock(nfs_node_hash_mutex);
+               *npp = 0;
+               FSDBG_BOT(263, dnp, *npp, 0x80000001, ENOENT);
+               return ENOENT;
+       }
+
        /*
        /*
-        * Obtain a lock to prevent a race condition if the getnewvnode()
-        * or MALLOC() below happens to block.
+        * allocate and initialize nfsnode and stick it in the hash
+        * before calling getnewvnode().  Anyone finding it in the
+        * hash before initialization is complete will wait for it.
         */
         */
-       if (nfs_node_hash_lock) {
-               while (nfs_node_hash_lock) {
-                       nfs_node_hash_lock = -1;
-                       tsleep(&nfs_node_hash_lock, PVM, "nfsngt", 0);
+       MALLOC_ZONE(np, nfsnode_t, sizeof *np, M_NFSNODE, M_WAITOK);
+       if (!np) {
+               lck_mtx_unlock(nfs_node_hash_mutex);
+               *npp = 0;
+               FSDBG_BOT(263, dnp, *npp, 0x80000001, ENOMEM);
+               return ENOMEM;
+       }
+       bzero(np, sizeof *np);
+       np->n_hflag |= (NHINIT | NHLOCKED);
+       np->n_mount = mp;
+       np->n_auth = auth;
+       TAILQ_INIT(&np->n_opens);
+       TAILQ_INIT(&np->n_lock_owners);
+       TAILQ_INIT(&np->n_locks);
+       np->n_dlink.tqe_next = NFSNOLIST;
+       np->n_dreturn.tqe_next = NFSNOLIST;
+       np->n_monlink.le_next = NFSNOLIST;
+
+       /* ugh... need to keep track of ".zfs" directories to workaround server bugs */
+       if ((nvap->nva_type == VDIR) && cnp && (cnp->cn_namelen == 4) &&
+           (cnp->cn_nameptr[0] == '.') && (cnp->cn_nameptr[1] == 'z') &&
+           (cnp->cn_nameptr[2] == 'f') && (cnp->cn_nameptr[3] == 's')) {
+               np->n_flag |= NISDOTZFS;
+       }
+       if (dnp && (dnp->n_flag & NISDOTZFS)) {
+               np->n_flag |= NISDOTZFSCHILD;
+       }
+
+       if (dnp && cnp && ((cnp->cn_namelen != 2) ||
+           (cnp->cn_nameptr[0] != '.') || (cnp->cn_nameptr[1] != '.'))) {
+               vnode_t dvp = NFSTOV(dnp);
+               if (!vnode_get(dvp)) {
+                       if (!vnode_ref(dvp)) {
+                               np->n_parent = dvp;
+                       }
+                       vnode_put(dvp);
                }
                }
-               goto loop;
        }
        }
-       nfs_node_hash_lock = 1;
 
 
-       /*
-        * Do the MALLOC before the getnewvnode since doing so afterward
-        * might cause a bogus v_data pointer to get dereferenced
-        * elsewhere if MALLOC should block.
-        */
-       MALLOC_ZONE(np, struct nfsnode *, sizeof *np, M_NFSNODE, M_WAITOK);
-               
-       error = getnewvnode(VT_NFS, mntp, nfsv2_vnodeop_p, &nvp);
+       /* setup node's file handle */
+       if (fhsize > NFS_SMALLFH) {
+               MALLOC_ZONE(np->n_fhp, u_char *,
+                   fhsize, M_NFSBIGFH, M_WAITOK);
+               if (!np->n_fhp) {
+                       lck_mtx_unlock(nfs_node_hash_mutex);
+                       FREE_ZONE(np, sizeof *np, M_NFSNODE);
+                       *npp = 0;
+                       FSDBG_BOT(263, dnp, *npp, 0x80000002, ENOMEM);
+                       return ENOMEM;
+               }
+       } else {
+               np->n_fhp = &np->n_fh[0];
+       }
+       bcopy(fhp, np->n_fhp, fhsize);
+       np->n_fhsize = fhsize;
+
+       /* Insert the nfsnode in the hash queue for its new file handle */
+       LIST_INSERT_HEAD(nhpp, np, n_hash);
+       np->n_hflag |= NHHASHED;
+       FSDBG(266, 0, np, np->n_flag, np->n_hflag);
+
+       /* lock the new nfsnode */
+       lck_mtx_init(&np->n_lock, nfs_node_lck_grp, LCK_ATTR_NULL);
+       lck_rw_init(&np->n_datalock, nfs_data_lck_grp, LCK_ATTR_NULL);
+       lck_mtx_init(&np->n_openlock, nfs_open_grp, LCK_ATTR_NULL);
+       lck_mtx_lock(&np->n_lock);
+
+       /* release lock on hash table */
+       lck_mtx_unlock(nfs_node_hash_mutex);
+
+       /* do initial loading of attributes */
+       NACLINVALIDATE(np);
+       NACCESSINVALIDATE(np);
+       error = nfs_loadattrcache(np, nvap, xidp, 1);
        if (error) {
        if (error) {
-               if (nfs_node_hash_lock < 0)
-                       wakeup(&nfs_node_hash_lock);
-               nfs_node_hash_lock = 0;
-               *npp = 0;
+               FSDBG(266, 0, np, np->n_flag, 0xb1eb1e);
+               nfs_node_unlock(np);
+               lck_mtx_lock(nfs_node_hash_mutex);
+               LIST_REMOVE(np, n_hash);
+               np->n_hflag &= ~(NHHASHED | NHINIT | NHLOCKED);
+               if (np->n_hflag & NHLOCKWANT) {
+                       np->n_hflag &= ~NHLOCKWANT;
+                       wakeup(np);
+               }
+               lck_mtx_unlock(nfs_node_hash_mutex);
+               if (np->n_parent) {
+                       if (!vnode_get(np->n_parent)) {
+                               vnode_rele(np->n_parent);
+                               vnode_put(np->n_parent);
+                       }
+                       np->n_parent = NULL;
+               }
+               lck_mtx_destroy(&np->n_lock, nfs_node_lck_grp);
+               lck_rw_destroy(&np->n_datalock, nfs_data_lck_grp);
+               lck_mtx_destroy(&np->n_openlock, nfs_open_grp);
+               if (np->n_fhsize > NFS_SMALLFH) {
+                       FREE_ZONE(np->n_fhp, np->n_fhsize, M_NFSBIGFH);
+               }
                FREE_ZONE(np, sizeof *np, M_NFSNODE);
                FREE_ZONE(np, sizeof *np, M_NFSNODE);
-               return (error);
+               *npp = 0;
+               FSDBG_BOT(263, dnp, *npp, 0x80000003, error);
+               return error;
        }
        }
-       vp = nvp;
-       bzero((caddr_t)np, sizeof *np);
-       vp->v_data = np;
-       np->n_vnode = vp;
-       /*
-        * Insert the nfsnode in the hash queue for its new file handle
-        */
-       LIST_INSERT_HEAD(nhpp, np, n_hash);
-       if (fhsize > NFS_SMALLFH) {
-               MALLOC_ZONE(np->n_fhp, nfsfh_t *,
-                               fhsize, M_NFSBIGFH, M_WAITOK);
+       NFS_CHANGED_UPDATE(nfsvers, np, nvap);
+       if (nvap->nva_type == VDIR) {
+               NFS_CHANGED_UPDATE_NC(nfsvers, np, nvap);
+       }
+
+       /* now, attempt to get a new vnode */
+       vfsp.vnfs_mp = mp;
+       vfsp.vnfs_vtype = nvap->nva_type;
+       vfsp.vnfs_str = "nfs";
+       vfsp.vnfs_dvp = dnp ? NFSTOV(dnp) : NULL;
+       vfsp.vnfs_fsnode = np;
+#if CONFIG_NFS4
+       if (nfsvers == NFS_VER4) {
+#if FIFO
+               if (nvap->nva_type == VFIFO) {
+                       vfsp.vnfs_vops = fifo_nfsv4nodeop_p;
+               } else
+#endif /* FIFO */
+               if (nvap->nva_type == VBLK || nvap->nva_type == VCHR) {
+                       vfsp.vnfs_vops = spec_nfsv4nodeop_p;
+               } else {
+                       vfsp.vnfs_vops = nfsv4_vnodeop_p;
+               }
        } else
        } else
-               np->n_fhp = &np->n_fh;
-       bcopy((caddr_t)fhp, (caddr_t)np->n_fhp, fhsize);
-       np->n_fhsize = fhsize;
-       *npp = np;
+#endif /* CONFIG_NFS4 */
+       {
+#if FIFO
+               if (nvap->nva_type == VFIFO) {
+                       vfsp.vnfs_vops = fifo_nfsv2nodeop_p;
+               } else
+#endif /* FIFO */
+               if (nvap->nva_type == VBLK || nvap->nva_type == VCHR) {
+                       vfsp.vnfs_vops = spec_nfsv2nodeop_p;
+               } else {
+                       vfsp.vnfs_vops = nfsv2_vnodeop_p;
+               }
+       }
+       vfsp.vnfs_markroot = (flags & NG_MARKROOT) ? 1 : 0;
+       vfsp.vnfs_marksystem = 0;
+       vfsp.vnfs_rdev = 0;
+       vfsp.vnfs_filesize = nvap->nva_size;
+       vfsp.vnfs_cnp = cnp;
+       vfsp.vnfs_flags = VNFS_ADDFSREF;
+       if (!dnp || !cnp || !(flags & NG_MAKEENTRY)) {
+               vfsp.vnfs_flags |= VNFS_NOCACHE;
+       }
 
 
-       if (nfs_node_hash_lock < 0)
-               wakeup(&nfs_node_hash_lock);
-       nfs_node_hash_lock = 0;
+#if CONFIG_TRIGGERS
+       if (((nfsvers >= NFS_VER4)
+           )
+           && (nvap->nva_type == VDIR) && (np->n_vattr.nva_flags & NFS_FFLAG_TRIGGER)
+           && !(flags & NG_MARKROOT)) {
+               struct vnode_trigger_param vtp;
+               bzero(&vtp, sizeof(vtp));
+               bcopy(&vfsp, &vtp.vnt_params, sizeof(vfsp));
+               vtp.vnt_resolve_func = nfs_mirror_mount_trigger_resolve;
+               vtp.vnt_unresolve_func = nfs_mirror_mount_trigger_unresolve;
+               vtp.vnt_rearm_func = nfs_mirror_mount_trigger_rearm;
+               vtp.vnt_flags = VNT_AUTO_REARM | VNT_KERN_RESOLVE;
+               error = vnode_create(VNCREATE_TRIGGER, VNCREATE_TRIGGER_SIZE, &vtp, &np->n_vnode);
+       } else
+#endif
+       {
+               error = vnode_create(VNCREATE_FLAVOR, VCREATESIZE, &vfsp, &np->n_vnode);
+       }
+notsup:
+       if (error) {
+               FSDBG(266, 0, np, np->n_flag, 0xb1eb1e);
+               nfs_node_unlock(np);
+               lck_mtx_lock(nfs_node_hash_mutex);
+               LIST_REMOVE(np, n_hash);
+               np->n_hflag &= ~(NHHASHED | NHINIT | NHLOCKED);
+               if (np->n_hflag & NHLOCKWANT) {
+                       np->n_hflag &= ~NHLOCKWANT;
+                       wakeup(np);
+               }
+               lck_mtx_unlock(nfs_node_hash_mutex);
+               if (np->n_parent) {
+                       if (!vnode_get(np->n_parent)) {
+                               vnode_rele(np->n_parent);
+                               vnode_put(np->n_parent);
+                       }
+                       np->n_parent = NULL;
+               }
+               lck_mtx_destroy(&np->n_lock, nfs_node_lck_grp);
+               lck_rw_destroy(&np->n_datalock, nfs_data_lck_grp);
+               lck_mtx_destroy(&np->n_openlock, nfs_open_grp);
+               if (np->n_fhsize > NFS_SMALLFH) {
+                       FREE_ZONE(np->n_fhp, np->n_fhsize, M_NFSBIGFH);
+               }
+               FREE_ZONE(np, sizeof *np, M_NFSNODE);
+               *npp = 0;
+               FSDBG_BOT(263, dnp, *npp, 0x80000004, error);
+               return error;
+       }
+       vp = np->n_vnode;
+       vnode_settag(vp, VT_NFS);
+       /* node is now initialized */
 
 
-       /*
-        * Lock the new nfsnode.
-        */
-       error = vn_lock(vp, LK_EXCLUSIVE | LK_RETRY, p);
+       /* check if anyone's waiting on this node */
+       lck_mtx_lock(nfs_node_hash_mutex);
+       np->n_hflag &= ~(NHINIT | NHLOCKED);
+       if (np->n_hflag & NHLOCKWANT) {
+               np->n_hflag &= ~NHLOCKWANT;
+               wakeup(np);
+       }
+       lck_mtx_unlock(nfs_node_hash_mutex);
 
 
-       return (error);
+       *npp = np;
+
+       FSDBG_BOT(263, dnp, vp, *npp, error);
+       return error;
 }
 
 }
 
+
 int
 int
-nfs_inactive(ap)
-       struct vop_inactive_args /* {
-               struct vnode *a_vp;
-               struct proc *a_p;
-       } */ *ap;
+nfs_vnop_inactive(
+       struct vnop_inactive_args /* {
+                                   *  struct vnodeop_desc *a_desc;
+                                   *  vnode_t a_vp;
+                                   *  vfs_context_t a_context;
+                                   *  } */*ap)
 {
 {
-       register struct nfsnode *np;
-       register struct sillyrename *sp;
-       struct proc *p = current_proc();        /* XXX */
-       extern int prtactive;
-       struct ucred *cred;
-
-       np = VTONFS(ap->a_vp);
-       if (prtactive && ap->a_vp->v_usecount != 0)
-               vprint("nfs_inactive: pushing active", ap->a_vp);
-       if (ap->a_vp->v_type != VDIR) {
-               sp = np->n_sillyrename;
-               np->n_sillyrename = (struct sillyrename *)0;
-       } else
-               sp = (struct sillyrename *)0;
+       vnode_t vp = ap->a_vp;
+       vfs_context_t ctx = ap->a_context;
+       nfsnode_t np;
+       struct nfs_sillyrename *nsp;
+       struct nfs_vattr nvattr;
+       int unhash, attrerr, busyerror, error, inuse, busied, force;
+       struct nfs_open_file *nofp;
+       struct componentname cn;
+       struct nfsmount *nmp;
+       mount_t mp;
+
+       if (vp == NULL) {
+               panic("nfs_vnop_inactive: vp == NULL");
+       }
+       np = VTONFS(vp);
+       if (np == NULL) {
+               panic("nfs_vnop_inactive: np == NULL");
+       }
+
+       nmp = NFSTONMP(np);
+       mp = vnode_mount(vp);
+
+restart:
+       force = (!mp || vfs_isforce(mp));
+       error = 0;
+       inuse = (nfs_mount_state_in_use_start(nmp, NULL) == 0);
 
 
-       if (sp) {
+       /* There shouldn't be any open or lock state at this point */
+       lck_mtx_lock(&np->n_openlock);
+       if (np->n_openrefcnt && !force) {
                /*
                /*
-                * Remove the silly file that was rename'd earlier
+                * vnode_rele and vnode_put drop the vnode lock before
+                * calling VNOP_INACTIVE, so there is a race were the
+                * vnode could become active again. Perhaps there are
+                * other places where this can happen, so if we've got
+                * here we need to get out.
                 */
                 */
-#if DIAGNOSTIC
-               kprintf("nfs_inactive removing %s, dvp=%x, a_vp=%x, ap=%x, np=%x, sp=%x\n", &sp->s_name[0], (unsigned)sp->s_dvp, (unsigned)ap->a_vp, (unsigned)ap, (unsigned)np, (unsigned)sp);
+#ifdef NFS_NODE_DEBUG
+               NP(np, "nfs_vnop_inactive: still open: %d", np->n_openrefcnt);
 #endif
 #endif
+               lck_mtx_unlock(&np->n_openlock);
+               return 0;
+       }
+
+       TAILQ_FOREACH(nofp, &np->n_opens, nof_link) {
+               lck_mtx_lock(&nofp->nof_lock);
+               if (nofp->nof_flags & NFS_OPEN_FILE_BUSY) {
+                       if (!force) {
+                               NP(np, "nfs_vnop_inactive: open file busy");
+                       }
+                       busied = 0;
+               } else {
+                       nofp->nof_flags |= NFS_OPEN_FILE_BUSY;
+                       busied = 1;
+               }
+               lck_mtx_unlock(&nofp->nof_lock);
+               if ((np->n_flag & NREVOKE) || (nofp->nof_flags & NFS_OPEN_FILE_LOST)) {
+                       if (busied) {
+                               nfs_open_file_clear_busy(nofp);
+                       }
+                       continue;
+               }
                /*
                /*
-                * We get a reference (vget) to ensure getnewvnode()
-                * doesn't recycle vp while we're asleep awaiting I/O.
-                * Note we don't need the reference unless usecount is
-                * already zero.  In the case of a forcible unmount it
-                * wont be zero and doing a vget would fail because
-                * vclean holds VXLOCK.
+                * If we just created the file, we already had it open in
+                * anticipation of getting a subsequent open call.  If the
+                * node has gone inactive without being open, we need to
+                * clean up (close) the open done in the create.
                 */
                 */
-                if (ap->a_vp->v_usecount > 0) {
-                        VREF(ap->a_vp);
-                } else if (vget(ap->a_vp, 0, ap->a_p))
-                       panic("nfs_inactive: vget failed");
-               (void) nfs_vinvalbuf(ap->a_vp, 0, sp->s_cred, p, 1);
-               np->n_size = 0;
-               ubc_setsize(ap->a_vp, (off_t)0);
-
-                /* We have a problem. The dvp could have gone away on us while
-                 * in the unmount path. Thus it appears as VBAD and we cannot
-                 * use it. If we tried locking the parent (future), for silly
-                 * rename files, it is unclear where we would lock. The unmount
-                 * code just pulls unlocked vnodes as it goes thru its list and
-                 * yanks them. Could unmount be smarter to see if a busy reg vnode has
-                 * a parent, and not yank it yet? Put in more passes at unmount
-                 * time? In the meantime, just check if it went away on us.
-                 * Could have gone away during the nfs_vinvalbuf or ubc_setsize
-                 * which block.  Or perhaps even before nfs_inactive got called.
-                 */
-                if ((sp->s_dvp)->v_type != VBAD) 
-                        nfs_removeit(sp); /* uses the dvp */
-               cred = sp->s_cred;
-               if (cred != NOCRED) {
-                       sp->s_cred = NOCRED;
-                       crfree(cred);
-               }
-               vrele(sp->s_dvp);
-               FREE_ZONE((caddr_t)sp, sizeof (struct sillyrename), M_NFSREQ);
-               vrele(ap->a_vp);
-       }
-       np->n_flag &= (NMODIFIED | NFLUSHINPROG | NFLUSHWANT | NQNFSEVICTED |
-               NQNFSNONCACHE | NQNFSWRITE);
-       VOP_UNLOCK(ap->a_vp, 0, ap->a_p);
-       return (0);
+#if CONFIG_NFS4
+               if ((nofp->nof_flags & NFS_OPEN_FILE_CREATE) && nofp->nof_creator && !force) {
+                       if (nofp->nof_flags & NFS_OPEN_FILE_REOPEN) {
+                               lck_mtx_unlock(&np->n_openlock);
+                               if (busied) {
+                                       nfs_open_file_clear_busy(nofp);
+                               }
+                               if (inuse) {
+                                       nfs_mount_state_in_use_end(nmp, 0);
+                               }
+                               if (!nfs4_reopen(nofp, NULL)) {
+                                       goto restart;
+                               }
+                       }
+                       nofp->nof_flags &= ~NFS_OPEN_FILE_CREATE;
+                       lck_mtx_unlock(&np->n_openlock);
+                       error = nfs_close(np, nofp, NFS_OPEN_SHARE_ACCESS_BOTH, NFS_OPEN_SHARE_DENY_NONE, ctx);
+                       if (error) {
+                               NP(np, "nfs_vnop_inactive: create close error: %d", error);
+                               nofp->nof_flags |= NFS_OPEN_FILE_CREATE;
+                       }
+                       if (busied) {
+                               nfs_open_file_clear_busy(nofp);
+                       }
+                       if (inuse) {
+                               nfs_mount_state_in_use_end(nmp, error);
+                       }
+                       goto restart;
+               }
+#endif
+               if (nofp->nof_flags & NFS_OPEN_FILE_NEEDCLOSE) {
+                       /*
+                        * If the file is marked as needing reopen, but this was the only
+                        * open on the file, just drop the open.
+                        */
+                       nofp->nof_flags &= ~NFS_OPEN_FILE_NEEDCLOSE;
+                       if ((nofp->nof_flags & NFS_OPEN_FILE_REOPEN) && (nofp->nof_opencnt == 1)) {
+                               nofp->nof_flags &= ~NFS_OPEN_FILE_REOPEN;
+                               nofp->nof_r--;
+                               nofp->nof_opencnt--;
+                               nofp->nof_access = 0;
+                       } else if (!force) {
+                               lck_mtx_unlock(&np->n_openlock);
+                               if (nofp->nof_flags & NFS_OPEN_FILE_REOPEN) {
+                                       if (busied) {
+                                               nfs_open_file_clear_busy(nofp);
+                                       }
+                                       if (inuse) {
+                                               nfs_mount_state_in_use_end(nmp, 0);
+                                       }
+#if CONFIG_NFS4
+                                       if (!nfs4_reopen(nofp, NULL)) {
+                                               goto restart;
+                                       }
+#endif
+                               }
+                               error = nfs_close(np, nofp, NFS_OPEN_SHARE_ACCESS_READ, NFS_OPEN_SHARE_DENY_NONE, ctx);
+                               if (error) {
+                                       NP(np, "nfs_vnop_inactive: need close error: %d", error);
+                                       nofp->nof_flags |= NFS_OPEN_FILE_NEEDCLOSE;
+                               }
+                               if (busied) {
+                                       nfs_open_file_clear_busy(nofp);
+                               }
+                               if (inuse) {
+                                       nfs_mount_state_in_use_end(nmp, error);
+                               }
+                               goto restart;
+                       }
+               }
+               if (nofp->nof_opencnt && !force) {
+                       NP(np, "nfs_vnop_inactive: file still open: %d", nofp->nof_opencnt);
+               }
+               if (!force && (nofp->nof_access || nofp->nof_deny ||
+                   nofp->nof_mmap_access || nofp->nof_mmap_deny ||
+                   nofp->nof_r || nofp->nof_w || nofp->nof_rw ||
+                   nofp->nof_r_dw || nofp->nof_w_dw || nofp->nof_rw_dw ||
+                   nofp->nof_r_drw || nofp->nof_w_drw || nofp->nof_rw_drw ||
+                   nofp->nof_d_r || nofp->nof_d_w || nofp->nof_d_rw ||
+                   nofp->nof_d_r_dw || nofp->nof_d_w_dw || nofp->nof_d_rw_dw ||
+                   nofp->nof_d_r_drw || nofp->nof_d_w_drw || nofp->nof_d_rw_drw)) {
+                       NP(np, "nfs_vnop_inactive: non-zero access: %d %d %d %d # %u.%u %u.%u %u.%u dw %u.%u %u.%u %u.%u drw %u.%u %u.%u %u.%u",
+                           nofp->nof_access, nofp->nof_deny,
+                           nofp->nof_mmap_access, nofp->nof_mmap_deny,
+                           nofp->nof_r, nofp->nof_d_r,
+                           nofp->nof_w, nofp->nof_d_w,
+                           nofp->nof_rw, nofp->nof_d_rw,
+                           nofp->nof_r_dw, nofp->nof_d_r_dw,
+                           nofp->nof_w_dw, nofp->nof_d_w_dw,
+                           nofp->nof_rw_dw, nofp->nof_d_rw_dw,
+                           nofp->nof_r_drw, nofp->nof_d_r_drw,
+                           nofp->nof_w_drw, nofp->nof_d_w_drw,
+                           nofp->nof_rw_drw, nofp->nof_d_rw_drw);
+               }
+               if (busied) {
+                       nfs_open_file_clear_busy(nofp);
+               }
+       }
+       lck_mtx_unlock(&np->n_openlock);
+
+       if (inuse && nfs_mount_state_in_use_end(nmp, error)) {
+               goto restart;
+       }
+
+       nfs_node_lock_force(np);
+
+       if (vnode_vtype(vp) != VDIR) {
+               nsp = np->n_sillyrename;
+               np->n_sillyrename = NULL;
+       } else {
+               nsp = NULL;
+       }
+
+       FSDBG_TOP(264, vp, np, np->n_flag, nsp);
+
+       if (!nsp) {
+               /* no silly file to clean up... */
+               /* clear all flags other than these */
+               np->n_flag &= (NMODIFIED);
+               nfs_node_unlock(np);
+               FSDBG_BOT(264, vp, np, np->n_flag, 0);
+               return 0;
+       }
+       nfs_node_unlock(np);
+
+       /* Remove the silly file that was rename'd earlier */
+
+       /* flush all the buffers */
+       nfs_vinvalbuf2(vp, V_SAVE, vfs_context_thread(ctx), nsp->nsr_cred, 1);
+
+       /* try to get the latest attributes */
+       attrerr = nfs_getattr(np, &nvattr, ctx, NGA_UNCACHED);
+
+       /* Check if we should remove it from the node hash. */
+       /* Leave it if inuse or it has multiple hard links. */
+       if (vnode_isinuse(vp, 0) || (!attrerr && (nvattr.nva_nlink > 1))) {
+               unhash = 0;
+       } else {
+               unhash = 1;
+               ubc_setsize(vp, 0);
+       }
+
+       /* mark this node and the directory busy while we do the remove */
+       busyerror = nfs_node_set_busy2(nsp->nsr_dnp, np, vfs_context_thread(ctx));
+
+       /* lock the node while we remove the silly file */
+       lck_mtx_lock(nfs_node_hash_mutex);
+       while (np->n_hflag & NHLOCKED) {
+               np->n_hflag |= NHLOCKWANT;
+               msleep(np, nfs_node_hash_mutex, PINOD, "nfs_inactive", NULL);
+       }
+       np->n_hflag |= NHLOCKED;
+       lck_mtx_unlock(nfs_node_hash_mutex);
+
+       /* purge the name cache to deter others from finding it */
+       bzero(&cn, sizeof(cn));
+       cn.cn_nameptr = nsp->nsr_name;
+       cn.cn_namelen = nsp->nsr_namlen;
+       nfs_name_cache_purge(nsp->nsr_dnp, np, &cn, ctx);
+
+       FSDBG(264, np, np->n_size, np->n_vattr.nva_size, 0xf00d00f1);
+
+       /* now remove the silly file */
+       nfs_removeit(nsp);
+
+       /* clear all flags other than these */
+       nfs_node_lock_force(np);
+       np->n_flag &= (NMODIFIED);
+       nfs_node_unlock(np);
+
+       if (!busyerror) {
+               nfs_node_clear_busy2(nsp->nsr_dnp, np);
+       }
+
+       if (unhash && vnode_isinuse(vp, 0)) {
+               /* vnode now inuse after silly remove? */
+               unhash = 0;
+               ubc_setsize(vp, np->n_size);
+       }
+
+       lck_mtx_lock(nfs_node_hash_mutex);
+       if (unhash) {
+               /*
+                * remove nfsnode from hash now so we can't accidentally find it
+                * again if another object gets created with the same filehandle
+                * before this vnode gets reclaimed
+                */
+               if (np->n_hflag & NHHASHED) {
+                       LIST_REMOVE(np, n_hash);
+                       np->n_hflag &= ~NHHASHED;
+                       FSDBG(266, 0, np, np->n_flag, 0xb1eb1e);
+               }
+               vnode_recycle(vp);
+       }
+       /* unlock the node */
+       np->n_hflag &= ~NHLOCKED;
+       if (np->n_hflag & NHLOCKWANT) {
+               np->n_hflag &= ~NHLOCKWANT;
+               wakeup(np);
+       }
+       lck_mtx_unlock(nfs_node_hash_mutex);
+
+       /* cleanup sillyrename info */
+       if (nsp->nsr_cred != NOCRED) {
+               kauth_cred_unref(&nsp->nsr_cred);
+       }
+       vnode_rele(NFSTOV(nsp->nsr_dnp));
+       FREE_ZONE(nsp, sizeof(*nsp), M_NFSREQ);
+
+       FSDBG_BOT(264, vp, np, np->n_flag, 0);
+       return 0;
 }
 
 /*
  * Reclaim an nfsnode so that it can be used for other purposes.
  */
 int
 }
 
 /*
  * Reclaim an nfsnode so that it can be used for other purposes.
  */
 int
-nfs_reclaim(ap)
-       struct vop_reclaim_args /* {
-               struct vnode *a_vp;
-       } */ *ap;
+nfs_vnop_reclaim(
+       struct vnop_reclaim_args /* {
+                                  *  struct vnodeop_desc *a_desc;
+                                  *  vnode_t a_vp;
+                                  *  vfs_context_t a_context;
+                                  *  } */*ap)
 {
 {
-       register struct vnode *vp = ap->a_vp;
-       register struct nfsnode *np = VTONFS(vp);
-       register struct nfsmount *nmp = VFSTONFS(vp->v_mount);
-       register struct nfsdmap *dp, *dp2;
-       extern int prtactive;
+       vnode_t vp = ap->a_vp;
+       nfsnode_t np = VTONFS(vp);
+       vfs_context_t ctx = ap->a_context;
+       struct nfs_open_file *nofp, *nextnofp;
+       struct nfs_file_lock *nflp, *nextnflp;
+       struct nfs_lock_owner *nlop, *nextnlop;
+       struct nfsmount *nmp = np->n_mount ? VFSTONFS(np->n_mount) : NFSTONMP(np);
+       mount_t mp = vnode_mount(vp);
+       int force;
 
 
-       if (prtactive && vp->v_usecount != 0)
-               vprint("nfs_reclaim: pushing active", vp);
+       FSDBG_TOP(265, vp, np, np->n_flag, 0);
+       force = (!mp || vfs_isforce(mp) || nfs_mount_gone(nmp));
 
 
-       LIST_REMOVE(np, n_hash);
 
 
-        /*
-         * In case we block during FREE_ZONEs below, get the entry out
-         * of tbe name cache now so subsequent lookups won't find it.
-         */ 
-        cache_purge(vp); 
+       /* There shouldn't be any open or lock state at this point */
+       lck_mtx_lock(&np->n_openlock);
 
 
-       /*
-        * For nqnfs, take it off the timer queue as required.
-        */
-       if ((nmp->nm_flag & NFSMNT_NQNFS) && np->n_timer.cqe_next != 0) {
-               CIRCLEQ_REMOVE(&nmp->nm_timerhead, np, n_timer);
+#if CONFIG_NFS4
+       if (nmp && (nmp->nm_vers >= NFS_VER4)) {
+               /* need to drop a delegation */
+               if (np->n_dreturn.tqe_next != NFSNOLIST) {
+                       /* remove this node from the delegation return list */
+                       lck_mtx_lock(&nmp->nm_lock);
+                       if (np->n_dreturn.tqe_next != NFSNOLIST) {
+                               TAILQ_REMOVE(&nmp->nm_dreturnq, np, n_dreturn);
+                               np->n_dreturn.tqe_next = NFSNOLIST;
+                       }
+                       lck_mtx_unlock(&nmp->nm_lock);
+               }
+               if (np->n_dlink.tqe_next != NFSNOLIST) {
+                       /* remove this node from the delegation list */
+                       lck_mtx_lock(&nmp->nm_lock);
+                       if (np->n_dlink.tqe_next != NFSNOLIST) {
+                               TAILQ_REMOVE(&nmp->nm_delegations, np, n_dlink);
+                               np->n_dlink.tqe_next = NFSNOLIST;
+                       }
+                       lck_mtx_unlock(&nmp->nm_lock);
+               }
+               if ((np->n_openflags & N_DELEG_MASK) && !force) {
+                       /* try to return the delegation */
+                       np->n_openflags &= ~N_DELEG_MASK;
+                       nfs4_delegreturn_rpc(nmp, np->n_fhp, np->n_fhsize, &np->n_dstateid,
+                           R_RECOVER, vfs_context_thread(ctx), vfs_context_ucred(ctx));
+               }
+               if (np->n_attrdirfh) {
+                       FREE(np->n_attrdirfh, M_TEMP);
+                       np->n_attrdirfh = NULL;
+               }
+       }
+#endif
+
+       /* clean up file locks */
+       TAILQ_FOREACH_SAFE(nflp, &np->n_locks, nfl_link, nextnflp) {
+               if (!(nflp->nfl_flags & NFS_FILE_LOCK_DEAD) && !force) {
+                       NP(np, "nfs_vnop_reclaim: lock 0x%llx 0x%llx 0x%x (bc %d)",
+                           nflp->nfl_start, nflp->nfl_end, nflp->nfl_flags, nflp->nfl_blockcnt);
+               }
+               if (!(nflp->nfl_flags & (NFS_FILE_LOCK_BLOCKED | NFS_FILE_LOCK_DEAD))) {
+                       /* try sending an unlock RPC if it wasn't delegated */
+                       if (!(nflp->nfl_flags & NFS_FILE_LOCK_DELEGATED) && !force) {
+                               nmp->nm_funcs->nf_unlock_rpc(np, nflp->nfl_owner, F_WRLCK, nflp->nfl_start, nflp->nfl_end, R_RECOVER,
+                                   NULL, nflp->nfl_owner->nlo_open_owner->noo_cred);
+                       }
+                       lck_mtx_lock(&nflp->nfl_owner->nlo_lock);
+                       TAILQ_REMOVE(&nflp->nfl_owner->nlo_locks, nflp, nfl_lolink);
+                       lck_mtx_unlock(&nflp->nfl_owner->nlo_lock);
+               }
+               TAILQ_REMOVE(&np->n_locks, nflp, nfl_link);
+               nfs_file_lock_destroy(nflp);
+       }
+       /* clean up lock owners */
+       TAILQ_FOREACH_SAFE(nlop, &np->n_lock_owners, nlo_link, nextnlop) {
+               if (!TAILQ_EMPTY(&nlop->nlo_locks) && !force) {
+                       NP(np, "nfs_vnop_reclaim: lock owner with locks");
+               }
+               TAILQ_REMOVE(&np->n_lock_owners, nlop, nlo_link);
+               nfs_lock_owner_destroy(nlop);
        }
        }
+       /* clean up open state */
+       if (np->n_openrefcnt && !force) {
+               NP(np, "nfs_vnop_reclaim: still open: %d", np->n_openrefcnt);
+       }
+       TAILQ_FOREACH_SAFE(nofp, &np->n_opens, nof_link, nextnofp) {
+               if (nofp->nof_flags & NFS_OPEN_FILE_BUSY) {
+                       NP(np, "nfs_vnop_reclaim: open file busy");
+               }
+               if (!(np->n_flag & NREVOKE) && !(nofp->nof_flags & NFS_OPEN_FILE_LOST)) {
+                       if (nofp->nof_opencnt && !force) {
+                               NP(np, "nfs_vnop_reclaim: file still open: %d", nofp->nof_opencnt);
+                       }
+                       if (!force && (nofp->nof_access || nofp->nof_deny ||
+                           nofp->nof_mmap_access || nofp->nof_mmap_deny ||
+                           nofp->nof_r || nofp->nof_w || nofp->nof_rw ||
+                           nofp->nof_r_dw || nofp->nof_w_dw || nofp->nof_rw_dw ||
+                           nofp->nof_r_drw || nofp->nof_w_drw || nofp->nof_rw_drw ||
+                           nofp->nof_d_r || nofp->nof_d_w || nofp->nof_d_rw ||
+                           nofp->nof_d_r_dw || nofp->nof_d_w_dw || nofp->nof_d_rw_dw ||
+                           nofp->nof_d_r_drw || nofp->nof_d_w_drw || nofp->nof_d_rw_drw)) {
+                               NP(np, "nfs_vnop_reclaim: non-zero access: %d %d %d %d # %u.%u %u.%u %u.%u dw %u.%u %u.%u %u.%u drw %u.%u %u.%u %u.%u",
+                                   nofp->nof_access, nofp->nof_deny,
+                                   nofp->nof_mmap_access, nofp->nof_mmap_deny,
+                                   nofp->nof_r, nofp->nof_d_r,
+                                   nofp->nof_w, nofp->nof_d_w,
+                                   nofp->nof_rw, nofp->nof_d_rw,
+                                   nofp->nof_r_dw, nofp->nof_d_r_dw,
+                                   nofp->nof_w_dw, nofp->nof_d_w_dw,
+                                   nofp->nof_rw_dw, nofp->nof_d_rw_dw,
+                                   nofp->nof_r_drw, nofp->nof_d_r_drw,
+                                   nofp->nof_w_drw, nofp->nof_d_w_drw,
+                                   nofp->nof_rw_drw, nofp->nof_d_rw_drw);
+#if CONFIG_NFS4
+                               /* try sending a close RPC if it wasn't delegated */
+                               if (nofp->nof_r || nofp->nof_w || nofp->nof_rw ||
+                                   nofp->nof_r_dw || nofp->nof_w_dw || nofp->nof_rw_dw ||
+                                   nofp->nof_r_drw || nofp->nof_w_drw || nofp->nof_rw_drw) {
+                                       nfs4_close_rpc(np, nofp, NULL, nofp->nof_owner->noo_cred, R_RECOVER);
+                               }
+#endif
+                       }
+               }
+               TAILQ_REMOVE(&np->n_opens, nofp, nof_link);
+               nfs_open_file_destroy(nofp);
+       }
+       lck_mtx_unlock(&np->n_openlock);
+
+       if (np->n_monlink.le_next != NFSNOLIST) {
+               /* Wait for any in-progress getattr to complete, */
+               /* then remove this node from the monitored node list. */
+               lck_mtx_lock(&nmp->nm_lock);
+               while (np->n_mflag & NMMONSCANINPROG) {
+                       struct timespec ts = { .tv_sec = 1, .tv_nsec = 0 };
+                       np->n_mflag |= NMMONSCANWANT;
+                       msleep(&np->n_mflag, &nmp->nm_lock, PZERO - 1, "nfswaitmonscan", &ts);
+               }
+               if (np->n_monlink.le_next != NFSNOLIST) {
+                       LIST_REMOVE(np, n_monlink);
+                       np->n_monlink.le_next = NFSNOLIST;
+               }
+               lck_mtx_unlock(&nmp->nm_lock);
+       }
+
+       lck_mtx_lock(nfs_buf_mutex);
+       if (!force && (!LIST_EMPTY(&np->n_dirtyblkhd) || !LIST_EMPTY(&np->n_cleanblkhd))) {
+               NP(np, "nfs_reclaim: dropping %s buffers", (!LIST_EMPTY(&np->n_dirtyblkhd) ? "dirty" : "clean"));
+       }
+       lck_mtx_unlock(nfs_buf_mutex);
+       nfs_vinvalbuf(vp, V_IGNORE_WRITEERR, ap->a_context, 0);
+
+       lck_mtx_lock(nfs_node_hash_mutex);
+
+       if ((vnode_vtype(vp) != VDIR) && np->n_sillyrename) {
+               if (!force) {
+                       NP(np, "nfs_reclaim: leaving unlinked file %s", np->n_sillyrename->nsr_name);
+               }
+               if (np->n_sillyrename->nsr_cred != NOCRED) {
+                       kauth_cred_unref(&np->n_sillyrename->nsr_cred);
+               }
+               vnode_rele(NFSTOV(np->n_sillyrename->nsr_dnp));
+               FREE_ZONE(np->n_sillyrename, sizeof(*np->n_sillyrename), M_NFSREQ);
+       }
+
+       vnode_removefsref(vp);
+
+       if (np->n_hflag & NHHASHED) {
+               LIST_REMOVE(np, n_hash);
+               np->n_hflag &= ~NHHASHED;
+               FSDBG(266, 0, np, np->n_flag, 0xb1eb1e);
+       }
+       lck_mtx_unlock(nfs_node_hash_mutex);
 
        /*
 
        /*
-        * Free up any directory cookie structures and
-        * large file handle structures that might be associated with
-        * this nfs node.
+        * Free up any directory cookie structures and large file handle
+        * structures that might be associated with this nfs node.
         */
         */
-       if (vp->v_type == VDIR) {
-               dp = np->n_cookies.lh_first;
-               while (dp) {
-                       dp2 = dp;
-                       dp = dp->ndm_list.le_next;
-                       FREE_ZONE((caddr_t)dp2,
-                                       sizeof (struct nfsdmap), M_NFSDIROFF);
-               }
+       nfs_node_lock_force(np);
+       if ((vnode_vtype(vp) == VDIR) && np->n_cookiecache) {
+               FREE_ZONE(np->n_cookiecache, sizeof(struct nfsdmap), M_NFSDIROFF);
        }
        if (np->n_fhsize > NFS_SMALLFH) {
        }
        if (np->n_fhsize > NFS_SMALLFH) {
-               FREE_ZONE((caddr_t)np->n_fhp, np->n_fhsize, M_NFSBIGFH);
+               FREE_ZONE(np->n_fhp, np->n_fhsize, M_NFSBIGFH);
        }
        }
+       if (np->n_vattr.nva_acl) {
+               kauth_acl_free(np->n_vattr.nva_acl);
+       }
+       nfs_node_unlock(np);
+       vnode_clearfsnode(vp);
+
+       if (np->n_parent) {
+               if (!vnode_get(np->n_parent)) {
+                       vnode_rele(np->n_parent);
+                       vnode_put(np->n_parent);
+               }
+               np->n_parent = NULL;
+       }
+
+       lck_mtx_destroy(&np->n_lock, nfs_node_lck_grp);
+       lck_rw_destroy(&np->n_datalock, nfs_data_lck_grp);
+       lck_mtx_destroy(&np->n_openlock, nfs_open_grp);
 
 
-       FREE_ZONE(vp->v_data, sizeof (struct nfsnode), M_NFSNODE);
-       vp->v_data = (void *)0;
-       return (0);
+       FSDBG_BOT(265, vp, np, np->n_flag, 0xd1ed1e);
+       FREE_ZONE(np, sizeof(struct nfsnode), M_NFSNODE);
+       return 0;
 }
 
 /*
 }
 
 /*
- * Lock an nfsnode
+ * Acquire an NFS node lock
  */
  */
+
 int
 int
-nfs_lock(ap)
-       struct vop_lock_args /* {
-                struct vnode *a_vp;
-                int a_flags;
-                struct proc *a_p;
-       } */ *ap;
+nfs_node_lock_internal(nfsnode_t np, int force)
 {
 {
-       register struct vnode *vp = ap->a_vp;
+       FSDBG_TOP(268, np, force, 0, 0);
+       lck_mtx_lock(&np->n_lock);
+       if (!force && !(np->n_hflag && NHHASHED)) {
+               FSDBG_BOT(268, np, 0xdead, 0, 0);
+               lck_mtx_unlock(&np->n_lock);
+               return ENOENT;
+       }
+       FSDBG_BOT(268, np, force, 0, 0);
+       return 0;
+}
 
 
-       /*
-        * Ugh, another place where interruptible mounts will get hung.
-        * If you make this call interruptible, then you have to fix all
-        * the VOP_LOCK() calls to expect interruptibility.
-        */
-       if (vp->v_tag == VT_NON)
-               return (ENOENT); /* ??? -- got to check something and error, but what? */
-        
-       return(lockmgr(&VTONFS(vp)->n_lock, ap->a_flags, &vp->v_interlock,
-                ap->a_p));
-       
+int
+nfs_node_lock(nfsnode_t np)
+{
+       return nfs_node_lock_internal(np, 0);
+}
+
+void
+nfs_node_lock_force(nfsnode_t np)
+{
+       nfs_node_lock_internal(np, 1);
 }
 
 /*
 }
 
 /*
- * Unlock an nfsnode
+ * Release an NFS node lock
+ */
+void
+nfs_node_unlock(nfsnode_t np)
+{
+       FSDBG(269, np, current_thread(), 0, 0);
+       lck_mtx_unlock(&np->n_lock);
+}
+
+/*
+ * Acquire 2 NFS node locks
+ *   - locks taken in reverse address order
+ *   - both or neither of the locks are taken
+ *   - only one lock taken per node (dup nodes are skipped)
  */
 int
  */
 int
-nfs_unlock(ap)
-        struct vop_unlock_args /* {
-                struct vnode *a_vp;
-                int a_flags;
-                struct proc *a_p;
-        } */ *ap;
+nfs_node_lock2(nfsnode_t np1, nfsnode_t np2)
 {
 {
-        struct vnode *vp = ap->a_vp;
+       nfsnode_t first, second;
+       int error;
+
+       first = (np1 > np2) ? np1 : np2;
+       second = (np1 > np2) ? np2 : np1;
+       if ((error = nfs_node_lock(first))) {
+               return error;
+       }
+       if (np1 == np2) {
+               return error;
+       }
+       if ((error = nfs_node_lock(second))) {
+               nfs_node_unlock(first);
+       }
+       return error;
+}
 
 
-        return (lockmgr(&VTONFS(vp)->n_lock, ap->a_flags | LK_RELEASE,
-                &vp->v_interlock, ap->a_p));
+void
+nfs_node_unlock2(nfsnode_t np1, nfsnode_t np2)
+{
+       nfs_node_unlock(np1);
+       if (np1 != np2) {
+               nfs_node_unlock(np2);
+       }
 }
 
 /*
 }
 
 /*
- * Check for a locked nfsnode
+ * Manage NFS node busy state.
+ * (Similar to NFS node locks above)
  */
 int
  */
 int
-nfs_islocked(ap)
-       struct vop_islocked_args /* {
-               struct vnode *a_vp;
-       } */ *ap;
+nfs_node_set_busy(nfsnode_t np, thread_t thd)
 {
 {
-       return (lockstatus(&VTONFS(ap->a_vp)->n_lock));
+       struct timespec ts = { .tv_sec = 2, .tv_nsec = 0 };
+       int error;
 
 
+       if ((error = nfs_node_lock(np))) {
+               return error;
+       }
+       while (ISSET(np->n_flag, NBUSY)) {
+               SET(np->n_flag, NBUSYWANT);
+               msleep(np, &np->n_lock, PZERO - 1, "nfsbusywant", &ts);
+               if ((error = nfs_sigintr(NFSTONMP(np), NULL, thd, 0))) {
+                       break;
+               }
+       }
+       if (!error) {
+               SET(np->n_flag, NBUSY);
+       }
+       nfs_node_unlock(np);
+       return error;
+}
+
+void
+nfs_node_clear_busy(nfsnode_t np)
+{
+       int wanted;
+
+       nfs_node_lock_force(np);
+       wanted = ISSET(np->n_flag, NBUSYWANT);
+       CLR(np->n_flag, NBUSY | NBUSYWANT);
+       nfs_node_unlock(np);
+       if (wanted) {
+               wakeup(np);
+       }
+}
+
+int
+nfs_node_set_busy2(nfsnode_t np1, nfsnode_t np2, thread_t thd)
+{
+       nfsnode_t first, second;
+       int error;
+
+       first = (np1 > np2) ? np1 : np2;
+       second = (np1 > np2) ? np2 : np1;
+       if ((error = nfs_node_set_busy(first, thd))) {
+               return error;
+       }
+       if (np1 == np2) {
+               return error;
+       }
+       if ((error = nfs_node_set_busy(second, thd))) {
+               nfs_node_clear_busy(first);
+       }
+       return error;
+}
+
+void
+nfs_node_clear_busy2(nfsnode_t np1, nfsnode_t np2)
+{
+       nfs_node_clear_busy(np1);
+       if (np1 != np2) {
+               nfs_node_clear_busy(np2);
+       }
+}
+
+/* helper function to sort four nodes in reverse address order (no dupes) */
+static void
+nfs_node_sort4(nfsnode_t np1, nfsnode_t np2, nfsnode_t np3, nfsnode_t np4, nfsnode_t *list, int *lcntp)
+{
+       nfsnode_t na[2], nb[2];
+       int a, b, i, lcnt;
+
+       /* sort pairs then merge */
+       na[0] = (np1 > np2) ? np1 : np2;
+       na[1] = (np1 > np2) ? np2 : np1;
+       nb[0] = (np3 > np4) ? np3 : np4;
+       nb[1] = (np3 > np4) ? np4 : np3;
+       for (a = b = i = lcnt = 0; i < 4; i++) {
+               if (a >= 2) {
+                       list[lcnt] = nb[b++];
+               } else if ((b >= 2) || (na[a] >= nb[b])) {
+                       list[lcnt] = na[a++];
+               } else {
+                       list[lcnt] = nb[b++];
+               }
+               if ((lcnt <= 0) || (list[lcnt] != list[lcnt - 1])) {
+                       lcnt++; /* omit dups */
+               }
+       }
+       if (list[lcnt - 1] == NULL) {
+               lcnt--;
+       }
+       *lcntp = lcnt;
+}
+
+int
+nfs_node_set_busy4(nfsnode_t np1, nfsnode_t np2, nfsnode_t np3, nfsnode_t np4, thread_t thd)
+{
+       nfsnode_t list[4];
+       int i, lcnt, error;
+
+       nfs_node_sort4(np1, np2, np3, np4, list, &lcnt);
+
+       /* Now we can lock using list[0 - lcnt-1] */
+       for (i = 0; i < lcnt; ++i) {
+               if ((error = nfs_node_set_busy(list[i], thd))) {
+                       /* Drop any locks we acquired. */
+                       while (--i >= 0) {
+                               nfs_node_clear_busy(list[i]);
+                       }
+                       return error;
+               }
+       }
+       return 0;
+}
+
+void
+nfs_node_clear_busy4(nfsnode_t np1, nfsnode_t np2, nfsnode_t np3, nfsnode_t np4)
+{
+       nfsnode_t list[4];
+       int lcnt;
+
+       nfs_node_sort4(np1, np2, np3, np4, list, &lcnt);
+       while (--lcnt >= 0) {
+               nfs_node_clear_busy(list[lcnt]);
+       }
+}
+
+/*
+ * Acquire an NFS node data lock
+ */
+void
+nfs_data_lock(nfsnode_t np, int locktype)
+{
+       nfs_data_lock_internal(np, locktype, 1);
+}
+void
+nfs_data_lock_noupdate(nfsnode_t np, int locktype)
+{
+       nfs_data_lock_internal(np, locktype, 0);
+}
+void
+nfs_data_lock_internal(nfsnode_t np, int locktype, int updatesize)
+{
+       FSDBG_TOP(270, np, locktype, np->n_datalockowner, 0);
+       if (locktype == NFS_DATA_LOCK_SHARED) {
+               if (updatesize && ISSET(np->n_flag, NUPDATESIZE)) {
+                       nfs_data_update_size(np, 0);
+               }
+               lck_rw_lock_shared(&np->n_datalock);
+       } else {
+               lck_rw_lock_exclusive(&np->n_datalock);
+               np->n_datalockowner = current_thread();
+               if (updatesize && ISSET(np->n_flag, NUPDATESIZE)) {
+                       nfs_data_update_size(np, 1);
+               }
+       }
+       FSDBG_BOT(270, np, locktype, np->n_datalockowner, 0);
+}
+
+/*
+ * Release an NFS node data lock
+ */
+void
+nfs_data_unlock(nfsnode_t np)
+{
+       nfs_data_unlock_internal(np, 1);
+}
+void
+nfs_data_unlock_noupdate(nfsnode_t np)
+{
+       nfs_data_unlock_internal(np, 0);
+}
+void
+nfs_data_unlock_internal(nfsnode_t np, int updatesize)
+{
+       int mine = (np->n_datalockowner == current_thread());
+       FSDBG_TOP(271, np, np->n_datalockowner, current_thread(), 0);
+       if (updatesize && mine && ISSET(np->n_flag, NUPDATESIZE)) {
+               nfs_data_update_size(np, 1);
+       }
+       np->n_datalockowner = NULL;
+       lck_rw_done(&np->n_datalock);
+       if (updatesize && !mine && ISSET(np->n_flag, NUPDATESIZE)) {
+               nfs_data_update_size(np, 0);
+       }
+       FSDBG_BOT(271, np, np->n_datalockowner, current_thread(), 0);
 }
 
 
 /*
 }
 
 
 /*
- * Nfs abort op, called after namei() when a CREATE/DELETE isn't actually
- * done. Currently nothing to do.
+ * update an NFS node's size
  */
  */
-/* ARGSUSED */
+void
+nfs_data_update_size(nfsnode_t np, int datalocked)
+{
+       int error;
+
+       FSDBG_TOP(272, np, np->n_flag, np->n_size, np->n_newsize);
+       if (!datalocked) {
+               nfs_data_lock(np, NFS_DATA_LOCK_EXCLUSIVE);
+               /* grabbing data lock will automatically update size */
+               nfs_data_unlock(np);
+               FSDBG_BOT(272, np, np->n_flag, np->n_size, np->n_newsize);
+               return;
+       }
+       error = nfs_node_lock(np);
+       if (error || !ISSET(np->n_flag, NUPDATESIZE)) {
+               if (!error) {
+                       nfs_node_unlock(np);
+               }
+               FSDBG_BOT(272, np, np->n_flag, np->n_size, np->n_newsize);
+               return;
+       }
+       CLR(np->n_flag, NUPDATESIZE);
+       np->n_size = np->n_newsize;
+       /* make sure we invalidate buffers the next chance we get */
+       SET(np->n_flag, NNEEDINVALIDATE);
+       nfs_node_unlock(np);
+       ubc_setsize(NFSTOV(np), (off_t)np->n_size); /* XXX error? */
+       FSDBG_BOT(272, np, np->n_flag, np->n_size, np->n_newsize);
+}
+
+#define DODEBUG 1
+
 int
 int
-nfs_abortop(ap)
-       struct vop_abortop_args /* {
-               struct vnode *a_dvp;
-               struct componentname *a_cnp;
-       } */ *ap;
+nfs_mount_is_dirty(mount_t mp)
 {
 {
+       u_long i;
+       nfsnode_t np;
+#ifdef DODEBUG
+       struct timeval now, then, diff;
+       u_long ncnt = 0;
+       microuptime(&now);
+#endif
+       lck_mtx_lock(nfs_node_hash_mutex);
+       for (i = 0; i <= nfsnodehash; i++) {
+               LIST_FOREACH(np, &nfsnodehashtbl[i], n_hash) {
+#ifdef DODEBUG
+                       ncnt++;
+#endif
+                       if (np->n_mount == mp && !LIST_EMPTY(&np->n_dirtyblkhd)) {
+                               goto out;
+                       }
+               }
+       }
+out:
+       lck_mtx_unlock(nfs_node_hash_mutex);
+#ifdef DODEBUG
+       microuptime(&then);
+       timersub(&then, &now, &diff);
+
+       NFS_DBG(NFS_FAC_SOCK, 7, "mount_is_dirty for %s took %lld mics for %ld slots and %ld nodes return %d\n",
+           vfs_statfs(mp)->f_mntfromname, (uint64_t)diff.tv_sec * 1000000LL + diff.tv_usec, i, ncnt, (i <= nfsnodehash));
+#endif
 
 
-       if ((ap->a_cnp->cn_flags & (HASBUF | SAVESTART)) == HASBUF)
-               FREE_ZONE(ap->a_cnp->cn_pnbuf, ap->a_cnp->cn_pnlen, M_NAMEI);
-       return (0);
+       return i <= nfsnodehash;
 }
 }