Merge commit 'v2.6.37-rc1' into kbuild/kbuild

16 files changed, 1110 insertions, 773 deletions

diff --git a/fs/nfsd/Kconfig b/fs/nfsd/Kconfig
index 503b9da..18b3e89 100644
--- a/fs/nfsd/Kconfig
+++ b/fs/nfsd/Kconfig
@@ -28,6 +28,18 @@ config NFSD
 
 	  If unsure, say N.
 
+config NFSD_DEPRECATED
+	bool "Include support for deprecated syscall interface to NFSD"
+	depends on NFSD
+	default y
+	help
+	  The syscall interface to nfsd was obsoleted in 2.6.0 by a new
+	  filesystem based interface.  The old interface is due for removal
+	  in 2.6.40.  If you wish to remove the interface before then
+	  say N.
+
+	  In unsure, say Y.
+
 config NFSD_V2_ACL
 	bool
 	depends on NFSD
@@ -69,7 +81,7 @@ config NFSD_V4
 	depends on NFSD && PROC_FS && EXPERIMENTAL
 	select NFSD_V3
 	select FS_POSIX_ACL
-	select RPCSEC_GSS_KRB5
+	select SUNRPC_GSS
 	help
 	  This option enables support in your system's NFS server for
 	  version 4 of the NFS protocol (RFC 3530).
diff --git a/fs/nfsd/export.c b/fs/nfsd/export.c
index c2a4f71..c0fcb7a 100644
--- a/fs/nfsd/export.c
+++ b/fs/nfsd/export.c
@@ -28,9 +28,6 @@
 typedef struct auth_domain	svc_client;
 typedef struct svc_export	svc_export;
 
-static void		exp_do_unexport(svc_export *unexp);
-static int		exp_verify_string(char *cp, int max);
-
 /*
  * We have two caches.
  * One maps client+vfsmnt+dentry to export options - the export map
@@ -802,6 +799,7 @@ exp_find_key(svc_client *clp, int fsid_type, u32 *fsidv, struct cache_req *reqp)
 	return ek;
 }
 
+#ifdef CONFIG_NFSD_DEPRECATED
 static int exp_set_key(svc_client *clp, int fsid_type, u32 *fsidv,
 		       struct svc_export *exp)
 {
@@ -852,6 +850,7 @@ exp_get_fsid_key(svc_client *clp, int fsid)
 
 	return exp_find_key(clp, FSID_NUM, fsidv, NULL);
 }
+#endif
 
 static svc_export *exp_get_by_name(svc_client *clp, const struct path *path,
 				     struct cache_req *reqp)
@@ -893,6 +892,7 @@ static struct svc_export *exp_parent(svc_client *clp, struct path *path)
 	return exp;
 }
 
+#ifdef CONFIG_NFSD_DEPRECATED
 /*
  * Hashtable locking. Write locks are placed only by user processes
  * wanting to modify export information.
@@ -925,6 +925,19 @@ exp_writeunlock(void)
 {
 	up_write(&hash_sem);
 }
+#else
+
+/* hash_sem not needed once deprecated interface is removed */
+void exp_readlock(void) {}
+static inline void exp_writelock(void){}
+void exp_readunlock(void) {}
+static inline void exp_writeunlock(void){}
+
+#endif
+
+#ifdef CONFIG_NFSD_DEPRECATED
+static void		exp_do_unexport(svc_export *unexp);
+static int		exp_verify_string(char *cp, int max);
 
 static void exp_fsid_unhash(struct svc_export *exp)
 {
@@ -935,10 +948,9 @@ static void exp_fsid_unhash(struct svc_export *exp)
 
 	ek = exp_get_fsid_key(exp->ex_client, exp->ex_fsid);
 	if (!IS_ERR(ek)) {
-		ek->h.expiry_time = get_seconds()-1;
+		sunrpc_invalidate(&ek->h, &svc_expkey_cache);
 		cache_put(&ek->h, &svc_expkey_cache);
 	}
-	svc_expkey_cache.nextcheck = get_seconds();
 }
 
 static int exp_fsid_hash(svc_client *clp, struct svc_export *exp)
@@ -973,10 +985,9 @@ static void exp_unhash(struct svc_export *exp)
 
 	ek = exp_get_key(exp->ex_client, inode->i_sb->s_dev, inode->i_ino);
 	if (!IS_ERR(ek)) {
-		ek->h.expiry_time = get_seconds()-1;
+		sunrpc_invalidate(&ek->h, &svc_expkey_cache);
 		cache_put(&ek->h, &svc_expkey_cache);
 	}
-	svc_expkey_cache.nextcheck = get_seconds();
 }
 	
 /*
@@ -1097,8 +1108,7 @@ out:
 static void
 exp_do_unexport(svc_export *unexp)
 {
-	unexp->h.expiry_time = get_seconds()-1;
-	svc_export_cache.nextcheck = get_seconds();
+	sunrpc_invalidate(&unexp->h, &svc_export_cache);
 	exp_unhash(unexp);
 	exp_fsid_unhash(unexp);
 }
@@ -1150,6 +1160,7 @@ out_unlock:
 	exp_writeunlock();
 	return err;
 }
+#endif /* CONFIG_NFSD_DEPRECATED */
 
 /*
  * Obtain the root fh on behalf of a client.
@@ -1459,25 +1470,43 @@ static void show_secinfo_flags(struct seq_file *m, int flags)
 	show_expflags(m, flags, NFSEXP_SECINFO_FLAGS);
 }
 
+static bool secinfo_flags_equal(int f, int g)
+{
+	f &= NFSEXP_SECINFO_FLAGS;
+	g &= NFSEXP_SECINFO_FLAGS;
+	return f == g;
+}
+
+static int show_secinfo_run(struct seq_file *m, struct exp_flavor_info **fp, struct exp_flavor_info *end)
+{
+	int flags;
+
+	flags = (*fp)->flags;
+	seq_printf(m, ",sec=%d", (*fp)->pseudoflavor);
+	(*fp)++;
+	while (*fp != end && secinfo_flags_equal(flags, (*fp)->flags)) {
+		seq_printf(m, ":%d", (*fp)->pseudoflavor);
+		(*fp)++;
+	}
+	return flags;
+}
+
 static void show_secinfo(struct seq_file *m, struct svc_export *exp)
 {
 	struct exp_flavor_info *f;
 	struct exp_flavor_info *end = exp->ex_flavors + exp->ex_nflavors;
-	int lastflags = 0, first = 0;
+	int flags;
 
 	if (exp->ex_nflavors == 0)
 		return;
-	for (f = exp->ex_flavors; f < end; f++) {
-		if (first || f->flags != lastflags) {
-			if (!first)
-				show_secinfo_flags(m, lastflags);
-			seq_printf(m, ",sec=%d", f->pseudoflavor);
-			lastflags = f->flags;
-		} else {
-			seq_printf(m, ":%d", f->pseudoflavor);
-		}
+	f = exp->ex_flavors;
+	flags = show_secinfo_run(m, &f, end);
+	if (!secinfo_flags_equal(flags, exp->ex_flags))
+		show_secinfo_flags(m, flags);
+	while (f != end) {
+		flags = show_secinfo_run(m, &f, end);
+		show_secinfo_flags(m, flags);
 	}
-	show_secinfo_flags(m, lastflags);
 }
 
 static void exp_flags(struct seq_file *m, int flag, int fsid,
@@ -1532,6 +1561,7 @@ const struct seq_operations nfs_exports_op = {
 	.show	= e_show,
 };
 
+#ifdef CONFIG_NFSD_DEPRECATED
 /*
  * Add or modify a client.
  * Change requests may involve the list of host addresses. The list of
@@ -1563,7 +1593,7 @@ exp_addclient(struct nfsctl_client *ncp)
 	/* Insert client into hashtable. */
 	for (i = 0; i < ncp->cl_naddr; i++) {
 		ipv6_addr_set_v4mapped(ncp->cl_addrlist[i].s_addr, &addr6);
-		auth_unix_add_addr(&addr6, dom);
+		auth_unix_add_addr(&init_net, &addr6, dom);
 	}
 	auth_unix_forget_old(dom);
 	auth_domain_put(dom);
@@ -1621,6 +1651,7 @@ exp_verify_string(char *cp, int max)
 	printk(KERN_NOTICE "nfsd: couldn't validate string %s\n", cp);
 	return 0;
 }
+#endif /* CONFIG_NFSD_DEPRECATED */
 
 /*
  * Initialize the exports module.
diff --git a/fs/nfsd/nfs3proc.c b/fs/nfsd/nfs3proc.c
index 3d68f45..5b7e302 100644
--- a/fs/nfsd/nfs3proc.c
+++ b/fs/nfsd/nfs3proc.c
@@ -168,7 +168,7 @@ nfsd3_proc_read(struct svc_rqst *rqstp, struct nfsd3_readargs *argp,
 	svc_reserve_auth(rqstp, ((1 + NFS3_POST_OP_ATTR_WORDS + 3)<<2) + resp->count +4);
 
 	fh_copy(&resp->fh, &argp->fh);
-	nfserr = nfsd_read(rqstp, &resp->fh, NULL,
+	nfserr = nfsd_read(rqstp, &resp->fh,
 				  argp->offset,
 			   	  rqstp->rq_vec, argp->vlen,
 				  &resp->count);
@@ -271,7 +271,7 @@ nfsd3_proc_mkdir(struct svc_rqst *rqstp, struct nfsd3_createargs *argp,
 	fh_init(&resp->fh, NFS3_FHSIZE);
 	nfserr = nfsd_create(rqstp, &resp->dirfh, argp->name, argp->len,
 				    &argp->attrs, S_IFDIR, 0, &resp->fh);
-
+	fh_unlock(&resp->dirfh);
 	RETURN_STATUS(nfserr);
 }
 
@@ -327,7 +327,7 @@ nfsd3_proc_mknod(struct svc_rqst *rqstp, struct nfsd3_mknodargs *argp,
 	type = nfs3_ftypes[argp->ftype];
 	nfserr = nfsd_create(rqstp, &resp->dirfh, argp->name, argp->len,
 				    &argp->attrs, type, rdev, &resp->fh);
-
+	fh_unlock(&resp->dirfh);
 	RETURN_STATUS(nfserr);
 }
 
@@ -348,6 +348,7 @@ nfsd3_proc_remove(struct svc_rqst *rqstp, struct nfsd3_diropargs *argp,
 	/* Unlink. -S_IFDIR means file must not be a directory */
 	fh_copy(&resp->fh, &argp->fh);
 	nfserr = nfsd_unlink(rqstp, &resp->fh, -S_IFDIR, argp->name, argp->len);
+	fh_unlock(&resp->fh);
 	RETURN_STATUS(nfserr);
 }
 
@@ -367,6 +368,7 @@ nfsd3_proc_rmdir(struct svc_rqst *rqstp, struct nfsd3_diropargs *argp,
 
 	fh_copy(&resp->fh, &argp->fh);
 	nfserr = nfsd_unlink(rqstp, &resp->fh, S_IFDIR, argp->name, argp->len);
+	fh_unlock(&resp->fh);
 	RETURN_STATUS(nfserr);
 }
 
diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c
index eb78e7e..143da2e 100644
--- a/fs/nfsd/nfs4callback.c
+++ b/fs/nfsd/nfs4callback.c
@@ -41,7 +41,6 @@
 
 #define NFSPROC4_CB_NULL 0
 #define NFSPROC4_CB_COMPOUND 1
-#define NFS4_STATEID_SIZE 16
 
 /* Index of predefined Linux callback client operations */
 
@@ -143,8 +142,6 @@ struct nfs4_cb_compound_hdr {
 	u32		minorversion;
 	/* res */
 	int		status;
-	u32		taglen;
-	char		*tag;
 };
 
 static struct {
@@ -205,6 +202,16 @@ nfs_cb_stat_to_errno(int stat)
  */
 
 static void
+encode_stateid(struct xdr_stream *xdr, stateid_t *sid)
+{
+	__be32 *p;
+
+	RESERVE_SPACE(sizeof(stateid_t));
+	WRITE32(sid->si_generation);
+	WRITEMEM(&sid->si_opaque, sizeof(stateid_opaque_t));
+}
+
+static void
 encode_cb_compound_hdr(struct xdr_stream *xdr, struct nfs4_cb_compound_hdr *hdr)
 {
 	__be32 * p;
@@ -229,10 +236,10 @@ encode_cb_recall(struct xdr_stream *xdr, struct nfs4_delegation *dp,
 	__be32 *p;
 	int len = dp->dl_fh.fh_size;
 
-	RESERVE_SPACE(12+sizeof(dp->dl_stateid) + len);
+	RESERVE_SPACE(4);
 	WRITE32(OP_CB_RECALL);
-	WRITE32(dp->dl_stateid.si_generation);
-	WRITEMEM(&dp->dl_stateid.si_opaque, sizeof(stateid_opaque_t));
+	encode_stateid(xdr, &dp->dl_stateid);
+	RESERVE_SPACE(8 + (XDR_QUADLEN(len) << 2));
 	WRITE32(0); /* truncate optimization not implemented */
 	WRITE32(len);
 	WRITEMEM(&dp->dl_fh.fh_base, len);
@@ -240,10 +247,11 @@ encode_cb_recall(struct xdr_stream *xdr, struct nfs4_delegation *dp,
 }
 
 static void
-encode_cb_sequence(struct xdr_stream *xdr, struct nfsd4_cb_sequence *args,
+encode_cb_sequence(struct xdr_stream *xdr, struct nfsd4_callback *cb,
 		   struct nfs4_cb_compound_hdr *hdr)
 {
 	__be32 *p;
+	struct nfsd4_session *ses = cb->cb_clp->cl_cb_session;
 
 	if (hdr->minorversion == 0)
 		return;
@@ -251,8 +259,8 @@ encode_cb_sequence(struct xdr_stream *xdr, struct nfsd4_cb_sequence *args,
 	RESERVE_SPACE(1 + NFS4_MAX_SESSIONID_LEN + 20);
 
 	WRITE32(OP_CB_SEQUENCE);
-	WRITEMEM(args->cbs_clp->cl_sessionid.data, NFS4_MAX_SESSIONID_LEN);
-	WRITE32(args->cbs_clp->cl_cb_seq_nr);
+	WRITEMEM(ses->se_sessionid.data, NFS4_MAX_SESSIONID_LEN);
+	WRITE32(ses->se_cb_seq_nr);
 	WRITE32(0);		/* slotid, always 0 */
 	WRITE32(0);		/* highest slotid always 0 */
 	WRITE32(0);		/* cachethis always 0 */
@@ -272,18 +280,18 @@ nfs4_xdr_enc_cb_null(struct rpc_rqst *req, __be32 *p)
 
 static int
 nfs4_xdr_enc_cb_recall(struct rpc_rqst *req, __be32 *p,
-		struct nfs4_rpc_args *rpc_args)
+		struct nfsd4_callback *cb)
 {
 	struct xdr_stream xdr;
-	struct nfs4_delegation *args = rpc_args->args_op;
+	struct nfs4_delegation *args = cb->cb_op;
 	struct nfs4_cb_compound_hdr hdr = {
-		.ident = args->dl_ident,
-		.minorversion = rpc_args->args_seq.cbs_minorversion,
+		.ident = cb->cb_clp->cl_cb_ident,
+		.minorversion = cb->cb_minorversion,
 	};
 
 	xdr_init_encode(&xdr, &req->rq_snd_buf, p);
 	encode_cb_compound_hdr(&xdr, &hdr);
-	encode_cb_sequence(&xdr, &rpc_args->args_seq, &hdr);
+	encode_cb_sequence(&xdr, cb, &hdr);
 	encode_cb_recall(&xdr, args, &hdr);
 	encode_cb_nops(&hdr);
 	return 0;
@@ -293,13 +301,14 @@ nfs4_xdr_enc_cb_recall(struct rpc_rqst *req, __be32 *p,
 static int
 decode_cb_compound_hdr(struct xdr_stream *xdr, struct nfs4_cb_compound_hdr *hdr){
         __be32 *p;
+	u32 taglen;
 
         READ_BUF(8);
         READ32(hdr->status);
-        READ32(hdr->taglen);
-        READ_BUF(hdr->taglen + 4);
-        hdr->tag = (char *)p;
-        p += XDR_QUADLEN(hdr->taglen);
+	/* We've got no use for the tag; ignore it: */
+        READ32(taglen);
+        READ_BUF(taglen + 4);
+        p += XDR_QUADLEN(taglen);
         READ32(hdr->nops);
         return 0;
 }
@@ -330,15 +339,16 @@ decode_cb_op_hdr(struct xdr_stream *xdr, enum nfs_opnum4 expected)
  * with a single slot.
  */
 static int
-decode_cb_sequence(struct xdr_stream *xdr, struct nfsd4_cb_sequence *res,
+decode_cb_sequence(struct xdr_stream *xdr, struct nfsd4_callback *cb,
 		   struct rpc_rqst *rqstp)
 {
+	struct nfsd4_session *ses = cb->cb_clp->cl_cb_session;
 	struct nfs4_sessionid id;
 	int status;
 	u32 dummy;
 	__be32 *p;
 
-	if (res->cbs_minorversion == 0)
+	if (cb->cb_minorversion == 0)
 		return 0;
 
 	status = decode_cb_op_hdr(xdr, OP_CB_SEQUENCE);
@@ -354,13 +364,12 @@ decode_cb_sequence(struct xdr_stream *xdr, struct nfsd4_cb_sequence *res,
 	READ_BUF(NFS4_MAX_SESSIONID_LEN + 16);
 	memcpy(id.data, p, NFS4_MAX_SESSIONID_LEN);
 	p += XDR_QUADLEN(NFS4_MAX_SESSIONID_LEN);
-	if (memcmp(id.data, res->cbs_clp->cl_sessionid.data,
-		   NFS4_MAX_SESSIONID_LEN)) {
+	if (memcmp(id.data, ses->se_sessionid.data, NFS4_MAX_SESSIONID_LEN)) {
 		dprintk("%s Invalid session id\n", __func__);
 		goto out;
 	}
 	READ32(dummy);
-	if (dummy != res->cbs_clp->cl_cb_seq_nr) {
+	if (dummy != ses->se_cb_seq_nr) {
 		dprintk("%s Invalid sequence number\n", __func__);
 		goto out;
 	}
@@ -384,7 +393,7 @@ nfs4_xdr_dec_cb_null(struct rpc_rqst *req, __be32 *p)
 
 static int
 nfs4_xdr_dec_cb_recall(struct rpc_rqst *rqstp, __be32 *p,
-		struct nfsd4_cb_sequence *seq)
+		struct nfsd4_callback *cb)
 {
 	struct xdr_stream xdr;
 	struct nfs4_cb_compound_hdr hdr;
@@ -394,8 +403,8 @@ nfs4_xdr_dec_cb_recall(struct rpc_rqst *rqstp, __be32 *p,
 	status = decode_cb_compound_hdr(&xdr, &hdr);
 	if (status)
 		goto out;
-	if (seq) {
-		status = decode_cb_sequence(&xdr, seq, rqstp);
+	if (cb) {
+		status = decode_cb_sequence(&xdr, cb, rqstp);
 		if (status)
 			goto out;
 	}
@@ -464,30 +473,34 @@ static int max_cb_time(void)
 /* Reference counting, callback cleanup, etc., all look racy as heck.
  * And why is cl_cb_set an atomic? */
 
-int setup_callback_client(struct nfs4_client *clp, struct nfs4_cb_conn *cb)
+int setup_callback_client(struct nfs4_client *clp, struct nfs4_cb_conn *conn)
 {
 	struct rpc_timeout	timeparms = {
 		.to_initval	= max_cb_time(),
 		.to_retries	= 0,
 	};
 	struct rpc_create_args args = {
-		.protocol	= XPRT_TRANSPORT_TCP,
-		.address	= (struct sockaddr *) &cb->cb_addr,
-		.addrsize	= cb->cb_addrlen,
+		.net		= &init_net,
+		.address	= (struct sockaddr *) &conn->cb_addr,
+		.addrsize	= conn->cb_addrlen,
 		.timeout	= &timeparms,
 		.program	= &cb_program,
-		.prognumber	= cb->cb_prog,
 		.version	= 0,
 		.authflavor	= clp->cl_flavor,
 		.flags		= (RPC_CLNT_CREATE_NOPING | RPC_CLNT_CREATE_QUIET),
-		.client_name    = clp->cl_principal,
 	};
 	struct rpc_clnt *client;
 
-	if (!clp->cl_principal && (clp->cl_flavor >= RPC_AUTH_GSS_KRB5))
-		return -EINVAL;
-	if (cb->cb_minorversion) {
-		args.bc_xprt = cb->cb_xprt;
+	if (clp->cl_minorversion == 0) {
+		if (!clp->cl_principal && (clp->cl_flavor >= RPC_AUTH_GSS_KRB5))
+			return -EINVAL;
+		args.client_name = clp->cl_principal;
+		args.prognumber	= conn->cb_prog,
+		args.protocol = XPRT_TRANSPORT_TCP;
+		clp->cl_cb_ident = conn->cb_ident;
+	} else {
+		args.bc_xprt = conn->cb_xprt;
+		args.prognumber = clp->cl_cb_session->se_cb_prog;
 		args.protocol = XPRT_TRANSPORT_BC_TCP;
 	}
 	/* Create RPC client */
@@ -497,7 +510,7 @@ int setup_callback_client(struct nfs4_client *clp, struct nfs4_cb_conn *cb)
 			PTR_ERR(client));
 		return PTR_ERR(client);
 	}
-	nfsd4_set_callback_client(clp, client);
+	clp->cl_cb_client = client;
 	return 0;
 
 }
@@ -510,7 +523,7 @@ static void warn_no_callback_path(struct nfs4_client *clp, int reason)
 
 static void nfsd4_cb_probe_done(struct rpc_task *task, void *calldata)
 {
-	struct nfs4_client *clp = calldata;
+	struct nfs4_client *clp = container_of(calldata, struct nfs4_client, cl_cb_null);
 
 	if (task->tk_status)
 		warn_no_callback_path(clp, task->tk_status);
@@ -519,6 +532,8 @@ static void nfsd4_cb_probe_done(struct rpc_task *task, void *calldata)
 }
 
 static const struct rpc_call_ops nfsd4_cb_probe_ops = {
+	/* XXX: release method to ensure we set the cb channel down if
+	 * necessary on early failure? */
 	.rpc_call_done = nfsd4_cb_probe_done,
 };
 
@@ -534,38 +549,42 @@ int set_callback_cred(void)
 	return 0;
 }
 
+static struct workqueue_struct *callback_wq;
 
-void do_probe_callback(struct nfs4_client *clp)
+static void do_probe_callback(struct nfs4_client *clp)
 {
-	struct rpc_message msg = {
-		.rpc_proc       = &nfs4_cb_procedures[NFSPROC4_CLNT_CB_NULL],
-		.rpc_argp       = clp,
-		.rpc_cred	= callback_cred
-	};
-	int status;
+	struct nfsd4_callback *cb = &clp->cl_cb_null;
 
-	status = rpc_call_async(clp->cl_cb_client, &msg,
-				RPC_TASK_SOFT | RPC_TASK_SOFTCONN,
-				&nfsd4_cb_probe_ops, (void *)clp);
-	if (status)
-		warn_no_callback_path(clp, status);
+	cb->cb_op = NULL;
+	cb->cb_clp = clp;
+
+	cb->cb_msg.rpc_proc = &nfs4_cb_procedures[NFSPROC4_CLNT_CB_NULL];
+	cb->cb_msg.rpc_argp = NULL;
+	cb->cb_msg.rpc_resp = NULL;
+	cb->cb_msg.rpc_cred = callback_cred;
+
+	cb->cb_ops = &nfsd4_cb_probe_ops;
+
+	queue_work(callback_wq, &cb->cb_work);
 }
 
 /*
- * Set up the callback client and put a NFSPROC4_CB_NULL on the wire...
+ * Poke the callback thread to process any updates to the callback
+ * parameters, and send a null probe.
  */
-void nfsd4_probe_callback(struct nfs4_client *clp, struct nfs4_cb_conn *cb)
+void nfsd4_probe_callback(struct nfs4_client *clp)
 {
-	int status;
+	set_bit(NFSD4_CLIENT_CB_UPDATE, &clp->cl_cb_flags);
+	do_probe_callback(clp);
+}
 
+void nfsd4_change_callback(struct nfs4_client *clp, struct nfs4_cb_conn *conn)
+{
 	BUG_ON(atomic_read(&clp->cl_cb_set));
 
-	status = setup_callback_client(clp, cb);
-	if (status) {
-		warn_no_callback_path(clp, status);
-		return;
-	}
-	do_probe_callback(clp);
+	spin_lock(&clp->cl_lock);
+	memcpy(&clp->cl_cb_conn, conn, sizeof(struct nfs4_cb_conn));
+	spin_unlock(&clp->cl_lock);
 }
 
 /*
@@ -576,8 +595,7 @@ void nfsd4_probe_callback(struct nfs4_client *clp, struct nfs4_cb_conn *cb)
 static int nfsd41_cb_setup_sequence(struct nfs4_client *clp,
 		struct rpc_task *task)
 {
-	struct nfs4_rpc_args *args = task->tk_msg.rpc_argp;
-	u32 *ptr = (u32 *)clp->cl_sessionid.data;
+	u32 *ptr = (u32 *)clp->cl_cb_session->se_sessionid.data;
 	int status = 0;
 
 	dprintk("%s: %u:%u:%u:%u\n", __func__,
@@ -589,14 +607,6 @@ static int nfsd41_cb_setup_sequence(struct nfs4_client *clp,
 		status = -EAGAIN;
 		goto out;
 	}
-
-	/*
-	 * We'll need the clp during XDR encoding and decoding,
-	 * and the sequence during decoding to verify the reply
-	 */
-	args->args_seq.cbs_clp = clp;
-	task->tk_msg.rpc_resp = &args->args_seq;
-
 out:
 	dprintk("%s status=%d\n", __func__, status);
 	return status;
@@ -608,13 +618,13 @@ out:
  */
 static void nfsd4_cb_prepare(struct rpc_task *task, void *calldata)
 {
-	struct nfs4_delegation *dp = calldata;
+	struct nfsd4_callback *cb = calldata;
+	struct nfs4_delegation *dp = container_of(cb, struct nfs4_delegation, dl_recall);
 	struct nfs4_client *clp = dp->dl_client;
-	struct nfs4_rpc_args *args = task->tk_msg.rpc_argp;
-	u32 minorversion = clp->cl_cb_conn.cb_minorversion;
+	u32 minorversion = clp->cl_minorversion;
 	int status = 0;
 
-	args->args_seq.cbs_minorversion = minorversion;
+	cb->cb_minorversion = minorversion;
 	if (minorversion) {
 		status = nfsd41_cb_setup_sequence(clp, task);
 		if (status) {
@@ -631,19 +641,20 @@ static void nfsd4_cb_prepare(struct rpc_task *task, void *calldata)
 
 static void nfsd4_cb_done(struct rpc_task *task, void *calldata)
 {
-	struct nfs4_delegation *dp = calldata;
+	struct nfsd4_callback *cb = calldata;
+	struct nfs4_delegation *dp = container_of(cb, struct nfs4_delegation, dl_recall);
 	struct nfs4_client *clp = dp->dl_client;
 
 	dprintk("%s: minorversion=%d\n", __func__,
-		clp->cl_cb_conn.cb_minorversion);
+		clp->cl_minorversion);
 
-	if (clp->cl_cb_conn.cb_minorversion) {
+	if (clp->cl_minorversion) {
 		/* No need for lock, access serialized in nfsd4_cb_prepare */
-		++clp->cl_cb_seq_nr;
+		++clp->cl_cb_session->se_cb_seq_nr;
 		clear_bit(0, &clp->cl_cb_slot_busy);
 		rpc_wake_up_next(&clp->cl_cb_waitq);
 		dprintk("%s: freed slot, new seqid=%d\n", __func__,
-			clp->cl_cb_seq_nr);
+			clp->cl_cb_session->se_cb_seq_nr);
 
 		/* We're done looking into the sequence information */
 		task->tk_msg.rpc_resp = NULL;
@@ -653,7 +664,8 @@ static void nfsd4_cb_done(struct rpc_task *task, void *calldata)
 
 static void nfsd4_cb_recall_done(struct rpc_task *task, void *calldata)
 {
-	struct nfs4_delegation *dp = calldata;
+	struct nfsd4_callback *cb = calldata;
+	struct nfs4_delegation *dp = container_of(cb, struct nfs4_delegation, dl_recall);
 	struct nfs4_client *clp = dp->dl_client;
 	struct rpc_clnt *current_rpc_client = clp->cl_cb_client;
 
@@ -667,28 +679,28 @@ static void nfsd4_cb_recall_done(struct rpc_task *task, void *calldata)
 	}
 
 	switch (task->tk_status) {
-	case -EIO:
+	case 0:
+		return;
+	case -EBADHANDLE:
+	case -NFS4ERR_BAD_STATEID:
+		/* Race: client probably got cb_recall
+		 * before open reply granting delegation */
+		break;
+	default:
 		/* Network partition? */
 		atomic_set(&clp->cl_cb_set, 0);
 		warn_no_callback_path(clp, task->tk_status);
 		if (current_rpc_client != task->tk_client) {
 			/* queue a callback on the new connection: */
+			atomic_inc(&dp->dl_count);
 			nfsd4_cb_recall(dp);
 			return;
 		}
-	case -EBADHANDLE:
-	case -NFS4ERR_BAD_STATEID:
-		/* Race: client probably got cb_recall
-		 * before open reply granting delegation */
-		break;
-	default:
-		/* success, or error we can't handle */
-		return;
 	}
 	if (dp->dl_retries--) {
 		rpc_delay(task, 2*HZ);
 		task->tk_status = 0;
-		rpc_restart_call(task);
+		rpc_restart_call_prepare(task);
 		return;
 	} else {
 		atomic_set(&clp->cl_cb_set, 0);
@@ -698,7 +710,8 @@ static void nfsd4_cb_recall_done(struct rpc_task *task, void *calldata)
 
 static void nfsd4_cb_recall_release(void *calldata)
 {
-	struct nfs4_delegation *dp = calldata;
+	struct nfsd4_callback *cb = calldata;
+	struct nfs4_delegation *dp = container_of(cb, struct nfs4_delegation, dl_recall);
 
 	nfs4_put_delegation(dp);
 }
@@ -709,8 +722,6 @@ static const struct rpc_call_ops nfsd4_cb_recall_ops = {
 	.rpc_release = nfsd4_cb_recall_release,
 };
 
-static struct workqueue_struct *callback_wq;
-
 int nfsd4_create_callback_queue(void)
 {
 	callback_wq = create_singlethread_workqueue("nfsd4_callbacks");
@@ -725,59 +736,88 @@ void nfsd4_destroy_callback_queue(void)
 }
 
 /* must be called under the state lock */
-void nfsd4_set_callback_client(struct nfs4_client *clp, struct rpc_clnt *new)
+void nfsd4_shutdown_callback(struct nfs4_client *clp)
 {
-	struct rpc_clnt *old = clp->cl_cb_client;
-
-	clp->cl_cb_client = new;
+	set_bit(NFSD4_CLIENT_KILL, &clp->cl_cb_flags);
 	/*
-	 * After this, any work that saw the old value of cl_cb_client will
-	 * be gone:
+	 * Note this won't actually result in a null callback;
+	 * instead, nfsd4_do_callback_rpc() will detect the killed
+	 * client, destroy the rpc client, and stop:
 	 */
+	do_probe_callback(clp);
 	flush_workqueue(callback_wq);
-	/* So we can safely shut it down: */
-	if (old)
-		rpc_shutdown_client(old);
 }
 
-/*
- * called with dp->dl_count inc'ed.
- */
-static void _nfsd4_cb_recall(struct nfs4_delegation *dp)
+void nfsd4_release_cb(struct nfsd4_callback *cb)
 {
-	struct nfs4_client *clp = dp->dl_client;
-	struct rpc_clnt *clnt = clp->cl_cb_client;
-	struct nfs4_rpc_args *args = &dp->dl_recall.cb_args;
-	struct rpc_message msg = {
-		.rpc_proc = &nfs4_cb_procedures[NFSPROC4_CLNT_CB_RECALL],
-		.rpc_cred = callback_cred
-	};
-	int status;
+	if (cb->cb_ops->rpc_release)
+		cb->cb_ops->rpc_release(cb);
+}
 
-	if (clnt == NULL)
-		return; /* Client is shutting down; give up. */
+void nfsd4_process_cb_update(struct nfsd4_callback *cb)
+{
+	struct nfs4_cb_conn conn;
+	struct nfs4_client *clp = cb->cb_clp;
+	int err;
 
-	args->args_op = dp;
-	msg.rpc_argp = args;
-	dp->dl_retries = 1;
-	status = rpc_call_async(clnt, &msg, RPC_TASK_SOFT,
-				&nfsd4_cb_recall_ops, dp);
-	if (status)
-		nfs4_put_delegation(dp);
+	/*
+	 * This is either an update, or the client dying; in either case,
+	 * kill the old client:
+	 */
+	if (clp->cl_cb_client) {
+		rpc_shutdown_client(clp->cl_cb_client);
+		clp->cl_cb_client = NULL;
+	}
+	if (test_bit(NFSD4_CLIENT_KILL, &clp->cl_cb_flags))
+		return;
+	spin_lock(&clp->cl_lock);
+	/*
+	 * Only serialized callback code is allowed to clear these
+	 * flags; main nfsd code can only set them:
+	 */
+	BUG_ON(!clp->cl_cb_flags);
+	clear_bit(NFSD4_CLIENT_CB_UPDATE, &clp->cl_cb_flags);
+	memcpy(&conn, &cb->cb_clp->cl_cb_conn, sizeof(struct nfs4_cb_conn));
+	spin_unlock(&clp->cl_lock);
+
+	err = setup_callback_client(clp, &conn);
+	if (err)
+		warn_no_callback_path(clp, err);
 }
 
 void nfsd4_do_callback_rpc(struct work_struct *w)
 {
-	/* XXX: for now, just send off delegation recall. */
-	/* In future, generalize to handle any sort of callback. */
-	struct nfsd4_callback *c = container_of(w, struct nfsd4_callback, cb_work);
-	struct nfs4_delegation *dp = container_of(c, struct nfs4_delegation, dl_recall);
+	struct nfsd4_callback *cb = container_of(w, struct nfsd4_callback, cb_work);
+	struct nfs4_client *clp = cb->cb_clp;
+	struct rpc_clnt *clnt;
 
-	_nfsd4_cb_recall(dp);
-}
+	if (clp->cl_cb_flags)
+		nfsd4_process_cb_update(cb);
 
+	clnt = clp->cl_cb_client;
+	if (!clnt) {
+		/* Callback channel broken, or client killed; give up: */
+		nfsd4_release_cb(cb);
+		return;
+	}
+	rpc_call_async(clnt, &cb->cb_msg, RPC_TASK_SOFT | RPC_TASK_SOFTCONN,
+			cb->cb_ops, cb);
+}
 
 void nfsd4_cb_recall(struct nfs4_delegation *dp)
 {
+	struct nfsd4_callback *cb = &dp->dl_recall;
+
+	dp->dl_retries = 1;
+	cb->cb_op = dp;
+	cb->cb_clp = dp->dl_client;
+	cb->cb_msg.rpc_proc = &nfs4_cb_procedures[NFSPROC4_CLNT_CB_RECALL];
+	cb->cb_msg.rpc_argp = cb;
+	cb->cb_msg.rpc_resp = cb;
+	cb->cb_msg.rpc_cred = callback_cred;
+
+	cb->cb_ops = &nfsd4_cb_recall_ops;
+	dp->dl_retries = 1;
+
 	queue_work(callback_wq, &dp->dl_recall.cb_work);
 }
diff --git a/fs/nfsd/nfs4idmap.c b/fs/nfsd/nfs4idmap.c
index c78dbf4..f0695e8 100644
--- a/fs/nfsd/nfs4idmap.c
+++ b/fs/nfsd/nfs4idmap.c
@@ -482,109 +482,26 @@ nfsd_idmap_shutdown(void)
 	cache_unregister(&nametoid_cache);
 }
 
-/*
- * Deferred request handling
- */
-
-struct idmap_defer_req {
-       struct cache_req		req;
-       struct cache_deferred_req deferred_req;
-       wait_queue_head_t	waitq;
-       atomic_t			count;
-};
-
-static inline void
-put_mdr(struct idmap_defer_req *mdr)
-{
-	if (atomic_dec_and_test(&mdr->count))
-		kfree(mdr);
-}
-
-static inline void
-get_mdr(struct idmap_defer_req *mdr)
-{
-	atomic_inc(&mdr->count);
-}
-
-static void
-idmap_revisit(struct cache_deferred_req *dreq, int toomany)
-{
-	struct idmap_defer_req *mdr =
-		container_of(dreq, struct idmap_defer_req, deferred_req);
-
-	wake_up(&mdr->waitq);
-	put_mdr(mdr);
-}
-
-static struct cache_deferred_req *
-idmap_defer(struct cache_req *req)
-{
-	struct idmap_defer_req *mdr =
-		container_of(req, struct idmap_defer_req, req);
-
-	mdr->deferred_req.revisit = idmap_revisit;
-	get_mdr(mdr);
-	return (&mdr->deferred_req);
-}
-
-static inline int
-do_idmap_lookup(struct ent *(*lookup_fn)(struct ent *), struct ent *key,
-		struct cache_detail *detail, struct ent **item,
-		struct idmap_defer_req *mdr)
-{
-	*item = lookup_fn(key);
-	if (!*item)
-		return -ENOMEM;
-	return cache_check(detail, &(*item)->h, &mdr->req);
-}
-
-static inline int
-do_idmap_lookup_nowait(struct ent *(*lookup_fn)(struct ent *),
-			struct ent *key, struct cache_detail *detail,
-			struct ent **item)
-{
-	int ret = -ENOMEM;
-
-	*item = lookup_fn(key);
-	if (!*item)
-		goto out_err;
-	ret = -ETIMEDOUT;
-	if (!test_bit(CACHE_VALID, &(*item)->h.flags)
-			|| (*item)->h.expiry_time < get_seconds()
-			|| detail->flush_time > (*item)->h.last_refresh)
-		goto out_put;
-	ret = -ENOENT;
-	if (test_bit(CACHE_NEGATIVE, &(*item)->h.flags))
-		goto out_put;
-	return 0;
-out_put:
-	cache_put(&(*item)->h, detail);
-out_err:
-	*item = NULL;
-	return ret;
-}
-
 static int
 idmap_lookup(struct svc_rqst *rqstp,
 		struct ent *(*lookup_fn)(struct ent *), struct ent *key,
 		struct cache_detail *detail, struct ent **item)
 {
-	struct idmap_defer_req *mdr;
 	int ret;
 
-	mdr = kzalloc(sizeof(*mdr), GFP_KERNEL);
-	if (!mdr)
+	*item = lookup_fn(key);
+	if (!*item)
 		return -ENOMEM;
-	atomic_set(&mdr->count, 1);
-	init_waitqueue_head(&mdr->waitq);
-	mdr->req.defer = idmap_defer;
-	ret = do_idmap_lookup(lookup_fn, key, detail, item, mdr);
-	if (ret == -EAGAIN) {
-		wait_event_interruptible_timeout(mdr->waitq,
-			test_bit(CACHE_VALID, &(*item)->h.flags), 1 * HZ);
-		ret = do_idmap_lookup_nowait(lookup_fn, key, detail, item);
+ retry:
+	ret = cache_check(detail, &(*item)->h, &rqstp->rq_chandle);
+
+	if (ret == -ETIMEDOUT) {
+		struct ent *prev_item = *item;
+		*item = lookup_fn(key);
+		if (*item != prev_item)
+			goto retry;
+		cache_put(&(*item)->h, detail);
 	}
-	put_mdr(mdr);
 	return ret;
 }
 
diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c
index 59ec449..0cdfd02 100644
--- a/fs/nfsd/nfs4proc.c
+++ b/fs/nfsd/nfs4proc.c
@@ -1031,8 +1031,11 @@ nfsd4_proc_compound(struct svc_rqst *rqstp,
 	resp->cstate.session = NULL;
 	fh_init(&resp->cstate.current_fh, NFS4_FHSIZE);
 	fh_init(&resp->cstate.save_fh, NFS4_FHSIZE);
-	/* Use the deferral mechanism only for NFSv4.0 compounds */
-	rqstp->rq_usedeferral = (args->minorversion == 0);
+	/*
+	 * Don't use the deferral mechanism for NFSv4; compounds make it
+	 * too hard to avoid non-idempotency problems.
+	 */
+	rqstp->rq_usedeferral = 0;
 
 	/*
 	 * According to RFC3010, this takes precedence over all other errors.
diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c
index 4a27347..f1e5ec6 100644
--- a/fs/nfsd/nfs4state.c
+++ b/fs/nfsd/nfs4state.c
@@ -33,7 +33,7 @@
 */
 
 #include <linux/file.h>
-#include <linux/smp_lock.h>
+#include <linux/fs.h>
 #include <linux/slab.h>
 #include <linux/namei.h>
 #include <linux/swap.h>
@@ -51,7 +51,6 @@ static time_t boot_time;
 static u32 current_ownerid = 1;
 static u32 current_fileid = 1;
 static u32 current_delegid = 1;
-static u32 nfs4_init;
 static stateid_t zerostateid;             /* bits all 0 */
 static stateid_t onestateid;              /* bits all 1 */
 static u64 current_sessionid = 1;
@@ -163,14 +162,60 @@ static struct list_head	ownerstr_hashtbl[OWNER_HASH_SIZE];
 static struct list_head file_hashtbl[FILE_HASH_SIZE];
 static struct list_head stateid_hashtbl[STATEID_HASH_SIZE];
 
+static void __nfs4_file_get_access(struct nfs4_file *fp, int oflag)
+{
+	BUG_ON(!(fp->fi_fds[oflag] || fp->fi_fds[O_RDWR]));
+	atomic_inc(&fp->fi_access[oflag]);
+}
+
+static void nfs4_file_get_access(struct nfs4_file *fp, int oflag)
+{
+	if (oflag == O_RDWR) {
+		__nfs4_file_get_access(fp, O_RDONLY);
+		__nfs4_file_get_access(fp, O_WRONLY);
+	} else
+		__nfs4_file_get_access(fp, oflag);
+}
+
+static void nfs4_file_put_fd(struct nfs4_file *fp, int oflag)
+{
+	if (fp->fi_fds[oflag]) {
+		fput(fp->fi_fds[oflag]);
+		fp->fi_fds[oflag] = NULL;
+	}
+}
+
+static void __nfs4_file_put_access(struct nfs4_file *fp, int oflag)
+{
+	if (atomic_dec_and_test(&fp->fi_access[oflag])) {
+		nfs4_file_put_fd(fp, O_RDWR);
+		nfs4_file_put_fd(fp, oflag);
+	}
+}
+
+static void nfs4_file_put_access(struct nfs4_file *fp, int oflag)
+{
+	if (oflag == O_RDWR) {
+		__nfs4_file_put_access(fp, O_RDONLY);
+		__nfs4_file_put_access(fp, O_WRONLY);
+	} else
+		__nfs4_file_put_access(fp, oflag);
+}
+
 static struct nfs4_delegation *
 alloc_init_deleg(struct nfs4_client *clp, struct nfs4_stateid *stp, struct svc_fh *current_fh, u32 type)
 {
 	struct nfs4_delegation *dp;
 	struct nfs4_file *fp = stp->st_file;
-	struct nfs4_cb_conn *cb = &stp->st_stateowner->so_client->cl_cb_conn;
 
 	dprintk("NFSD alloc_init_deleg\n");
+	/*
+	 * Major work on the lease subsystem (for example, to support
+	 * calbacks on stat) will be required before we can support
+	 * write delegations properly.
+	 */
+	if (type != NFS4_OPEN_DELEGATE_READ)
+		return NULL;
 	if (fp->fi_had_conflict)
 		return NULL;
 	if (num_delegations > max_delegations)
@@ -185,11 +230,9 @@ alloc_init_deleg(struct nfs4_client *clp, struct nfs4_stateid *stp, struct svc_f
 	dp->dl_client = clp;
 	get_nfs4_file(fp);
 	dp->dl_file = fp;
+	nfs4_file_get_access(fp, O_RDONLY);
 	dp->dl_flock = NULL;
-	get_file(stp->st_vfs_file);
-	dp->dl_vfs_file = stp->st_vfs_file;
 	dp->dl_type = type;
-	dp->dl_ident = cb->cb_ident;
 	dp->dl_stateid.si_boot = boot_time;
 	dp->dl_stateid.si_stateownerid = current_delegid++;
 	dp->dl_stateid.si_fileid = 0;
@@ -222,15 +265,12 @@ nfs4_put_delegation(struct nfs4_delegation *dp)
 static void
 nfs4_close_delegation(struct nfs4_delegation *dp)
 {
-	struct file *filp = dp->dl_vfs_file;
+	struct file *filp = find_readable_file(dp->dl_file);
 
 	dprintk("NFSD: close_delegation dp %p\n",dp);
-	dp->dl_vfs_file = NULL;
-	/* The following nfsd_close may not actually close the file,
-	 * but we want to remove the lease in any case. */
 	if (dp->dl_flock)
 		vfs_setlease(filp, F_UNLCK, &dp->dl_flock);
-	nfsd_close(filp);
+	nfs4_file_put_access(dp->dl_file, O_RDONLY);
 }
 
 /* Called under the state lock. */
@@ -302,8 +342,12 @@ static void free_generic_stateid(struct nfs4_stateid *stp)
 
 static void release_lock_stateid(struct nfs4_stateid *stp)
 {
+	struct file *file;
+
 	unhash_generic_stateid(stp);
-	locks_remove_posix(stp->st_vfs_file, (fl_owner_t)stp->st_stateowner);
+	file = find_any_file(stp->st_file);
+	if (file)
+		locks_remove_posix(file, (fl_owner_t)stp->st_stateowner);
 	free_generic_stateid(stp);
 }
 
@@ -341,11 +385,85 @@ release_stateid_lockowners(struct nfs4_stateid *open_stp)
 	}
 }
 
+/*
+ * We store the NONE, READ, WRITE, and BOTH bits separately in the
+ * st_{access,deny}_bmap field of the stateid, in order to track not
+ * only what share bits are currently in force, but also what
+ * combinations of share bits previous opens have used.  This allows us
+ * to enforce the recommendation of rfc 3530 14.2.19 that the server
+ * return an error if the client attempt to downgrade to a combination
+ * of share bits not explicable by closing some of its previous opens.
+ *
+ * XXX: This enforcement is actually incomplete, since we don't keep
+ * track of access/deny bit combinations; so, e.g., we allow:
+ *
+ *	OPEN allow read, deny write
+ *	OPEN allow both, deny none
+ *	DOWNGRADE allow read, deny none
+ *
+ * which we should reject.
+ */
+static void
+set_access(unsigned int *access, unsigned long bmap) {
+	int i;
+
+	*access = 0;
+	for (i = 1; i < 4; i++) {
+		if (test_bit(i, &bmap))
+			*access |= i;
+	}
+}
+
+static void
+set_deny(unsigned int *deny, unsigned long bmap) {
+	int i;
+
+	*deny = 0;
+	for (i = 0; i < 4; i++) {
+		if (test_bit(i, &bmap))
+			*deny |= i ;
+	}
+}
+
+static int
+test_share(struct nfs4_stateid *stp, struct nfsd4_open *open) {
+	unsigned int access, deny;
+
+	set_access(&access, stp->st_access_bmap);
+	set_deny(&deny, stp->st_deny_bmap);
+	if ((access & open->op_share_deny) || (deny & open->op_share_access))
+		return 0;
+	return 1;
+}
+
+static int nfs4_access_to_omode(u32 access)
+{
+	switch (access & NFS4_SHARE_ACCESS_BOTH) {
+	case NFS4_SHARE_ACCESS_READ:
+		return O_RDONLY;
+	case NFS4_SHARE_ACCESS_WRITE:
+		return O_WRONLY;
+	case NFS4_SHARE_ACCESS_BOTH:
+		return O_RDWR;
+	}
+	BUG();
+}
+
+static int nfs4_access_bmap_to_omode(struct nfs4_stateid *stp)
+{
+	unsigned int access;
+
+	set_access(&access, stp->st_access_bmap);
+	return nfs4_access_to_omode(access);
+}
+
 static void release_open_stateid(struct nfs4_stateid *stp)
 {
+	int oflag = nfs4_access_bmap_to_omode(stp);
+
 	unhash_generic_stateid(stp);
 	release_stateid_lockowners(stp);
-	nfsd_close(stp->st_vfs_file);
+	nfs4_file_put_access(stp->st_file, oflag);
 	free_generic_stateid(stp);
 }
 
@@ -415,171 +533,258 @@ gen_sessionid(struct nfsd4_session *ses)
  */
 #define NFSD_MIN_HDR_SEQ_SZ  (24 + 12 + 44)
 
+static void
+free_session_slots(struct nfsd4_session *ses)
+{
+	int i;
+
+	for (i = 0; i < ses->se_fchannel.maxreqs; i++)
+		kfree(ses->se_slots[i]);
+}
+
 /*
- * Give the client the number of ca_maxresponsesize_cached slots it
- * requests, of size bounded by NFSD_SLOT_CACHE_SIZE,
- * NFSD_MAX_MEM_PER_SESSION, and nfsd_drc_max_mem. Do not allow more
- * than NFSD_MAX_SLOTS_PER_SESSION.
- *
- * If we run out of reserved DRC memory we should (up to a point)
+ * We don't actually need to cache the rpc and session headers, so we
+ * can allocate a little less for each slot:
+ */
+static inline int slot_bytes(struct nfsd4_channel_attrs *ca)
+{
+	return ca->maxresp_cached - NFSD_MIN_HDR_SEQ_SZ;
+}
+
+static int nfsd4_sanitize_slot_size(u32 size)
+{
+	size -= NFSD_MIN_HDR_SEQ_SZ; /* We don't cache the rpc header */
+	size = min_t(u32, size, NFSD_SLOT_CACHE_SIZE);
+
+	return size;
+}
+
+/*
+ * XXX: If we run out of reserved DRC memory we could (up to a point)
  * re-negotiate active sessions and reduce their slot usage to make
  * rooom for new connections. For now we just fail the create session.
  */
-static int set_forechannel_drc_size(struct nfsd4_channel_attrs *fchan)
+static int nfsd4_get_drc_mem(int slotsize, u32 num)
 {
-	int mem, size = fchan->maxresp_cached;
+	int avail;
 
-	if (fchan->maxreqs < 1)
-		return nfserr_inval;
+	num = min_t(u32, num, NFSD_MAX_SLOTS_PER_SESSION);
 
-	if (size < NFSD_MIN_HDR_SEQ_SZ)
-		size = NFSD_MIN_HDR_SEQ_SZ;
-	size -= NFSD_MIN_HDR_SEQ_SZ;
-	if (size > NFSD_SLOT_CACHE_SIZE)
-		size = NFSD_SLOT_CACHE_SIZE;
-
-	/* bound the maxreqs by NFSD_MAX_MEM_PER_SESSION */
-	mem = fchan->maxreqs * size;
-	if (mem > NFSD_MAX_MEM_PER_SESSION) {
-		fchan->maxreqs = NFSD_MAX_MEM_PER_SESSION / size;
-		if (fchan->maxreqs > NFSD_MAX_SLOTS_PER_SESSION)
-			fchan->maxreqs = NFSD_MAX_SLOTS_PER_SESSION;
-		mem = fchan->maxreqs * size;
-	}
+	spin_lock(&nfsd_drc_lock);
+	avail = min_t(int, NFSD_MAX_MEM_PER_SESSION,
+			nfsd_drc_max_mem - nfsd_drc_mem_used);
+	num = min_t(int, num, avail / slotsize);
+	nfsd_drc_mem_used += num * slotsize;
+	spin_unlock(&nfsd_drc_lock);
+
+	return num;
+}
 
+static void nfsd4_put_drc_mem(int slotsize, int num)
+{
 	spin_lock(&nfsd_drc_lock);
-	/* bound the total session drc memory ussage */
-	if (mem + nfsd_drc_mem_used > nfsd_drc_max_mem) {
-		fchan->maxreqs = (nfsd_drc_max_mem - nfsd_drc_mem_used) / size;
-		mem = fchan->maxreqs * size;
-	}
-	nfsd_drc_mem_used += mem;
+	nfsd_drc_mem_used -= slotsize * num;
 	spin_unlock(&nfsd_drc_lock);
+}
 
-	if (fchan->maxreqs == 0)
-		return nfserr_serverfault;
+static struct nfsd4_session *alloc_session(int slotsize, int numslots)
+{
+	struct nfsd4_session *new;
+	int mem, i;
 
-	fchan->maxresp_cached = size + NFSD_MIN_HDR_SEQ_SZ;
-	return 0;
+	BUILD_BUG_ON(NFSD_MAX_SLOTS_PER_SESSION * sizeof(struct nfsd4_slot *)
+			+ sizeof(struct nfsd4_session) > PAGE_SIZE);
+	mem = numslots * sizeof(struct nfsd4_slot *);
+
+	new = kzalloc(sizeof(*new) + mem, GFP_KERNEL);
+	if (!new)
+		return NULL;
+	/* allocate each struct nfsd4_slot and data cache in one piece */
+	for (i = 0; i < numslots; i++) {
+		mem = sizeof(struct nfsd4_slot) + slotsize;
+		new->se_slots[i] = kzalloc(mem, GFP_KERNEL);
+		if (!new->se_slots[i])
+			goto out_free;
+	}
+	return new;
+out_free:
+	while (i--)
+		kfree(new->se_slots[i]);
+	kfree(new);
+	return NULL;
 }
 
-/*
- * fchan holds the client values on input, and the server values on output
- * sv_max_mesg is the maximum payload plus one page for overhead.
- */
-static int init_forechannel_attrs(struct svc_rqst *rqstp,
-				  struct nfsd4_channel_attrs *session_fchan,
-				  struct nfsd4_channel_attrs *fchan)
+static void init_forechannel_attrs(struct nfsd4_channel_attrs *new, struct nfsd4_channel_attrs *req, int numslots, int slotsize)
 {
-	int status = 0;
-	__u32   maxcount = nfsd_serv->sv_max_mesg;
+	u32 maxrpc = nfsd_serv->sv_max_mesg;
 
-	/* headerpadsz set to zero in encode routine */
+	new->maxreqs = numslots;
+	new->maxresp_cached = slotsize + NFSD_MIN_HDR_SEQ_SZ;
+	new->maxreq_sz = min_t(u32, req->maxreq_sz, maxrpc);
+	new->maxresp_sz = min_t(u32, req->maxresp_sz, maxrpc);
+	new->maxops = min_t(u32, req->maxops, NFSD_MAX_OPS_PER_COMPOUND);
+}
 
-	/* Use the client's max request and max response size if possible */
-	if (fchan->maxreq_sz > maxcount)
-		fchan->maxreq_sz = maxcount;
-	session_fchan->maxreq_sz = fchan->maxreq_sz;
+static void free_conn(struct nfsd4_conn *c)
+{
+	svc_xprt_put(c->cn_xprt);
+	kfree(c);
+}
 
-	if (fchan->maxresp_sz > maxcount)
-		fchan->maxresp_sz = maxcount;
-	session_fchan->maxresp_sz = fchan->maxresp_sz;
+static void nfsd4_conn_lost(struct svc_xpt_user *u)
+{
+	struct nfsd4_conn *c = container_of(u, struct nfsd4_conn, cn_xpt_user);
+	struct nfs4_client *clp = c->cn_session->se_client;
 
-	/* Use the client's maxops if possible */
-	if (fchan->maxops > NFSD_MAX_OPS_PER_COMPOUND)
-		fchan->maxops = NFSD_MAX_OPS_PER_COMPOUND;
-	session_fchan->maxops = fchan->maxops;
+	spin_lock(&clp->cl_lock);
+	if (!list_empty(&c->cn_persession)) {
+		list_del(&c->cn_persession);
+		free_conn(c);
+	}
+	spin_unlock(&clp->cl_lock);
+}
 
-	/* FIXME: Error means no more DRC pages so the server should
-	 * recover pages from existing sessions. For now fail session
-	 * creation.
-	 */
-	status = set_forechannel_drc_size(fchan);
+static struct nfsd4_conn *alloc_conn(struct svc_rqst *rqstp, u32 flags)
+{
+	struct nfsd4_conn *conn;
 
-	session_fchan->maxresp_cached = fchan->maxresp_cached;
-	session_fchan->maxreqs = fchan->maxreqs;
+	conn = kmalloc(sizeof(struct nfsd4_conn), GFP_KERNEL);
+	if (!conn)
+		return NULL;
+	svc_xprt_get(rqstp->rq_xprt);
+	conn->cn_xprt = rqstp->rq_xprt;
+	conn->cn_flags = flags;
+	INIT_LIST_HEAD(&conn->cn_xpt_user.list);
+	return conn;
+}
 
-	dprintk("%s status %d\n", __func__, status);
-	return status;
+static void __nfsd4_hash_conn(struct nfsd4_conn *conn, struct nfsd4_session *ses)
+{
+	conn->cn_session = ses;
+	list_add(&conn->cn_persession, &ses->se_conns);
 }
 
-static void
-free_session_slots(struct nfsd4_session *ses)
+static void nfsd4_hash_conn(struct nfsd4_conn *conn, struct nfsd4_session *ses)
 {
-	int i;
+	struct nfs4_client *clp = ses->se_client;
 
-	for (i = 0; i < ses->se_fchannel.maxreqs; i++)
-		kfree(ses->se_slots[i]);
+	spin_lock(&clp->cl_lock);
+	__nfsd4_hash_conn(conn, ses);
+	spin_unlock(&clp->cl_lock);
 }
 
-/*
- * We don't actually need to cache the rpc and session headers, so we
- * can allocate a little less for each slot:
- */
-static inline int slot_bytes(struct nfsd4_channel_attrs *ca)
+static void nfsd4_register_conn(struct nfsd4_conn *conn)
 {
-	return ca->maxresp_cached - NFSD_MIN_HDR_SEQ_SZ;
+	conn->cn_xpt_user.callback = nfsd4_conn_lost;
+	register_xpt_user(conn->cn_xprt, &conn->cn_xpt_user);
 }
 
-static int
-alloc_init_session(struct svc_rqst *rqstp, struct nfs4_client *clp,
-		   struct nfsd4_create_session *cses)
+static __be32 nfsd4_new_conn(struct svc_rqst *rqstp, struct nfsd4_session *ses)
 {
-	struct nfsd4_session *new, tmp;
-	struct nfsd4_slot *sp;
-	int idx, slotsize, cachesize, i;
-	int status;
+	struct nfsd4_conn *conn;
+	u32 flags = NFS4_CDFC4_FORE;
 
-	memset(&tmp, 0, sizeof(tmp));
+	if (ses->se_flags & SESSION4_BACK_CHAN)
+		flags |= NFS4_CDFC4_BACK;
+	conn = alloc_conn(rqstp, flags);
+	if (!conn)
+		return nfserr_jukebox;
+	nfsd4_hash_conn(conn, ses);
+	nfsd4_register_conn(conn);
+	return nfs_ok;
+}
 
-	/* FIXME: For now, we just accept the client back channel attributes. */
-	tmp.se_bchannel = cses->back_channel;
-	status = init_forechannel_attrs(rqstp, &tmp.se_fchannel,
-					&cses->fore_channel);
-	if (status)
-		goto out;
+static void nfsd4_del_conns(struct nfsd4_session *s)
+{
+	struct nfs4_client *clp = s->se_client;
+	struct nfsd4_conn *c;
 
-	BUILD_BUG_ON(NFSD_MAX_SLOTS_PER_SESSION * sizeof(struct nfsd4_slot)
-		     + sizeof(struct nfsd4_session) > PAGE_SIZE);
+	spin_lock(&clp->cl_lock);
+	while (!list_empty(&s->se_conns)) {
+		c = list_first_entry(&s->se_conns, struct nfsd4_conn, cn_persession);
+		list_del_init(&c->cn_persession);
+		spin_unlock(&clp->cl_lock);
 
-	status = nfserr_serverfault;
-	/* allocate struct nfsd4_session and slot table pointers in one piece */
-	slotsize = tmp.se_fchannel.maxreqs * sizeof(struct nfsd4_slot *);
-	new = kzalloc(sizeof(*new) + slotsize, GFP_KERNEL);
-	if (!new)
-		goto out;
+		unregister_xpt_user(c->cn_xprt, &c->cn_xpt_user);
+		free_conn(c);
 
-	memcpy(new, &tmp, sizeof(*new));
+		spin_lock(&clp->cl_lock);
+	}
+	spin_unlock(&clp->cl_lock);
+}
 
-	/* allocate each struct nfsd4_slot and data cache in one piece */
-	cachesize = slot_bytes(&new->se_fchannel);
-	for (i = 0; i < new->se_fchannel.maxreqs; i++) {
-		sp = kzalloc(sizeof(*sp) + cachesize, GFP_KERNEL);
-		if (!sp)
-			goto out_free;
-		new->se_slots[i] = sp;
+void free_session(struct kref *kref)
+{
+	struct nfsd4_session *ses;
+	int mem;
+
+	ses = container_of(kref, struct nfsd4_session, se_ref);
+	nfsd4_del_conns(ses);
+	spin_lock(&nfsd_drc_lock);
+	mem = ses->se_fchannel.maxreqs * slot_bytes(&ses->se_fchannel);
+	nfsd_drc_mem_used -= mem;
+	spin_unlock(&nfsd_drc_lock);
+	free_session_slots(ses);
+	kfree(ses);
+}
+
+static struct nfsd4_session *alloc_init_session(struct svc_rqst *rqstp, struct nfs4_client *clp, struct nfsd4_create_session *cses)
+{
+	struct nfsd4_session *new;
+	struct nfsd4_channel_attrs *fchan = &cses->fore_channel;
+	int numslots, slotsize;
+	int status;
+	int idx;
+
+	/*
+	 * Note decreasing slot size below client's request may
+	 * make it difficult for client to function correctly, whereas
+	 * decreasing the number of slots will (just?) affect
+	 * performance.  When short on memory we therefore prefer to
+	 * decrease number of slots instead of their size.
+	 */
+	slotsize = nfsd4_sanitize_slot_size(fchan->maxresp_cached);
+	numslots = nfsd4_get_drc_mem(slotsize, fchan->maxreqs);
+
+	new = alloc_session(slotsize, numslots);
+	if (!new) {
+		nfsd4_put_drc_mem(slotsize, fchan->maxreqs);
+		return NULL;
 	}
+	init_forechannel_attrs(&new->se_fchannel, fchan, numslots, slotsize);
 
 	new->se_client = clp;
 	gen_sessionid(new);
-	idx = hash_sessionid(&new->se_sessionid);
-	memcpy(clp->cl_sessionid.data, new->se_sessionid.data,
-	       NFS4_MAX_SESSIONID_LEN);
 
+	INIT_LIST_HEAD(&new->se_conns);
+
+	new->se_cb_seq_nr = 1;
 	new->se_flags = cses->flags;
+	new->se_cb_prog = cses->callback_prog;
 	kref_init(&new->se_ref);
+	idx = hash_sessionid(&new->se_sessionid);
 	spin_lock(&client_lock);
 	list_add(&new->se_hash, &sessionid_hashtbl[idx]);
 	list_add(&new->se_perclnt, &clp->cl_sessions);
 	spin_unlock(&client_lock);
 
-	status = nfs_ok;
-out:
-	return status;
-out_free:
-	free_session_slots(new);
-	kfree(new);
-	goto out;
+	status = nfsd4_new_conn(rqstp, new);
+	/* whoops: benny points out, status is ignored! (err, or bogus) */
+	if (status) {
+		free_session(&new->se_ref);
+		return NULL;
+	}
+	if (!clp->cl_cb_session && (cses->flags & SESSION4_BACK_CHAN)) {
+		struct sockaddr *sa = svc_addr(rqstp);
+
+		clp->cl_cb_session = new;
+		clp->cl_cb_conn.cb_xprt = rqstp->rq_xprt;
+		svc_xprt_get(rqstp->rq_xprt);
+		rpc_copy_addr((struct sockaddr *)&clp->cl_cb_conn.cb_addr, sa);
+		clp->cl_cb_conn.cb_addrlen = svc_addr_len(sa);
+		nfsd4_probe_callback(clp);
+	}
+	return new;
 }
 
 /* caller must hold client_lock */
@@ -591,10 +796,8 @@ find_in_sessionid_hashtbl(struct nfs4_sessionid *sessionid)
 
 	dump_sessionid(__func__, sessionid);
 	idx = hash_sessionid(sessionid);
-	dprintk("%s: idx is %d\n", __func__, idx);
 	/* Search in the appropriate list */
 	list_for_each_entry(elem, &sessionid_hashtbl[idx], se_hash) {
-		dump_sessionid("list traversal", &elem->se_sessionid);
 		if (!memcmp(elem->se_sessionid.data, sessionid->data,
 			    NFS4_MAX_SESSIONID_LEN)) {
 			return elem;
@@ -613,21 +816,6 @@ unhash_session(struct nfsd4_session *ses)
 	list_del(&ses->se_perclnt);
 }
 
-void
-free_session(struct kref *kref)
-{
-	struct nfsd4_session *ses;
-	int mem;
-
-	ses = container_of(kref, struct nfsd4_session, se_ref);
-	spin_lock(&nfsd_drc_lock);
-	mem = ses->se_fchannel.maxreqs * slot_bytes(&ses->se_fchannel);
-	nfsd_drc_mem_used -= mem;
-	spin_unlock(&nfsd_drc_lock);
-	free_session_slots(ses);
-	kfree(ses);
-}
-
 /* must be called under the client_lock */
 static inline void
 renew_client_locked(struct nfs4_client *clp)
@@ -694,6 +882,13 @@ static struct nfs4_client *alloc_client(struct xdr_netobj name)
 static inline void
 free_client(struct nfs4_client *clp)
 {
+	while (!list_empty(&clp->cl_sessions)) {
+		struct nfsd4_session *ses;
+		ses = list_entry(clp->cl_sessions.next, struct nfsd4_session,
+				se_perclnt);
+		list_del(&ses->se_perclnt);
+		nfsd4_put_session(ses);
+	}
 	if (clp->cl_cred.cr_group_info)
 		put_group_info(clp->cl_cred.cr_group_info);
 	kfree(clp->cl_principal);
@@ -714,22 +909,18 @@ release_session_client(struct nfsd4_session *session)
 	} else
 		renew_client_locked(clp);
 	spin_unlock(&client_lock);
-	nfsd4_put_session(session);
 }
 
 /* must be called under the client_lock */
 static inline void
 unhash_client_locked(struct nfs4_client *clp)
 {
+	struct nfsd4_session *ses;
+
 	mark_client_expired(clp);
 	list_del(&clp->cl_lru);
-	while (!list_empty(&clp->cl_sessions)) {
-		struct nfsd4_session  *ses;
-		ses = list_entry(clp->cl_sessions.next, struct nfsd4_session,
-				 se_perclnt);
-		unhash_session(ses);
-		nfsd4_put_session(ses);
-	}
+	list_for_each_entry(ses, &clp->cl_sessions, se_perclnt)
+		list_del_init(&ses->se_hash);
 }
 
 static void
@@ -758,7 +949,7 @@ expire_client(struct nfs4_client *clp)
 		sop = list_entry(clp->cl_openowners.next, struct nfs4_stateowner, so_perclient);
 		release_openowner(sop);
 	}
-	nfsd4_set_callback_client(clp, NULL);
+	nfsd4_shutdown_callback(clp);
 	if (clp->cl_cb_conn.cb_xprt)
 		svc_xprt_put(clp->cl_cb_conn.cb_xprt);
 	list_del(&clp->cl_idhash);
@@ -843,6 +1034,8 @@ static struct nfs4_client *create_client(struct xdr_netobj name, char *recdir,
 	if (clp == NULL)
 		return NULL;
 
+	INIT_LIST_HEAD(&clp->cl_sessions);
+
 	princ = svc_gss_principal(rqstp);
 	if (princ) {
 		clp->cl_principal = kstrdup(princ, GFP_KERNEL);
@@ -859,8 +1052,9 @@ static struct nfs4_client *create_client(struct xdr_netobj name, char *recdir,
 	INIT_LIST_HEAD(&clp->cl_strhash);
 	INIT_LIST_HEAD(&clp->cl_openowners);
 	INIT_LIST_HEAD(&clp->cl_delegations);
-	INIT_LIST_HEAD(&clp->cl_sessions);
 	INIT_LIST_HEAD(&clp->cl_lru);
+	spin_lock_init(&clp->cl_lock);
+	INIT_WORK(&clp->cl_cb_null.cb_work, nfsd4_do_callback_rpc);
 	clp->cl_time = get_seconds();
 	clear_bit(0, &clp->cl_cb_slot_busy);
 	rpc_init_wait_queue(&clp->cl_cb_waitq, "Backchannel slot table");
@@ -869,7 +1063,7 @@ static struct nfs4_client *create_client(struct xdr_netobj name, char *recdir,
 	clp->cl_flavor = rqstp->rq_flavor;
 	copy_cred(&clp->cl_cred, &rqstp->rq_cred);
 	gen_confirm(clp);
-
+	clp->cl_cb_session = NULL;
 	return clp;
 }
 
@@ -981,7 +1175,7 @@ find_unconfirmed_client_by_str(const char *dname, unsigned int hashval,
 static void
 gen_callback(struct nfs4_client *clp, struct nfsd4_setclientid *se, u32 scopeid)
 {
-	struct nfs4_cb_conn *cb = &clp->cl_cb_conn;
+	struct nfs4_cb_conn *conn = &clp->cl_cb_conn;
 	unsigned short expected_family;
 
 	/* Currently, we only support tcp and tcp6 for the callback channel */
@@ -994,24 +1188,23 @@ gen_callback(struct nfs4_client *clp, struct nfsd4_setclientid *se, u32 scopeid)
 	else
 		goto out_err;
 
-	cb->cb_addrlen = rpc_uaddr2sockaddr(se->se_callback_addr_val,
+	conn->cb_addrlen = rpc_uaddr2sockaddr(se->se_callback_addr_val,
 					    se->se_callback_addr_len,
-					    (struct sockaddr *) &cb->cb_addr,
-					    sizeof(cb->cb_addr));
+					    (struct sockaddr *)&conn->cb_addr,
+					    sizeof(conn->cb_addr));
 
-	if (!cb->cb_addrlen || cb->cb_addr.ss_family != expected_family)
+	if (!conn->cb_addrlen || conn->cb_addr.ss_family != expected_family)
 		goto out_err;
 
-	if (cb->cb_addr.ss_family == AF_INET6)
-		((struct sockaddr_in6 *) &cb->cb_addr)->sin6_scope_id = scopeid;
+	if (conn->cb_addr.ss_family == AF_INET6)
+		((struct sockaddr_in6 *)&conn->cb_addr)->sin6_scope_id = scopeid;
 
-	cb->cb_minorversion = 0;
-	cb->cb_prog = se->se_callback_prog;
-	cb->cb_ident = se->se_callback_ident;
+	conn->cb_prog = se->se_callback_prog;
+	conn->cb_ident = se->se_callback_ident;
 	return;
 out_err:
-	cb->cb_addr.ss_family = AF_UNSPEC;
-	cb->cb_addrlen = 0;
+	conn->cb_addr.ss_family = AF_UNSPEC;
+	conn->cb_addrlen = 0;
 	dprintk(KERN_INFO "NFSD: this client (clientid %08x/%08x) "
 		"will not receive delegations\n",
 		clp->cl_clientid.cl_boot, clp->cl_clientid.cl_id);
@@ -1220,7 +1413,7 @@ out_new:
 	/* Normal case */
 	new = create_client(exid->clname, dname, rqstp, &verf);
 	if (new == NULL) {
-		status = nfserr_serverfault;
+		status = nfserr_jukebox;
 		goto out;
 	}
 
@@ -1298,7 +1491,9 @@ nfsd4_create_session(struct svc_rqst *rqstp,
 {
 	struct sockaddr *sa = svc_addr(rqstp);
 	struct nfs4_client *conf, *unconf;
+	struct nfsd4_session *new;
 	struct nfsd4_clid_slot *cs_slot = NULL;
+	bool confirm_me = false;
 	int status = 0;
 
 	nfs4_lock_state();
@@ -1321,7 +1516,6 @@ nfsd4_create_session(struct svc_rqst *rqstp,
 				cs_slot->sl_seqid, cr_ses->seqid);
 			goto out;
 		}
-		cs_slot->sl_seqid++;
 	} else if (unconf) {
 		if (!same_creds(&unconf->cl_cred, &rqstp->rq_cred) ||
 		    !rpc_cmp_addr(sa, (struct sockaddr *) &unconf->cl_addr)) {
@@ -1334,25 +1528,10 @@ nfsd4_create_session(struct svc_rqst *rqstp,
 		if (status) {
 			/* an unconfirmed replay returns misordered */
 			status = nfserr_seq_misordered;
-			goto out_cache;
+			goto out;
 		}
 
-		cs_slot->sl_seqid++; /* from 0 to 1 */
-		move_to_confirmed(unconf);
-
-		if (cr_ses->flags & SESSION4_BACK_CHAN) {
-			unconf->cl_cb_conn.cb_xprt = rqstp->rq_xprt;
-			svc_xprt_get(rqstp->rq_xprt);
-			rpc_copy_addr(
-				(struct sockaddr *)&unconf->cl_cb_conn.cb_addr,
-				sa);
-			unconf->cl_cb_conn.cb_addrlen = svc_addr_len(sa);
-			unconf->cl_cb_conn.cb_minorversion =
-				cstate->minorversion;
-			unconf->cl_cb_conn.cb_prog = cr_ses->callback_prog;
-			unconf->cl_cb_seq_nr = 1;
-			nfsd4_probe_callback(unconf, &unconf->cl_cb_conn);
-		}
+		confirm_me = true;
 		conf = unconf;
 	} else {
 		status = nfserr_stale_clientid;
@@ -1360,22 +1539,30 @@ nfsd4_create_session(struct svc_rqst *rqstp,
 	}
 
 	/*
+	 * XXX: we should probably set this at creation time, and check
+	 * for consistent minorversion use throughout:
+	 */
+	conf->cl_minorversion = 1;
+	/*
 	 * We do not support RDMA or persistent sessions
 	 */
 	cr_ses->flags &= ~SESSION4_PERSIST;
 	cr_ses->flags &= ~SESSION4_RDMA;
 
-	status = alloc_init_session(rqstp, conf, cr_ses);
-	if (status)
+	status = nfserr_jukebox;
+	new = alloc_init_session(rqstp, conf, cr_ses);
+	if (!new)
 		goto out;
-
-	memcpy(cr_ses->sessionid.data, conf->cl_sessionid.data,
+	status = nfs_ok;
+	memcpy(cr_ses->sessionid.data, new->se_sessionid.data,
 	       NFS4_MAX_SESSIONID_LEN);
+	cs_slot->sl_seqid++;
 	cr_ses->seqid = cs_slot->sl_seqid;
 
-out_cache:
 	/* cache solo and embedded create sessions under the state lock */
 	nfsd4_cache_create_session(cr_ses, cs_slot, status);
+	if (confirm_me)
+		move_to_confirmed(conf);
 out:
 	nfs4_unlock_state();
 	dprintk("%s returns %d\n", __func__, ntohl(status));
@@ -1429,8 +1616,11 @@ nfsd4_destroy_session(struct svc_rqst *r,
 
 	nfs4_lock_state();
 	/* wait for callbacks */
-	nfsd4_set_callback_client(ses->se_client, NULL);
+	nfsd4_shutdown_callback(ses->se_client);
 	nfs4_unlock_state();
+
+	nfsd4_del_conns(ses);
+
 	nfsd4_put_session(ses);
 	status = nfs_ok;
 out:
@@ -1438,6 +1628,36 @@ out:
 	return status;
 }
 
+static struct nfsd4_conn *__nfsd4_find_conn(struct svc_xprt *xpt, struct nfsd4_session *s)
+{
+	struct nfsd4_conn *c;
+
+	list_for_each_entry(c, &s->se_conns, cn_persession) {
+		if (c->cn_xprt == xpt) {
+			return c;
+		}
+	}
+	return NULL;
+}
+
+static void nfsd4_sequence_check_conn(struct nfsd4_conn *new, struct nfsd4_session *ses)
+{
+	struct nfs4_client *clp = ses->se_client;
+	struct nfsd4_conn *c;
+
+	spin_lock(&clp->cl_lock);
+	c = __nfsd4_find_conn(new->cn_xprt, ses);
+	if (c) {
+		spin_unlock(&clp->cl_lock);
+		free_conn(new);
+		return;
+	}
+	__nfsd4_hash_conn(new, ses);
+	spin_unlock(&clp->cl_lock);
+	nfsd4_register_conn(new);
+	return;
+}
+
 __be32
 nfsd4_sequence(struct svc_rqst *rqstp,
 	       struct nfsd4_compound_state *cstate,
@@ -1446,11 +1666,20 @@ nfsd4_sequence(struct svc_rqst *rqstp,
 	struct nfsd4_compoundres *resp = rqstp->rq_resp;
 	struct nfsd4_session *session;
 	struct nfsd4_slot *slot;
+	struct nfsd4_conn *conn;
 	int status;
 
 	if (resp->opcnt != 1)
 		return nfserr_sequence_pos;
 
+	/*
+	 * Will be either used or freed by nfsd4_sequence_check_conn
+	 * below.
+	 */
+	conn = alloc_conn(rqstp, NFS4_CDFC4_FORE);
+	if (!conn)
+		return nfserr_jukebox;
+
 	spin_lock(&client_lock);
 	status = nfserr_badsession;
 	session = find_in_sessionid_hashtbl(&seq->sessionid);
@@ -1482,6 +1711,9 @@ nfsd4_sequence(struct svc_rqst *rqstp,
 	if (status)
 		goto out;
 
+	nfsd4_sequence_check_conn(conn, session);
+	conn = NULL;
+
 	/* Success! bump slot seqid */
 	slot->sl_inuse = true;
 	slot->sl_seqid = seq->seqid;
@@ -1496,6 +1728,7 @@ out:
 		nfsd4_get_session(cstate->session);
 		atomic_inc(&session->se_client->cl_refcount);
 	}
+	kfree(conn);
 	spin_unlock(&client_lock);
 	dprintk("%s: return %d\n", __func__, ntohl(status));
 	return status;
@@ -1630,6 +1863,11 @@ nfsd4_setclientid(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 			goto out;
 		gen_clid(new);
 	}
+	/*
+	 * XXX: we should probably set this at creation time, and check
+	 * for consistent minorversion use throughout:
+	 */
+	new->cl_minorversion = 0;
 	gen_callback(new, setclid, rpc_get_scope_id(sa));
 	add_to_unconfirmed(new, strhashval);
 	setclid->se_clientid.cl_boot = new->cl_clientid.cl_boot;
@@ -1690,7 +1928,8 @@ nfsd4_setclientid_confirm(struct svc_rqst *rqstp,
 			status = nfserr_clid_inuse;
 		else {
 			atomic_set(&conf->cl_cb_set, 0);
-			nfsd4_probe_callback(conf, &unconf->cl_cb_conn);
+			nfsd4_change_callback(conf, &unconf->cl_cb_conn);
+			nfsd4_probe_callback(conf);
 			expire_client(unconf);
 			status = nfs_ok;
 
@@ -1724,7 +1963,7 @@ nfsd4_setclientid_confirm(struct svc_rqst *rqstp,
 			}
 			move_to_confirmed(unconf);
 			conf = unconf;
-			nfsd4_probe_callback(conf, &conf->cl_cb_conn);
+			nfsd4_probe_callback(conf);
 			status = nfs_ok;
 		}
 	} else if ((!conf || (conf && !same_verf(&conf->cl_confirm, &confirm)))
@@ -1760,6 +1999,8 @@ alloc_init_file(struct inode *ino)
 		fp->fi_inode = igrab(ino);
 		fp->fi_id = current_fileid++;
 		fp->fi_had_conflict = false;
+		memset(fp->fi_fds, 0, sizeof(fp->fi_fds));
+		memset(fp->fi_access, 0, sizeof(fp->fi_access));
 		spin_lock(&recall_lock);
 		list_add(&fp->fi_hash, &file_hashtbl[hashval]);
 		spin_unlock(&recall_lock);
@@ -1971,57 +2212,6 @@ static inline int deny_valid(u32 x)
 }
 
 /*
- * We store the NONE, READ, WRITE, and BOTH bits separately in the
- * st_{access,deny}_bmap field of the stateid, in order to track not
- * only what share bits are currently in force, but also what
- * combinations of share bits previous opens have used.  This allows us
- * to enforce the recommendation of rfc 3530 14.2.19 that the server
- * return an error if the client attempt to downgrade to a combination
- * of share bits not explicable by closing some of its previous opens.
- *
- * XXX: This enforcement is actually incomplete, since we don't keep
- * track of access/deny bit combinations; so, e.g., we allow:
- *
- *	OPEN allow read, deny write
- *	OPEN allow both, deny none
- *	DOWNGRADE allow read, deny none
- *
- * which we should reject.
- */
-static void
-set_access(unsigned int *access, unsigned long bmap) {
-	int i;
-
-	*access = 0;
-	for (i = 1; i < 4; i++) {
-		if (test_bit(i, &bmap))
-			*access |= i;
-	}
-}
-
-static void
-set_deny(unsigned int *deny, unsigned long bmap) {
-	int i;
-
-	*deny = 0;
-	for (i = 0; i < 4; i++) {
-		if (test_bit(i, &bmap))
-			*deny |= i ;
-	}
-}
-
-static int
-test_share(struct nfs4_stateid *stp, struct nfsd4_open *open) {
-	unsigned int access, deny;
-
-	set_access(&access, stp->st_access_bmap);
-	set_deny(&deny, stp->st_deny_bmap);
-	if ((access & open->op_share_deny) || (deny & open->op_share_access))
-		return 0;
-	return 1;
-}
-
-/*
  * Called to check deny when READ with all zero stateid or
  * WRITE with all zero or all one stateid
  */
@@ -2052,14 +2242,12 @@ out:
 }
 
 static inline void
-nfs4_file_downgrade(struct file *filp, unsigned int share_access)
+nfs4_file_downgrade(struct nfs4_file *fp, unsigned int share_access)
 {
-	if (share_access & NFS4_SHARE_ACCESS_WRITE) {
-		drop_file_write_access(filp);
-		spin_lock(&filp->f_lock);
-		filp->f_mode = (filp->f_mode | FMODE_READ) & ~FMODE_WRITE;
-		spin_unlock(&filp->f_lock);
-	}
+	if (share_access & NFS4_SHARE_ACCESS_WRITE)
+		nfs4_file_put_access(fp, O_WRONLY);
+	if (share_access & NFS4_SHARE_ACCESS_READ)
+		nfs4_file_put_access(fp, O_RDONLY);
 }
 
 /*
@@ -2122,22 +2310,6 @@ void nfsd_release_deleg_cb(struct file_lock *fl)
 }
 
 /*
- * Set the delegation file_lock back pointer.
- *
- * Called from setlease() with lock_kernel() held.
- */
-static
-void nfsd_copy_lock_deleg_cb(struct file_lock *new, struct file_lock *fl)
-{
-	struct nfs4_delegation *dp = (struct nfs4_delegation *)new->fl_owner;
-
-	dprintk("NFSD: nfsd_copy_lock_deleg_cb: new fl %p dp %p\n", new, dp);
-	if (!dp)
-		return;
-	dp->dl_flock = new;
-}
-
-/*
  * Called from setlease() with lock_kernel() held
  */
 static
@@ -2167,7 +2339,6 @@ int nfsd_change_deleg_cb(struct file_lock **onlist, int arg)
 static const struct lock_manager_operations nfsd_lease_mng_ops = {
 	.fl_break = nfsd_break_deleg_cb,
 	.fl_release_private = nfsd_release_deleg_cb,
-	.fl_copy_lock = nfsd_copy_lock_deleg_cb,
 	.fl_mylease = nfsd_same_client_deleg_cb,
 	.fl_change = nfsd_change_deleg_cb,
 };
@@ -2255,6 +2426,13 @@ find_delegation_file(struct nfs4_file *fp, stateid_t *stid)
 	return NULL;
 }
 
+int share_access_to_flags(u32 share_access)
+{
+	share_access &= ~NFS4_SHARE_WANT_MASK;
+
+	return share_access == NFS4_SHARE_ACCESS_READ ? RD_STATE : WR_STATE;
+}
+
 static __be32
 nfs4_check_deleg(struct nfs4_file *fp, struct nfsd4_open *open,
 		struct nfs4_delegation **dp)
@@ -2265,8 +2443,7 @@ nfs4_check_deleg(struct nfs4_file *fp, struct nfsd4_open *open,
 	*dp = find_delegation_file(fp, &open->op_delegate_stateid);
 	if (*dp == NULL)
 		goto out;
-	flags = open->op_share_access == NFS4_SHARE_ACCESS_READ ?
-						RD_STATE : WR_STATE;
+	flags = share_access_to_flags(open->op_share_access);
 	status = nfs4_check_delegmode(*dp, flags);
 	if (status)
 		*dp = NULL;
@@ -2308,30 +2485,53 @@ nfs4_alloc_stateid(void)
 	return kmem_cache_alloc(stateid_slab, GFP_KERNEL);
 }
 
+static inline int nfs4_access_to_access(u32 nfs4_access)
+{
+	int flags = 0;
+
+	if (nfs4_access & NFS4_SHARE_ACCESS_READ)
+		flags |= NFSD_MAY_READ;
+	if (nfs4_access & NFS4_SHARE_ACCESS_WRITE)
+		flags |= NFSD_MAY_WRITE;
+	return flags;
+}
+
+static __be32 nfs4_get_vfs_file(struct svc_rqst *rqstp, struct nfs4_file
+*fp, struct svc_fh *cur_fh, u32 nfs4_access)
+{
+	__be32 status;
+	int oflag = nfs4_access_to_omode(nfs4_access);
+	int access = nfs4_access_to_access(nfs4_access);
+
+	if (!fp->fi_fds[oflag]) {
+		status = nfsd_open(rqstp, cur_fh, S_IFREG, access,
+			&fp->fi_fds[oflag]);
+		if (status == nfserr_dropit)
+			status = nfserr_jukebox;
+		if (status)
+			return status;
+	}
+	nfs4_file_get_access(fp, oflag);
+
+	return nfs_ok;
+}
+
 static __be32
 nfs4_new_open(struct svc_rqst *rqstp, struct nfs4_stateid **stpp,
-		struct nfs4_delegation *dp,
-		struct svc_fh *cur_fh, int flags)
+		struct nfs4_file *fp, struct svc_fh *cur_fh,
+		struct nfsd4_open *open)
 {
 	struct nfs4_stateid *stp;
+	__be32 status;
 
 	stp = nfs4_alloc_stateid();
 	if (stp == NULL)
 		return nfserr_resource;
 
-	if (dp) {
-		get_file(dp->dl_vfs_file);
-		stp->st_vfs_file = dp->dl_vfs_file;
-	} else {
-		__be32 status;
-		status = nfsd_open(rqstp, cur_fh, S_IFREG, flags,
-				&stp->st_vfs_file);
-		if (status) {
-			if (status == nfserr_dropit)
-				status = nfserr_jukebox;
-			kmem_cache_free(stateid_slab, stp);
-			return status;
-		}
+	status = nfs4_get_vfs_file(rqstp, fp, cur_fh, open->op_share_access);
+	if (status) {
+		kmem_cache_free(stateid_slab, stp);
+		return status;
 	}
 	*stpp = stp;
 	return 0;
@@ -2353,35 +2553,28 @@ nfsd4_truncate(struct svc_rqst *rqstp, struct svc_fh *fh,
 }
 
 static __be32
-nfs4_upgrade_open(struct svc_rqst *rqstp, struct svc_fh *cur_fh, struct nfs4_stateid *stp, struct nfsd4_open *open)
+nfs4_upgrade_open(struct svc_rqst *rqstp, struct nfs4_file *fp, struct svc_fh *cur_fh, struct nfs4_stateid *stp, struct nfsd4_open *open)
 {
-	struct file *filp = stp->st_vfs_file;
-	struct inode *inode = filp->f_path.dentry->d_inode;
-	unsigned int share_access, new_writer;
+	u32 op_share_access = open->op_share_access & ~NFS4_SHARE_WANT_MASK;
+	bool new_access;
 	__be32 status;
 
-	set_access(&share_access, stp->st_access_bmap);
-	new_writer = (~share_access) & open->op_share_access
-			& NFS4_SHARE_ACCESS_WRITE;
-
-	if (new_writer) {
-		int err = get_write_access(inode);
-		if (err)
-			return nfserrno(err);
-		err = mnt_want_write(cur_fh->fh_export->ex_path.mnt);
-		if (err)
-			return nfserrno(err);
-		file_take_write(filp);
+	new_access = !test_bit(op_share_access, &stp->st_access_bmap);
+	if (new_access) {
+		status = nfs4_get_vfs_file(rqstp, fp, cur_fh, op_share_access);
+		if (status)
+			return status;
 	}
 	status = nfsd4_truncate(rqstp, cur_fh, open);
 	if (status) {
-		if (new_writer)
-			put_write_access(inode);
+		if (new_access) {
+			int oflag = nfs4_access_to_omode(new_access);
+			nfs4_file_put_access(fp, oflag);
+		}
 		return status;
 	}
 	/* remember the open */
-	filp->f_mode |= open->op_share_access;
-	__set_bit(open->op_share_access, &stp->st_access_bmap);
+	__set_bit(op_share_access, &stp->st_access_bmap);
 	__set_bit(open->op_share_deny, &stp->st_deny_bmap);
 
 	return nfs_ok;
@@ -2404,7 +2597,7 @@ nfs4_open_delegation(struct svc_fh *fh, struct nfsd4_open *open, struct nfs4_sta
 	struct nfs4_delegation *dp;
 	struct nfs4_stateowner *sop = stp->st_stateowner;
 	int cb_up = atomic_read(&sop->so_client->cl_cb_set);
-	struct file_lock fl, *flp = &fl;
+	struct file_lock *fl;
 	int status, flag = 0;
 
 	flag = NFS4_OPEN_DELEGATE_NONE;
@@ -2438,20 +2631,28 @@ nfs4_open_delegation(struct svc_fh *fh, struct nfsd4_open *open, struct nfs4_sta
 		flag = NFS4_OPEN_DELEGATE_NONE;
 		goto out;
 	}
-	locks_init_lock(&fl);
-	fl.fl_lmops = &nfsd_lease_mng_ops;
-	fl.fl_flags = FL_LEASE;
-	fl.fl_type = flag == NFS4_OPEN_DELEGATE_READ? F_RDLCK: F_WRLCK;
-	fl.fl_end = OFFSET_MAX;
-	fl.fl_owner =  (fl_owner_t)dp;
-	fl.fl_file = stp->st_vfs_file;
-	fl.fl_pid = current->tgid;
+	status = -ENOMEM;
+	fl = locks_alloc_lock();
+	if (!fl)
+		goto out;
+	locks_init_lock(fl);
+	fl->fl_lmops = &nfsd_lease_mng_ops;
+	fl->fl_flags = FL_LEASE;
+	fl->fl_type = flag == NFS4_OPEN_DELEGATE_READ? F_RDLCK: F_WRLCK;
+	fl->fl_end = OFFSET_MAX;
+	fl->fl_owner =  (fl_owner_t)dp;
+	fl->fl_file = find_readable_file(stp->st_file);
+	BUG_ON(!fl->fl_file);
+	fl->fl_pid = current->tgid;
+	dp->dl_flock = fl;
 
 	/* vfs_setlease checks to see if delegation should be handed out.
 	 * the lock_manager callbacks fl_mylease and fl_change are used
 	 */
-	if ((status = vfs_setlease(stp->st_vfs_file, fl.fl_type, &flp))) {
+	if ((status = vfs_setlease(fl->fl_file, fl->fl_type, &fl))) {
 		dprintk("NFSD: setlease failed [%d], no delegation\n", status);
+		dp->dl_flock = NULL;
+		locks_free_lock(fl);
 		unhash_delegation(dp);
 		flag = NFS4_OPEN_DELEGATE_NONE;
 		goto out;
@@ -2514,18 +2715,12 @@ nfsd4_process_open2(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct nf
 	 */
 	if (stp) {
 		/* Stateid was found, this is an OPEN upgrade */
-		status = nfs4_upgrade_open(rqstp, current_fh, stp, open);
+		status = nfs4_upgrade_open(rqstp, fp, current_fh, stp, open);
 		if (status)
 			goto out;
 		update_stateid(&stp->st_stateid);
 	} else {
-		/* Stateid was not found, this is a new OPEN */
-		int flags = 0;
-		if (open->op_share_access & NFS4_SHARE_ACCESS_READ)
-			flags |= NFSD_MAY_READ;
-		if (open->op_share_access & NFS4_SHARE_ACCESS_WRITE)
-			flags |= NFSD_MAY_WRITE;
-		status = nfs4_new_open(rqstp, &stp, dp, current_fh, flags);
+		status = nfs4_new_open(rqstp, &stp, fp, current_fh, open);
 		if (status)
 			goto out;
 		init_stateid(stp, fp, open);
@@ -2727,7 +2922,7 @@ search_close_lru(u32 st_id, int flags)
 static inline int
 nfs4_check_fh(struct svc_fh *fhp, struct nfs4_stateid *stp)
 {
-	return fhp->fh_dentry->d_inode != stp->st_vfs_file->f_path.dentry->d_inode;
+	return fhp->fh_dentry->d_inode != stp->st_file->fi_inode;
 }
 
 static int
@@ -2760,6 +2955,9 @@ __be32 nfs4_check_openmode(struct nfs4_stateid *stp, int flags)
 {
         __be32 status = nfserr_openmode;
 
+	/* For lock stateid's, we test the parent open, not the lock: */
+	if (stp->st_openstp)
+		stp = stp->st_openstp;
 	if ((flags & WR_STATE) && (!access_permit_write(stp->st_access_bmap)))
                 goto out;
 	if ((flags & RD_STATE) && (!access_permit_read(stp->st_access_bmap)))
@@ -2858,7 +3056,11 @@ nfs4_preprocess_stateid_op(struct nfsd4_compound_state *cstate,
 	if (STALE_STATEID(stateid)) 
 		goto out;
 
-	status = nfserr_bad_stateid;
+	/*
+	 * We assume that any stateid that has the current boot time,
+	 * but that we can't find, is expired:
+	 */
+	status = nfserr_expired;
 	if (is_delegation_stateid(stateid)) {
 		dp = find_delegation_stateid(ino, stateid);
 		if (!dp)
@@ -2872,11 +3074,13 @@ nfs4_preprocess_stateid_op(struct nfsd4_compound_state *cstate,
 			goto out;
 		renew_client(dp->dl_client);
 		if (filpp)
-			*filpp = dp->dl_vfs_file;
+			*filpp = find_readable_file(dp->dl_file);
+		BUG_ON(!*filpp);
 	} else { /* open or lock stateid */
 		stp = find_stateid(stateid, flags);
 		if (!stp)
 			goto out;
+		status = nfserr_bad_stateid;
 		if (nfs4_check_fh(current_fh, stp))
 			goto out;
 		if (!stp->st_stateowner->so_confirmed)
@@ -2889,8 +3093,12 @@ nfs4_preprocess_stateid_op(struct nfsd4_compound_state *cstate,
 		if (status)
 			goto out;
 		renew_client(stp->st_stateowner->so_client);
-		if (filpp)
-			*filpp = stp->st_vfs_file;
+		if (filpp) {
+			if (flags & RD_STATE)
+				*filpp = find_readable_file(stp->st_file);
+			else
+				*filpp = find_writeable_file(stp->st_file);
+		}
 	}
 	status = nfs_ok;
 out:
@@ -2947,8 +3155,9 @@ nfs4_preprocess_seqid_op(struct nfsd4_compound_state *cstate, u32 seqid,
 		 * a replayed close:
 		 */
 		sop = search_close_lru(stateid->si_stateownerid, flags);
+		/* It's not stale; let's assume it's expired: */
 		if (sop == NULL)
-			return nfserr_bad_stateid;
+			return nfserr_expired;
 		*sopp = sop;
 		goto check_replay;
 	}
@@ -3126,8 +3335,7 @@ nfsd4_open_downgrade(struct svc_rqst *rqstp,
 		goto out;
 	}
 	set_access(&share_access, stp->st_access_bmap);
-	nfs4_file_downgrade(stp->st_vfs_file,
-	                    share_access & ~od->od_share_access);
+	nfs4_file_downgrade(stp->st_file, share_access & ~od->od_share_access);
 
 	reset_union_bmap_access(od->od_share_access, &stp->st_access_bmap);
 	reset_union_bmap_deny(od->od_share_deny, &stp->st_deny_bmap);
@@ -3214,6 +3422,7 @@ nfsd4_delegreturn(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	status = nfserr_bad_stateid;
 	if (!is_delegation_stateid(stateid))
 		goto out;
+	status = nfserr_expired;
 	dp = find_delegation_stateid(inode, stateid);
 	if (!dp)
 		goto out;
@@ -3346,11 +3555,9 @@ static inline void
 nfs4_set_lock_denied(struct file_lock *fl, struct nfsd4_lock_denied *deny)
 {
 	struct nfs4_stateowner *sop;
-	unsigned int hval;
 
 	if (fl->fl_lmops == &nfsd_posix_mng_ops) {
 		sop = (struct nfs4_stateowner *) fl->fl_owner;
-		hval = lockownerid_hashval(sop->so_id);
 		kref_get(&sop->so_ref);
 		deny->ld_sop = sop;
 		deny->ld_clientid = sop->so_client->cl_clientid;
@@ -3446,8 +3653,6 @@ alloc_init_lock_stateid(struct nfs4_stateowner *sop, struct nfs4_file *fp, struc
 	stp->st_stateid.si_stateownerid = sop->so_id;
 	stp->st_stateid.si_fileid = fp->fi_id;
 	stp->st_stateid.si_generation = 0;
-	stp->st_vfs_file = open_stp->st_vfs_file; /* FIXME refcount?? */
-	stp->st_access_bmap = open_stp->st_access_bmap;
 	stp->st_deny_bmap = open_stp->st_deny_bmap;
 	stp->st_openstp = open_stp;
 
@@ -3472,7 +3677,8 @@ nfsd4_lock(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	struct nfs4_stateowner *open_sop = NULL;
 	struct nfs4_stateowner *lock_sop = NULL;
 	struct nfs4_stateid *lock_stp;
-	struct file *filp;
+	struct nfs4_file *fp;
+	struct file *filp = NULL;
 	struct file_lock file_lock;
 	struct file_lock conflock;
 	__be32 status = 0;
@@ -3502,7 +3708,6 @@ nfsd4_lock(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 		 * lock stateid.
 		 */
 		struct nfs4_stateid *open_stp = NULL;
-		struct nfs4_file *fp;
 		
 		status = nfserr_stale_clientid;
 		if (!nfsd4_has_session(cstate) &&
@@ -3545,9 +3750,9 @@ nfsd4_lock(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 		if (status)
 			goto out;
 		lock_sop = lock->lk_replay_owner;
+		fp = lock_stp->st_file;
 	}
 	/* lock->lk_replay_owner and lock_stp have been created or found */
-	filp = lock_stp->st_vfs_file;
 
 	status = nfserr_grace;
 	if (locks_in_grace() && !lock->lk_reclaim)
@@ -3560,11 +3765,19 @@ nfsd4_lock(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	switch (lock->lk_type) {
 		case NFS4_READ_LT:
 		case NFS4_READW_LT:
+			if (find_readable_file(lock_stp->st_file)) {
+				nfs4_get_vfs_file(rqstp, fp, &cstate->current_fh, NFS4_SHARE_ACCESS_READ);
+				filp = find_readable_file(lock_stp->st_file);
+			}
 			file_lock.fl_type = F_RDLCK;
 			cmd = F_SETLK;
 		break;
 		case NFS4_WRITE_LT:
 		case NFS4_WRITEW_LT:
+			if (find_writeable_file(lock_stp->st_file)) {
+				nfs4_get_vfs_file(rqstp, fp, &cstate->current_fh, NFS4_SHARE_ACCESS_WRITE);
+				filp = find_writeable_file(lock_stp->st_file);
+			}
 			file_lock.fl_type = F_WRLCK;
 			cmd = F_SETLK;
 		break;
@@ -3572,6 +3785,10 @@ nfsd4_lock(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 			status = nfserr_inval;
 		goto out;
 	}
+	if (!filp) {
+		status = nfserr_openmode;
+		goto out;
+	}
 	file_lock.fl_owner = (fl_owner_t)lock_sop;
 	file_lock.fl_pid = current->tgid;
 	file_lock.fl_file = filp;
@@ -3740,7 +3957,11 @@ nfsd4_locku(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 					&locku->lu_stateowner, &stp, NULL)))
 		goto out;
 
-	filp = stp->st_vfs_file;
+	filp = find_any_file(stp->st_file);
+	if (!filp) {
+		status = nfserr_lock_range;
+		goto out;
+	}
 	BUG_ON(!filp);
 	locks_init_lock(&file_lock);
 	file_lock.fl_type = F_UNLCK;
@@ -3787,13 +4008,13 @@ out_nfserr:
  * 	0: no locks held by lockowner
  */
 static int
-check_for_locks(struct file *filp, struct nfs4_stateowner *lowner)
+check_for_locks(struct nfs4_file *filp, struct nfs4_stateowner *lowner)
 {
 	struct file_lock **flpp;
-	struct inode *inode = filp->f_path.dentry->d_inode;
+	struct inode *inode = filp->fi_inode;
 	int status = 0;
 
-	lock_kernel();
+	lock_flocks();
 	for (flpp = &inode->i_flock; *flpp != NULL; flpp = &(*flpp)->fl_next) {
 		if ((*flpp)->fl_owner == (fl_owner_t)lowner) {
 			status = 1;
@@ -3801,7 +4022,7 @@ check_for_locks(struct file *filp, struct nfs4_stateowner *lowner)
 		}
 	}
 out:
-	unlock_kernel();
+	unlock_flocks();
 	return status;
 }
 
@@ -3841,7 +4062,7 @@ nfsd4_release_lockowner(struct svc_rqst *rqstp,
 				continue;
 			list_for_each_entry(stp, &sop->so_stateids,
 					st_perstateowner) {
-				if (check_for_locks(stp->st_vfs_file, sop))
+				if (check_for_locks(stp->st_file, sop))
 					goto out;
 				/* Note: so_perclient unused for lockowners,
 				 * so it's OK to fool with here. */
@@ -4066,16 +4287,8 @@ out_free_laundry:
 int
 nfs4_state_start(void)
 {
-	int ret;
-
-	if (nfs4_init)
-		return 0;
 	nfsd4_load_reboot_recovery_data();
-	ret = __nfs4_state_start();
-	if (ret)
-		return ret;
-	nfs4_init = 1;
-	return 0;
+	return __nfs4_state_start();
 }
 
 static void
@@ -4110,7 +4323,6 @@ __nfs4_state_shutdown(void)
 	}
 
 	nfsd4_shutdown_recdir();
-	nfs4_init = 0;
 }
 
 void
diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c
index ac17a70..f35a94a 100644
--- a/fs/nfsd/nfs4xdr.c
+++ b/fs/nfsd/nfs4xdr.c
@@ -1756,6 +1756,10 @@ nfsd4_encode_fattr(struct svc_fh *fhp, struct svc_export *exp,
 	struct nfs4_acl *acl = NULL;
 	struct nfsd4_compoundres *resp = rqstp->rq_resp;
 	u32 minorversion = resp->cstate.minorversion;
+	struct path path = {
+		.mnt	= exp->ex_path.mnt,
+		.dentry	= dentry,
+	};
 
 	BUG_ON(bmval1 & NFSD_WRITEONLY_ATTRS_WORD1);
 	BUG_ON(bmval0 & ~nfsd_suppattrs0(minorversion));
@@ -1776,7 +1780,7 @@ nfsd4_encode_fattr(struct svc_fh *fhp, struct svc_export *exp,
 			FATTR4_WORD0_MAXNAME)) ||
 	    (bmval1 & (FATTR4_WORD1_SPACE_AVAIL | FATTR4_WORD1_SPACE_FREE |
 		       FATTR4_WORD1_SPACE_TOTAL))) {
-		err = vfs_statfs(dentry, &statfs);
+		err = vfs_statfs(&path, &statfs);
 		if (err)
 			goto out_nfserr;
 	}
@@ -1801,19 +1805,23 @@ nfsd4_encode_fattr(struct svc_fh *fhp, struct svc_export *exp,
 				goto out_nfserr;
 		}
 	}
-	if ((buflen -= 16) < 0)
-		goto out_resource;
 
-	if (unlikely(bmval2)) {
+	if (bmval2) {
+		if ((buflen -= 16) < 0)
+			goto out_resource;
 		WRITE32(3);
 		WRITE32(bmval0);
 		WRITE32(bmval1);
 		WRITE32(bmval2);
-	} else if (likely(bmval1)) {
+	} else if (bmval1) {
+		if ((buflen -= 12) < 0)
+			goto out_resource;
 		WRITE32(2);
 		WRITE32(bmval0);
 		WRITE32(bmval1);
 	} else {
+		if ((buflen -= 8) < 0)
+			goto out_resource;
 		WRITE32(1);
 		WRITE32(bmval0);
 	}
@@ -1824,15 +1832,17 @@ nfsd4_encode_fattr(struct svc_fh *fhp, struct svc_export *exp,
 		u32 word1 = nfsd_suppattrs1(minorversion);
 		u32 word2 = nfsd_suppattrs2(minorversion);
 
-		if ((buflen -= 12) < 0)
-			goto out_resource;
 		if (!aclsupport)
 			word0 &= ~FATTR4_WORD0_ACL;
 		if (!word2) {
+			if ((buflen -= 12) < 0)
+				goto out_resource;
 			WRITE32(2);
 			WRITE32(word0);
 			WRITE32(word1);
 		} else {
+			if ((buflen -= 16) < 0)
+				goto out_resource;
 			WRITE32(3);
 			WRITE32(word0);
 			WRITE32(word1);
@@ -2630,7 +2640,7 @@ nfsd4_encode_read(struct nfsd4_compoundres *resp, __be32 nfserr,
 	}
 	read->rd_vlen = v;
 
-	nfserr = nfsd_read(read->rd_rqstp, read->rd_fhp, read->rd_filp,
+	nfserr = nfsd_read_file(read->rd_rqstp, read->rd_fhp, read->rd_filp,
 			read->rd_offset, resp->rqstp->rq_vec, read->rd_vlen,
 			&maxcount);
 
@@ -3325,6 +3335,7 @@ nfs4svc_encode_compoundres(struct svc_rqst *rqstp, __be32 *p, struct nfsd4_compo
 		}
 		/* Renew the clientid on success and on replay */
 		release_session_client(cs->session);
+		nfsd4_put_session(cs->session);
 	}
 	return 1;
 }
diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c
index 508941c..4514ebb 100644
--- a/fs/nfsd/nfsctl.c
+++ b/fs/nfsd/nfsctl.c
@@ -22,6 +22,7 @@
  */
 enum {
 	NFSD_Root = 1,
+#ifdef CONFIG_NFSD_DEPRECATED
 	NFSD_Svc,
 	NFSD_Add,
 	NFSD_Del,
@@ -29,6 +30,7 @@ enum {
 	NFSD_Unexport,
 	NFSD_Getfd,
 	NFSD_Getfs,
+#endif
 	NFSD_List,
 	NFSD_Export_features,
 	NFSD_Fh,
@@ -54,6 +56,7 @@ enum {
 /*
  * write() for these nodes.
  */
+#ifdef CONFIG_NFSD_DEPRECATED
 static ssize_t write_svc(struct file *file, char *buf, size_t size);
 static ssize_t write_add(struct file *file, char *buf, size_t size);
 static ssize_t write_del(struct file *file, char *buf, size_t size);
@@ -61,6 +64,7 @@ static ssize_t write_export(struct file *file, char *buf, size_t size);
 static ssize_t write_unexport(struct file *file, char *buf, size_t size);
 static ssize_t write_getfd(struct file *file, char *buf, size_t size);
 static ssize_t write_getfs(struct file *file, char *buf, size_t size);
+#endif
 static ssize_t write_filehandle(struct file *file, char *buf, size_t size);
 static ssize_t write_unlock_ip(struct file *file, char *buf, size_t size);
 static ssize_t write_unlock_fs(struct file *file, char *buf, size_t size);
@@ -76,6 +80,7 @@ static ssize_t write_recoverydir(struct file *file, char *buf, size_t size);
 #endif
 
 static ssize_t (*write_op[])(struct file *, char *, size_t) = {
+#ifdef CONFIG_NFSD_DEPRECATED
 	[NFSD_Svc] = write_svc,
 	[NFSD_Add] = write_add,
 	[NFSD_Del] = write_del,
@@ -83,6 +88,7 @@ static ssize_t (*write_op[])(struct file *, char *, size_t) = {
 	[NFSD_Unexport] = write_unexport,
 	[NFSD_Getfd] = write_getfd,
 	[NFSD_Getfs] = write_getfs,
+#endif
 	[NFSD_Fh] = write_filehandle,
 	[NFSD_FO_UnlockIP] = write_unlock_ip,
 	[NFSD_FO_UnlockFS] = write_unlock_fs,
@@ -121,6 +127,14 @@ static ssize_t nfsctl_transaction_write(struct file *file, const char __user *bu
 
 static ssize_t nfsctl_transaction_read(struct file *file, char __user *buf, size_t size, loff_t *pos)
 {
+	static int warned;
+	if (file->f_dentry->d_name.name[0] == '.' && !warned) {
+		printk(KERN_INFO
+		       "Warning: \"%s\" uses deprecated NFSD interface: %s."
+		       "  This will be removed in 2.6.40\n",
+		       current->comm, file->f_dentry->d_name.name);
+		warned = 1;
+	}
 	if (! file->private_data) {
 		/* An attempt to read a transaction file without writing
 		 * causes a 0-byte write so that the file can return
@@ -137,6 +151,7 @@ static const struct file_operations transaction_ops = {
 	.write		= nfsctl_transaction_write,
 	.read		= nfsctl_transaction_read,
 	.release	= simple_transaction_release,
+	.llseek		= default_llseek,
 };
 
 static int exports_open(struct inode *inode, struct file *file)
@@ -186,6 +201,7 @@ static const struct file_operations pool_stats_operations = {
  * payload - write methods
  */
 
+#ifdef CONFIG_NFSD_DEPRECATED
 /**
  * write_svc - Start kernel's NFSD server
  *
@@ -401,7 +417,7 @@ static ssize_t write_getfs(struct file *file, char *buf, size_t size)
 
 	ipv6_addr_set_v4mapped(sin->sin_addr.s_addr, &in6);
 
-	clp = auth_unix_lookup(&in6);
+	clp = auth_unix_lookup(&init_net, &in6);
 	if (!clp)
 		err = -EPERM;
 	else {
@@ -464,7 +480,7 @@ static ssize_t write_getfd(struct file *file, char *buf, size_t size)
 
 	ipv6_addr_set_v4mapped(sin->sin_addr.s_addr, &in6);
 
-	clp = auth_unix_lookup(&in6);
+	clp = auth_unix_lookup(&init_net, &in6);
 	if (!clp)
 		err = -EPERM;
 	else {
@@ -481,6 +497,7 @@ static ssize_t write_getfd(struct file *file, char *buf, size_t size)
  out:
 	return err;
 }
+#endif /* CONFIG_NFSD_DEPRECATED */
 
 /**
  * write_unlock_ip - Release all locks used by a client
@@ -949,15 +966,12 @@ static ssize_t __write_ports_addfd(char *buf)
 	if (err != 0)
 		return err;
 
-	err = lockd_up();
-	if (err != 0)
-		goto out;
-
 	err = svc_addsock(nfsd_serv, fd, buf, SIMPLE_TRANSACTION_LIMIT);
-	if (err < 0)
-		lockd_down();
+	if (err < 0) {
+		svc_destroy(nfsd_serv);
+		return err;
+	}
 
-out:
 	/* Decrease the count, but don't shut down the service */
 	nfsd_serv->sv_nrthreads--;
 	return err;
@@ -978,9 +992,6 @@ static ssize_t __write_ports_delfd(char *buf)
 	if (nfsd_serv != NULL)
 		len = svc_sock_names(nfsd_serv, buf,
 					SIMPLE_TRANSACTION_LIMIT, toclose);
-	if (len >= 0)
-		lockd_down();
-
 	kfree(toclose);
 	return len;
 }
@@ -1005,15 +1016,18 @@ static ssize_t __write_ports_addxprt(char *buf)
 	if (err != 0)
 		return err;
 
-	err = svc_create_xprt(nfsd_serv, transport,
+	err = svc_create_xprt(nfsd_serv, transport, &init_net,
 				PF_INET, port, SVC_SOCK_ANONYMOUS);
 	if (err < 0)
 		goto out_err;
 
-	err = svc_create_xprt(nfsd_serv, transport,
+	err = svc_create_xprt(nfsd_serv, transport, &init_net,
 				PF_INET6, port, SVC_SOCK_ANONYMOUS);
 	if (err < 0 && err != -EAFNOSUPPORT)
 		goto out_close;
+
+	/* Decrease the count, but don't shut down the service */
+	nfsd_serv->sv_nrthreads--;
 	return 0;
 out_close:
 	xprt = svc_find_xprt(nfsd_serv, transport, PF_INET, port);
@@ -1022,8 +1036,7 @@ out_close:
 		svc_xprt_put(xprt);
 	}
 out_err:
-	/* Decrease the count, but don't shut down the service */
-	nfsd_serv->sv_nrthreads--;
+	svc_destroy(nfsd_serv);
 	return err;
 }
 
@@ -1194,7 +1207,7 @@ static ssize_t write_maxblksize(struct file *file, char *buf, size_t size)
 			bsize = NFSSVC_MAXBLKSIZE;
 		bsize &= ~(1024-1);
 		mutex_lock(&nfsd_mutex);
-		if (nfsd_serv && nfsd_serv->sv_nrthreads) {
+		if (nfsd_serv) {
 			mutex_unlock(&nfsd_mutex);
 			return -EBUSY;
 		}
@@ -1310,6 +1323,8 @@ static ssize_t __write_recoverydir(struct file *file, char *buf, size_t size)
 			return -EINVAL;
 
 		status = nfs4_reset_recoverydir(recdir);
+		if (status)
+			return status;
 	}
 
 	return scnprintf(buf, SIMPLE_TRANSACTION_LIMIT, "%s\n",
@@ -1357,6 +1372,7 @@ static ssize_t write_recoverydir(struct file *file, char *buf, size_t size)
 static int nfsd_fill_super(struct super_block * sb, void * data, int silent)
 {
 	static struct tree_descr nfsd_files[] = {
+#ifdef CONFIG_NFSD_DEPRECATED
 		[NFSD_Svc] = {".svc", &transaction_ops, S_IWUSR},
 		[NFSD_Add] = {".add", &transaction_ops, S_IWUSR},
 		[NFSD_Del] = {".del", &transaction_ops, S_IWUSR},
@@ -1364,6 +1380,7 @@ static int nfsd_fill_super(struct super_block * sb, void * data, int silent)
 		[NFSD_Unexport] = {".unexport", &transaction_ops, S_IWUSR},
 		[NFSD_Getfd] = {".getfd", &transaction_ops, S_IWUSR|S_IRUSR},
 		[NFSD_Getfs] = {".getfs", &transaction_ops, S_IWUSR|S_IRUSR},
+#endif
 		[NFSD_List] = {"exports", &exports_operations, S_IRUGO},
 		[NFSD_Export_features] = {"export_features",
 					&export_features_operations, S_IRUGO},
@@ -1388,16 +1405,16 @@ static int nfsd_fill_super(struct super_block * sb, void * data, int silent)
 	return simple_fill_super(sb, 0x6e667364, nfsd_files);
 }
 
-static int nfsd_get_sb(struct file_system_type *fs_type,
-	int flags, const char *dev_name, void *data, struct vfsmount *mnt)
+static struct dentry *nfsd_mount(struct file_system_type *fs_type,
+	int flags, const char *dev_name, void *data)
 {
-	return get_sb_single(fs_type, flags, data, nfsd_fill_super, mnt);
+	return mount_single(fs_type, flags, data, nfsd_fill_super);
 }
 
 static struct file_system_type nfsd_fs_type = {
 	.owner		= THIS_MODULE,
 	.name		= "nfsd",
-	.get_sb		= nfsd_get_sb,
+	.mount		= nfsd_mount,
 	.kill_sb	= kill_litter_super,
 };
 
diff --git a/fs/nfsd/nfsd.h b/fs/nfsd/nfsd.h
index 7237776..6b641cf 100644
--- a/fs/nfsd/nfsd.h
+++ b/fs/nfsd/nfsd.h
@@ -153,6 +153,7 @@ void		nfsd_lockd_shutdown(void);
 #define nfserr_bad_seqid	cpu_to_be32(NFSERR_BAD_SEQID)
 #define	nfserr_symlink		cpu_to_be32(NFSERR_SYMLINK)
 #define	nfserr_not_same		cpu_to_be32(NFSERR_NOT_SAME)
+#define nfserr_lock_range	cpu_to_be32(NFSERR_LOCK_RANGE)
 #define	nfserr_restorefh	cpu_to_be32(NFSERR_RESTOREFH)
 #define	nfserr_attrnotsupp	cpu_to_be32(NFSERR_ATTRNOTSUPP)
 #define	nfserr_bad_xdr		cpu_to_be32(NFSERR_BAD_XDR)
@@ -248,7 +249,7 @@ extern time_t nfsd4_grace;
 #define	COMPOUND_SLACK_SPACE		140    /* OP_GETFH */
 #define COMPOUND_ERR_SLACK_SPACE	12     /* OP_SETATTR */
 
-#define NFSD_LAUNDROMAT_MINTIMEOUT      10   /* seconds */
+#define NFSD_LAUNDROMAT_MINTIMEOUT      1   /* seconds */
 
 /*
  * The following attributes are currently not supported by the NFSv4 server:
diff --git a/fs/nfsd/nfsfh.h b/fs/nfsd/nfsfh.h
index cdfb8c6..c16f8d8 100644
--- a/fs/nfsd/nfsfh.h
+++ b/fs/nfsd/nfsfh.h
@@ -196,8 +196,6 @@ fh_lock(struct svc_fh *fhp)
 static inline void
 fh_unlock(struct svc_fh *fhp)
 {
-	BUG_ON(!fhp->fh_dentry);
-
 	if (fhp->fh_locked) {
 		fill_post_wcc(fhp);
 		mutex_unlock(&fhp->fh_dentry->d_inode->i_mutex);
diff --git a/fs/nfsd/nfsproc.c b/fs/nfsd/nfsproc.c
index a047ad6..08e1726 100644
--- a/fs/nfsd/nfsproc.c
+++ b/fs/nfsd/nfsproc.c
@@ -144,7 +144,7 @@ nfsd_proc_read(struct svc_rqst *rqstp, struct nfsd_readargs *argp,
 	svc_reserve_auth(rqstp, (19<<2) + argp->count + 4);
 
 	resp->count = argp->count;
-	nfserr = nfsd_read(rqstp, fh_copy(&resp->fh, &argp->fh), NULL,
+	nfserr = nfsd_read(rqstp, fh_copy(&resp->fh, &argp->fh),
 				  argp->offset,
 			   	  rqstp->rq_vec, argp->vlen,
 				  &resp->count);
@@ -290,7 +290,6 @@ nfsd_proc_create(struct svc_rqst *rqstp, struct nfsd_createargs *argp,
 	 * gospel of sun micro
 	 */
 	if (type != S_IFREG) {
-		int	is_borc = 0;
 		if (type != S_IFBLK && type != S_IFCHR) {
 			rdev = 0;
 		} else if (type == S_IFCHR && !(attr->ia_valid & ATTR_SIZE)) {
@@ -298,7 +297,6 @@ nfsd_proc_create(struct svc_rqst *rqstp, struct nfsd_createargs *argp,
 			type = S_IFIFO;
 		} else {
 			/* Okay, char or block special */
-			is_borc = 1;
 			if (!rdev)
 				rdev = wanted;
 		}
diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c
index 06b2a26..2bae1d8 100644
--- a/fs/nfsd/nfssvc.c
+++ b/fs/nfsd/nfssvc.c
@@ -16,6 +16,7 @@
 #include <linux/lockd/bind.h>
 #include <linux/nfsacl.h>
 #include <linux/seq_file.h>
+#include <net/net_namespace.h>
 #include "nfsd.h"
 #include "cache.h"
 #include "vfs.h"
@@ -180,15 +181,80 @@ int nfsd_nrthreads(void)
 	return rv;
 }
 
+static int nfsd_init_socks(int port)
+{
+	int error;
+	if (!list_empty(&nfsd_serv->sv_permsocks))
+		return 0;
+
+	error = svc_create_xprt(nfsd_serv, "udp", &init_net, PF_INET, port,
+					SVC_SOCK_DEFAULTS);
+	if (error < 0)
+		return error;
+
+	error = svc_create_xprt(nfsd_serv, "tcp", &init_net, PF_INET, port,
+					SVC_SOCK_DEFAULTS);
+	if (error < 0)
+		return error;
+
+	return 0;
+}
+
+static bool nfsd_up = false;
+
+static int nfsd_startup(unsigned short port, int nrservs)
+{
+	int ret;
+
+	if (nfsd_up)
+		return 0;
+	/*
+	 * Readahead param cache - will no-op if it already exists.
+	 * (Note therefore results will be suboptimal if number of
+	 * threads is modified after nfsd start.)
+	 */
+	ret = nfsd_racache_init(2*nrservs);
+	if (ret)
+		return ret;
+	ret = nfsd_init_socks(port);
+	if (ret)
+		goto out_racache;
+	ret = lockd_up();
+	if (ret)
+		goto out_racache;
+	ret = nfs4_state_start();
+	if (ret)
+		goto out_lockd;
+	nfsd_up = true;
+	return 0;
+out_lockd:
+	lockd_down();
+out_racache:
+	nfsd_racache_shutdown();
+	return ret;
+}
+
+static void nfsd_shutdown(void)
+{
+	/*
+	 * write_ports can create the server without actually starting
+	 * any threads--if we get shut down before any threads are
+	 * started, then nfsd_last_thread will be run before any of this
+	 * other initialization has been done.
+	 */
+	if (!nfsd_up)
+		return;
+	nfs4_state_shutdown();
+	lockd_down();
+	nfsd_racache_shutdown();
+	nfsd_up = false;
+}
+
 static void nfsd_last_thread(struct svc_serv *serv)
 {
 	/* When last nfsd thread exits we need to do some clean-up */
-	struct svc_xprt *xprt;
-	list_for_each_entry(xprt, &serv->sv_permsocks, xpt_list)
-		lockd_down();
 	nfsd_serv = NULL;
-	nfsd_racache_shutdown();
-	nfs4_state_shutdown();
+	nfsd_shutdown();
 
 	printk(KERN_WARNING "nfsd: last server has exited, flushing export "
 			    "cache\n");
@@ -263,45 +329,18 @@ int nfsd_create_serv(void)
 		       nfsd_max_blksize >= 8*1024*2)
 			nfsd_max_blksize /= 2;
 	}
+	nfsd_reset_versions();
 
 	nfsd_serv = svc_create_pooled(&nfsd_program, nfsd_max_blksize,
 				      nfsd_last_thread, nfsd, THIS_MODULE);
 	if (nfsd_serv == NULL)
-		err = -ENOMEM;
-	else
-		set_max_drc();
+		return -ENOMEM;
 
+	set_max_drc();
 	do_gettimeofday(&nfssvc_boot);		/* record boot time */
 	return err;
 }
 
-static int nfsd_init_socks(int port)
-{
-	int error;
-	if (!list_empty(&nfsd_serv->sv_permsocks))
-		return 0;
-
-	error = svc_create_xprt(nfsd_serv, "udp", PF_INET, port,
-					SVC_SOCK_DEFAULTS);
-	if (error < 0)
-		return error;
-
-	error = lockd_up();
-	if (error < 0)
-		return error;
-
-	error = svc_create_xprt(nfsd_serv, "tcp", PF_INET, port,
-					SVC_SOCK_DEFAULTS);
-	if (error < 0)
-		return error;
-
-	error = lockd_up();
-	if (error < 0)
-		return error;
-
-	return 0;
-}
-
 int nfsd_nrpools(void)
 {
 	if (nfsd_serv == NULL)
@@ -376,10 +415,16 @@ int nfsd_set_nrthreads(int n, int *nthreads)
 	return err;
 }
 
+/*
+ * Adjust the number of threads and return the new number of threads.
+ * This is also the function that starts the server if necessary, if
+ * this is the first time nrservs is nonzero.
+ */
 int
 nfsd_svc(unsigned short port, int nrservs)
 {
 	int	error;
+	bool	nfsd_up_before;
 
 	mutex_lock(&nfsd_mutex);
 	dprintk("nfsd: creating service\n");
@@ -391,34 +436,29 @@ nfsd_svc(unsigned short port, int nrservs)
 	if (nrservs == 0 && nfsd_serv == NULL)
 		goto out;
 
-	/* Readahead param cache - will no-op if it already exists */
-	error =	nfsd_racache_init(2*nrservs);
-	if (error<0)
-		goto out;
-	error = nfs4_state_start();
+	error = nfsd_create_serv();
 	if (error)
 		goto out;
 
-	nfsd_reset_versions();
-
-	error = nfsd_create_serv();
+	nfsd_up_before = nfsd_up;
 
+	error = nfsd_startup(port, nrservs);
 	if (error)
-		goto out;
-	error = nfsd_init_socks(port);
-	if (error)
-		goto failure;
-
+		goto out_destroy;
 	error = svc_set_num_threads(nfsd_serv, NULL, nrservs);
-	if (error == 0)
-		/* We are holding a reference to nfsd_serv which
-		 * we don't want to count in the return value,
-		 * so subtract 1
-		 */
-		error = nfsd_serv->sv_nrthreads - 1;
- failure:
+	if (error)
+		goto out_shutdown;
+	/* We are holding a reference to nfsd_serv which
+	 * we don't want to count in the return value,
+	 * so subtract 1
+	 */
+	error = nfsd_serv->sv_nrthreads - 1;
+out_shutdown:
+	if (error < 0 && !nfsd_up_before)
+		nfsd_shutdown();
+out_destroy:
 	svc_destroy(nfsd_serv);		/* Release server */
- out:
+out:
 	mutex_unlock(&nfsd_mutex);
 	return error;
 }
diff --git a/fs/nfsd/state.h b/fs/nfsd/state.h
index 006c842..39adc27 100644
--- a/fs/nfsd/state.h
+++ b/fs/nfsd/state.h
@@ -35,6 +35,7 @@
 #ifndef _NFSD4_STATE_H
 #define _NFSD4_STATE_H
 
+#include <linux/sunrpc/svc_xprt.h>
 #include <linux/nfsd/nfsfh.h>
 #include "nfsfh.h"
 
@@ -64,19 +65,12 @@ typedef struct {
 	(s)->si_fileid, \
 	(s)->si_generation
 
-struct nfsd4_cb_sequence {
-	/* args/res */
-	u32			cbs_minorversion;
-	struct nfs4_client	*cbs_clp;
-};
-
-struct nfs4_rpc_args {
-	void				*args_op;
-	struct nfsd4_cb_sequence	args_seq;
-};
-
 struct nfsd4_callback {
-	struct nfs4_rpc_args cb_args;
+	void *cb_op;
+	struct nfs4_client *cb_clp;
+	u32 cb_minorversion;
+	struct rpc_message cb_msg;
+	const struct rpc_call_ops *cb_ops;
 	struct work_struct cb_work;
 };
 
@@ -88,11 +82,9 @@ struct nfs4_delegation {
 	struct nfs4_client	*dl_client;
 	struct nfs4_file	*dl_file;
 	struct file_lock	*dl_flock;
-	struct file		*dl_vfs_file;
 	u32			dl_type;
 	time_t			dl_time;
 /* For recall: */
-	u32			dl_ident;
 	stateid_t		dl_stateid;
 	struct knfsd_fh		dl_fh;
 	int			dl_retries;
@@ -104,8 +96,8 @@ struct nfs4_cb_conn {
 	/* SETCLIENTID info */
 	struct sockaddr_storage	cb_addr;
 	size_t			cb_addrlen;
-	u32                     cb_prog;
-	u32			cb_minorversion;
+	u32                     cb_prog; /* used only in 4.0 case;
+					    per-session otherwise */
 	u32                     cb_ident;	/* minorversion 0 only */
 	struct svc_xprt		*cb_xprt;	/* minorversion 1 only */
 };
@@ -161,6 +153,15 @@ struct nfsd4_clid_slot {
 	struct nfsd4_create_session	sl_cr_ses;
 };
 
+struct nfsd4_conn {
+	struct list_head cn_persession;
+	struct svc_xprt *cn_xprt;
+	struct svc_xpt_user cn_xpt_user;
+	struct nfsd4_session *cn_session;
+/* CDFC4_FORE, CDFC4_BACK: */
+	unsigned char cn_flags;
+};
+
 struct nfsd4_session {
 	struct kref		se_ref;
 	struct list_head	se_hash;	/* hash by sessionid */
@@ -170,6 +171,9 @@ struct nfsd4_session {
 	struct nfs4_sessionid	se_sessionid;
 	struct nfsd4_channel_attrs se_fchannel;
 	struct nfsd4_channel_attrs se_bchannel;
+	struct list_head	se_conns;
+	u32			se_cb_prog;
+	u32			se_cb_seq_nr;
 	struct nfsd4_slot	*se_slots[];	/* forward channel slots */
 };
 
@@ -222,24 +226,32 @@ struct nfs4_client {
 	clientid_t		cl_clientid;	/* generated by server */
 	nfs4_verifier		cl_confirm;	/* generated by server */
 	u32			cl_firststate;	/* recovery dir creation */
+	u32			cl_minorversion;
 
 	/* for v4.0 and v4.1 callbacks: */
 	struct nfs4_cb_conn	cl_cb_conn;
+#define NFSD4_CLIENT_CB_UPDATE	1
+#define NFSD4_CLIENT_KILL	2
+	unsigned long		cl_cb_flags;
 	struct rpc_clnt		*cl_cb_client;
+	u32			cl_cb_ident;
 	atomic_t		cl_cb_set;
+	struct nfsd4_callback	cl_cb_null;
+	struct nfsd4_session	*cl_cb_session;
+
+	/* for all client information that callback code might need: */
+	spinlock_t		cl_lock;
 
 	/* for nfs41 */
 	struct list_head	cl_sessions;
 	struct nfsd4_clid_slot	cl_cs_slot;	/* create_session slot */
 	u32			cl_exchange_flags;
-	struct nfs4_sessionid	cl_sessionid;
 	/* number of rpc's in progress over an associated session: */
 	atomic_t		cl_refcount;
 
 	/* for nfs41 callbacks */
 	/* We currently support a single back channel with a single slot */
 	unsigned long		cl_cb_slot_busy;
-	u32			cl_cb_seq_nr;
 	struct rpc_wait_queue	cl_cb_waitq;	/* backchannel callers may */
 						/* wait here for slots */
 };
@@ -342,12 +354,50 @@ struct nfs4_file {
 	struct list_head        fi_hash;    /* hash by "struct inode *" */
 	struct list_head        fi_stateids;
 	struct list_head	fi_delegations;
+	/* One each for O_RDONLY, O_WRONLY, O_RDWR: */
+	struct file *		fi_fds[3];
+	/* One each for O_RDONLY, O_WRONLY: */
+	atomic_t		fi_access[2];
+	/*
+	 * Each open stateid contributes 1 to either fi_readers or
+	 * fi_writers, or both, depending on the open mode.  A
+	 * delegation also takes an fi_readers reference.  Lock
+	 * stateid's take none.
+	 */
+	atomic_t		fi_readers;
+	atomic_t		fi_writers;
 	struct inode		*fi_inode;
 	u32                     fi_id;      /* used with stateowner->so_id 
 					     * for stateid_hashtbl hash */
 	bool			fi_had_conflict;
 };
 
+/* XXX: for first cut may fall back on returning file that doesn't work
+ * at all? */
+static inline struct file *find_writeable_file(struct nfs4_file *f)
+{
+	if (f->fi_fds[O_WRONLY])
+		return f->fi_fds[O_WRONLY];
+	return f->fi_fds[O_RDWR];
+}
+
+static inline struct file *find_readable_file(struct nfs4_file *f)
+{
+	if (f->fi_fds[O_RDONLY])
+		return f->fi_fds[O_RDONLY];
+	return f->fi_fds[O_RDWR];
+}
+
+static inline struct file *find_any_file(struct nfs4_file *f)
+{
+	if (f->fi_fds[O_RDWR])
+		return f->fi_fds[O_RDWR];
+	else if (f->fi_fds[O_WRONLY])
+		return f->fi_fds[O_WRONLY];
+	else
+		return f->fi_fds[O_RDONLY];
+}
+
 /*
 * nfs4_stateid can either be an open stateid or (eventually) a lock stateid
 *
@@ -373,7 +423,6 @@ struct nfs4_stateid {
 	struct nfs4_stateowner      * st_stateowner;
 	struct nfs4_file            * st_file;
 	stateid_t                     st_stateid;
-	struct file                 * st_vfs_file;
 	unsigned long                 st_access_bmap;
 	unsigned long                 st_deny_bmap;
 	struct nfs4_stateid         * st_openstp;
@@ -404,12 +453,13 @@ extern int nfs4_in_grace(void);
 extern __be32 nfs4_check_open_reclaim(clientid_t *clid);
 extern void nfs4_free_stateowner(struct kref *kref);
 extern int set_callback_cred(void);
-extern void nfsd4_probe_callback(struct nfs4_client *clp, struct nfs4_cb_conn *);
+extern void nfsd4_probe_callback(struct nfs4_client *clp);
+extern void nfsd4_change_callback(struct nfs4_client *clp, struct nfs4_cb_conn *);
 extern void nfsd4_do_callback_rpc(struct work_struct *);
 extern void nfsd4_cb_recall(struct nfs4_delegation *dp);
 extern int nfsd4_create_callback_queue(void);
 extern void nfsd4_destroy_callback_queue(void);
-extern void nfsd4_set_callback_client(struct nfs4_client *, struct rpc_clnt *);
+extern void nfsd4_shutdown_callback(struct nfs4_client *);
 extern void nfs4_put_delegation(struct nfs4_delegation *dp);
 extern __be32 nfs4_make_rec_clidname(char *clidname, struct xdr_netobj *clname);
 extern void nfsd4_init_recdir(char *recdir_name);
diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c
index 3c11112..184938f 100644
--- a/fs/nfsd/vfs.c
+++ b/fs/nfsd/vfs.c
@@ -281,23 +281,13 @@ commit_metadata(struct svc_fh *fhp)
 {
 	struct inode *inode = fhp->fh_dentry->d_inode;
 	const struct export_operations *export_ops = inode->i_sb->s_export_op;
-	int error = 0;
 
 	if (!EX_ISSYNC(fhp->fh_export))
 		return 0;
 
-	if (export_ops->commit_metadata) {
-		error = export_ops->commit_metadata(inode);
-	} else {
-		struct writeback_control wbc = {
-			.sync_mode = WB_SYNC_ALL,
-			.nr_to_write = 0, /* metadata only */
-		};
-
-		error = sync_inode(inode, &wbc);
-	}
-
-	return error;
+	if (export_ops->commit_metadata)
+		return export_ops->commit_metadata(inode);
+	return sync_inode_metadata(inode, 1);
 }
 
 /*
@@ -604,7 +594,7 @@ nfsd4_get_nfs4_acl(struct svc_rqst *rqstp, struct dentry *dentry, struct nfs4_ac
 	return error;
 }
 
-#endif /* defined(CONFIG_NFS_V4) */
+#endif /* defined(CONFIG_NFSD_V4) */
 
 #ifdef CONFIG_NFSD_V3
 /*
@@ -903,7 +893,6 @@ nfsd_vfs_read(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file,
               loff_t offset, struct kvec *vec, int vlen, unsigned long *count)
 {
 	struct inode *inode;
-	struct raparms	*ra;
 	mm_segment_t	oldfs;
 	__be32		err;
 	int		host_err;
@@ -914,12 +903,6 @@ nfsd_vfs_read(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file,
 	if (svc_msnfs(fhp) && !lock_may_read(inode, offset, *count))
 		goto out;
 
-	/* Get readahead parameters */
-	ra = nfsd_get_raparms(inode->i_sb->s_dev, inode->i_ino);
-
-	if (ra && ra->p_set)
-		file->f_ra = ra->p_ra;
-
 	if (file->f_op->splice_read && rqstp->rq_splice_ok) {
 		struct splice_desc sd = {
 			.len		= 0,
@@ -937,21 +920,11 @@ nfsd_vfs_read(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file,
 		set_fs(oldfs);
 	}
 
-	/* Write back readahead params */
-	if (ra) {
-		struct raparm_hbucket *rab = &raparm_hash[ra->p_hindex];
-		spin_lock(&rab->pb_lock);
-		ra->p_ra = file->f_ra;
-		ra->p_set = 1;
-		ra->p_count--;
-		spin_unlock(&rab->pb_lock);
-	}
-
 	if (host_err >= 0) {
 		nfsdstats.io_read += host_err;
 		*count = host_err;
 		err = 0;
-		fsnotify_access(file->f_path.dentry);
+		fsnotify_access(file);
 	} else 
 		err = nfserrno(host_err);
 out:
@@ -1062,7 +1035,7 @@ nfsd_vfs_write(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file,
 		goto out_nfserr;
 	*cnt = host_err;
 	nfsdstats.io_write += host_err;
-	fsnotify_modify(file->f_path.dentry);
+	fsnotify_modify(file);
 
 	/* clear setuid/setgid flag after write */
 	if (inode->i_mode & (S_ISUID | S_ISGID))
@@ -1086,8 +1059,45 @@ out:
  * on entry. On return, *count contains the number of bytes actually read.
  * N.B. After this call fhp needs an fh_put
  */
+__be32 nfsd_read(struct svc_rqst *rqstp, struct svc_fh *fhp,
+	loff_t offset, struct kvec *vec, int vlen, unsigned long *count)
+{
+	struct file *file;
+	struct inode *inode;
+	struct raparms	*ra;
+	__be32 err;
+
+	err = nfsd_open(rqstp, fhp, S_IFREG, NFSD_MAY_READ, &file);
+	if (err)
+		return err;
+
+	inode = file->f_path.dentry->d_inode;
+
+	/* Get readahead parameters */
+	ra = nfsd_get_raparms(inode->i_sb->s_dev, inode->i_ino);
+
+	if (ra && ra->p_set)
+		file->f_ra = ra->p_ra;
+
+	err = nfsd_vfs_read(rqstp, fhp, file, offset, vec, vlen, count);
+
+	/* Write back readahead params */
+	if (ra) {
+		struct raparm_hbucket *rab = &raparm_hash[ra->p_hindex];
+		spin_lock(&rab->pb_lock);
+		ra->p_ra = file->f_ra;
+		ra->p_set = 1;
+		ra->p_count--;
+		spin_unlock(&rab->pb_lock);
+	}
+
+	nfsd_close(file);
+	return err;
+}
+
+/* As above, but use the provided file descriptor. */
 __be32
-nfsd_read(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file,
+nfsd_read_file(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file,
 		loff_t offset, struct kvec *vec, int vlen,
 		unsigned long *count)
 {
@@ -1099,13 +1109,8 @@ nfsd_read(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file,
 		if (err)
 			goto out;
 		err = nfsd_vfs_read(rqstp, fhp, file, offset, vec, vlen, count);
-	} else {
-		err = nfsd_open(rqstp, fhp, S_IFREG, NFSD_MAY_READ, &file);
-		if (err)
-			goto out;
-		err = nfsd_vfs_read(rqstp, fhp, file, offset, vec, vlen, count);
-		nfsd_close(file);
-	}
+	} else /* Note file may still be NULL in NFSv4 special stateid case: */
+		err = nfsd_read(rqstp, fhp, offset, vec, vlen, count);
 out:
 	return err;
 }
@@ -1631,7 +1636,7 @@ nfsd_link(struct svc_rqst *rqstp, struct svc_fh *ffhp,
 				char *name, int len, struct svc_fh *tfhp)
 {
 	struct dentry	*ddir, *dnew, *dold;
-	struct inode	*dirp, *dest;
+	struct inode	*dirp;
 	__be32		err;
 	int		host_err;
 
@@ -1659,7 +1664,6 @@ nfsd_link(struct svc_rqst *rqstp, struct svc_fh *ffhp,
 		goto out_nfserr;
 
 	dold = tfhp->fh_dentry;
-	dest = dold->d_inode;
 
 	host_err = mnt_want_write(tfhp->fh_export->ex_path.mnt);
 	if (host_err) {
@@ -2019,9 +2023,17 @@ out:
 __be32
 nfsd_statfs(struct svc_rqst *rqstp, struct svc_fh *fhp, struct kstatfs *stat, int access)
 {
-	__be32 err = fh_verify(rqstp, fhp, 0, NFSD_MAY_NOP | access);
-	if (!err && vfs_statfs(fhp->fh_dentry,stat))
-		err = nfserr_io;
+	__be32 err;
+
+	err = fh_verify(rqstp, fhp, 0, NFSD_MAY_NOP | access);
+	if (!err) {
+		struct path path = {
+			.mnt	= fhp->fh_export->ex_path.mnt,
+			.dentry	= fhp->fh_dentry,
+		};
+		if (vfs_statfs(&path, stat))
+			err = nfserr_io;
+	}
 	return err;
 }
 
@@ -2038,7 +2050,6 @@ nfsd_permission(struct svc_rqst *rqstp, struct svc_export *exp,
 					struct dentry *dentry, int acc)
 {
 	struct inode	*inode = dentry->d_inode;
-	struct path	path;
 	int		err;
 
 	if (acc == NFSD_MAY_NOP)
@@ -2111,15 +2122,7 @@ nfsd_permission(struct svc_rqst *rqstp, struct svc_export *exp,
 	if (err == -EACCES && S_ISREG(inode->i_mode) &&
 	    acc == (NFSD_MAY_READ | NFSD_MAY_OWNER_OVERRIDE))
 		err = inode_permission(inode, MAY_EXEC);
-	if (err)
-		goto nfsd_out;
 
-	/* Do integrity (permission) checking now, but defer incrementing
-	 * IMA counts to the actual file open.
-	 */
-	path.mnt = exp->ex_path.mnt;
-	path.dentry = dentry;
-nfsd_out:
 	return err? nfserrno(err) : 0;
 }
 
diff --git a/fs/nfsd/vfs.h b/fs/nfsd/vfs.h
index 217a62c..9a370a5 100644
--- a/fs/nfsd/vfs.h
+++ b/fs/nfsd/vfs.h
@@ -64,7 +64,9 @@ __be32		nfsd_commit(struct svc_rqst *, struct svc_fh *,
 __be32		nfsd_open(struct svc_rqst *, struct svc_fh *, int,
 				int, struct file **);
 void		nfsd_close(struct file *);
-__be32 		nfsd_read(struct svc_rqst *, struct svc_fh *, struct file *,
+__be32 		nfsd_read(struct svc_rqst *, struct svc_fh *,
+				loff_t, struct kvec *, int, unsigned long *);
+__be32 		nfsd_read_file(struct svc_rqst *, struct svc_fh *, struct file *,
 				loff_t, struct kvec *, int, unsigned long *);
 __be32 		nfsd_write(struct svc_rqst *, struct svc_fh *,struct file *,
 				loff_t, struct kvec *,int, unsigned long *, int *);