Merge branch 'for-2.6.35' of git://linux-nfs.org/~bfields/linux

* 'for-2.6.35' of git://linux-nfs.org/~bfields/linux: (45 commits)
  Revert "nfsd4: distinguish expired from stale stateids"
  nfsd: safer initialization order in find_file()
  nfs4: minor callback code simplification, comment
  NFSD: don't report compiled-out versions as present
  nfsd4: implement reclaim_complete
  nfsd4: nfsd4_destroy_session must set callback client under the state lock
  nfsd4: keep a reference count on client while in use
  nfsd4: mark_client_expired
  nfsd4: introduce nfs4_client.cl_refcount
  nfsd4: refactor expire_client
  nfsd4: extend the client_lock to cover cl_lru
  nfsd4: use list_move in move_to_confirmed
  nfsd4: fold release_session into expire_client
  nfsd4: rename sessionid_lock to client_lock
  nfsd4: fix bare destroy_session null dereference
  nfsd4: use local variable in nfs4svc_encode_compoundres
  nfsd: further comment typos
  sunrpc: centralise most calls to svc_xprt_received
  nfsd4: fix unlikely race in session replay case
  nfsd4: fix filehandle comment
  ...

Documentation/filesystems/nfs/nfs41-server.txt

@@ -137,7 +137,7 @@ NS*| OPENATTR             | OPT        |              | Section 18.17  |
    | READ                 | REQ        |              | Section 18.22  |
    | READDIR              | REQ        |              | Section 18.23  |
    | READLINK             | OPT        |              | Section 18.24  |
-NS | RECLAIM_COMPLETE     | REQ        |              | Section 18.51  |
+   | RECLAIM_COMPLETE     | REQ        |              | Section 18.51  |
    | RELEASE_LOCKOWNER    | MNI        |              | N/A            |
    | REMOVE               | REQ        |              | Section 18.25  |
    | RENAME               | REQ        |              | Section 18.26  |

fs/nfsd/export.c

@@ -259,10 +259,9 @@ static struct cache_detail svc_expkey_cache = {
 	.alloc		= expkey_alloc,
 };
 
-static struct svc_expkey *
-svc_expkey_lookup(struct svc_expkey *item)
+static int
+svc_expkey_hash(struct svc_expkey *item)
 {
-	struct cache_head *ch;
 	int hash = item->ek_fsidtype;
 	char * cp = (char*)item->ek_fsid;
 	int len = key_len(item->ek_fsidtype);
@@ -270,6 +269,14 @@ svc_expkey_lookup(struct svc_expkey *item)
 	hash ^= hash_mem(cp, len, EXPKEY_HASHBITS);
 	hash ^= hash_ptr(item->ek_client, EXPKEY_HASHBITS);
 	hash &= EXPKEY_HASHMASK;
+	return hash;
+}
+
+static struct svc_expkey *
+svc_expkey_lookup(struct svc_expkey *item)
+{
+	struct cache_head *ch;
+	int hash = svc_expkey_hash(item);
 
 	ch = sunrpc_cache_lookup(&svc_expkey_cache, &item->h,
 				 hash);
@@ -283,13 +290,7 @@ static struct svc_expkey *
 svc_expkey_update(struct svc_expkey *new, struct svc_expkey *old)
 {
 	struct cache_head *ch;
-	int hash = new->ek_fsidtype;
-	char * cp = (char*)new->ek_fsid;
-	int len = key_len(new->ek_fsidtype);
-
-	hash ^= hash_mem(cp, len, EXPKEY_HASHBITS);
-	hash ^= hash_ptr(new->ek_client, EXPKEY_HASHBITS);
-	hash &= EXPKEY_HASHMASK;
+	int hash = svc_expkey_hash(new);
 
 	ch = sunrpc_cache_update(&svc_expkey_cache, &new->h,
 				 &old->h, hash);
@@ -738,14 +739,22 @@ struct cache_detail svc_export_cache = {
 	.alloc		= svc_export_alloc,
 };
 
-static struct svc_export *
-svc_export_lookup(struct svc_export *exp)
+static int
+svc_export_hash(struct svc_export *exp)
 {
-	struct cache_head *ch;
 	int hash;
+
 	hash = hash_ptr(exp->ex_client, EXPORT_HASHBITS);
 	hash ^= hash_ptr(exp->ex_path.dentry, EXPORT_HASHBITS);
 	hash ^= hash_ptr(exp->ex_path.mnt, EXPORT_HASHBITS);
+	return hash;
+}
+
+static struct svc_export *
+svc_export_lookup(struct svc_export *exp)
+{
+	struct cache_head *ch;
+	int hash = svc_export_hash(exp);
 
 	ch = sunrpc_cache_lookup(&svc_export_cache, &exp->h,
 				 hash);
@@ -759,10 +768,7 @@ static struct svc_export *
 svc_export_update(struct svc_export *new, struct svc_export *old)
 {
 	struct cache_head *ch;
-	int hash;
-	hash = hash_ptr(old->ex_client, EXPORT_HASHBITS);
-	hash ^= hash_ptr(old->ex_path.dentry, EXPORT_HASHBITS);
-	hash ^= hash_ptr(old->ex_path.mnt, EXPORT_HASHBITS);
+	int hash = svc_export_hash(old);
 
 	ch = sunrpc_cache_update(&svc_export_cache, &new->h,
 				 &old->h,
@@ -1071,9 +1077,9 @@ exp_export(struct nfsctl_export *nxp)
 		err = 0;
 finish:
 	kfree(new.ex_pathname);
-	if (exp)
+	if (!IS_ERR_OR_NULL(exp))
 		exp_put(exp);
-	if (fsid_key && !IS_ERR(fsid_key))
+	if (!IS_ERR_OR_NULL(fsid_key))
 		cache_put(&fsid_key->h, &svc_expkey_cache);
 	path_put(&path);
 out_put_clp:

fs/nfsd/nfs4callback.c

@@ -32,6 +32,7 @@
  */
 
 #include <linux/sunrpc/clnt.h>
+#include <linux/sunrpc/svc_xprt.h>
 #include <linux/slab.h>
 #include "nfsd.h"
 #include "state.h"
@@ -79,11 +80,6 @@ enum nfs_cb_opnum4 {
 					cb_sequence_dec_sz +            \
 					op_dec_sz)
 
-struct nfs4_rpc_args {
-	void				*args_op;
-	struct nfsd4_cb_sequence	args_seq;
-};
-
 /*
 * Generic encode routines from fs/nfs/nfs4xdr.c
 */
@@ -428,13 +424,19 @@ static struct rpc_procinfo     nfs4_cb_procedures[] = {
 };
 
 static struct rpc_version       nfs_cb_version4 = {
+/*
+ * Note on the callback rpc program version number: despite language in rfc
+ * 5661 section 18.36.3 requiring servers to use 4 in this field, the
+ * official xdr descriptions for both 4.0 and 4.1 specify version 1, and
+ * in practice that appears to be what implementations use.  The section
+ * 18.36.3 language is expected to be fixed in an erratum.
+ */
         .number                 = 1,
         .nrprocs                = ARRAY_SIZE(nfs4_cb_procedures),
         .procs                  = nfs4_cb_procedures
 };
 
 static struct rpc_version *	nfs_cb_version[] = {
-	NULL,
 	&nfs_cb_version4,
 };
 
@@ -456,15 +458,14 @@ static struct rpc_program cb_program = {
 
 static int max_cb_time(void)
 {
-	return max(NFSD_LEASE_TIME/10, (time_t)1) * HZ;
+	return max(nfsd4_lease/10, (time_t)1) * HZ;
 }
 
 /* Reference counting, callback cleanup, etc., all look racy as heck.
- * And why is cb_set an atomic? */
+ * And why is cl_cb_set an atomic? */
 
-int setup_callback_client(struct nfs4_client *clp)
+int setup_callback_client(struct nfs4_client *clp, struct nfs4_cb_conn *cb)
 {
-	struct nfs4_cb_conn *cb = &clp->cl_cb_conn;
 	struct rpc_timeout	timeparms = {
 		.to_initval	= max_cb_time(),
 		.to_retries	= 0,
@@ -476,7 +477,7 @@ int setup_callback_client(struct nfs4_client *clp)
 		.timeout	= &timeparms,
 		.program	= &cb_program,
 		.prognumber	= cb->cb_prog,
-		.version	= nfs_cb_version[1]->number,
+		.version	= 0,
 		.authflavor	= clp->cl_flavor,
 		.flags		= (RPC_CLNT_CREATE_NOPING | RPC_CLNT_CREATE_QUIET),
 		.client_name    = clp->cl_principal,
@@ -486,7 +487,7 @@ int setup_callback_client(struct nfs4_client *clp)
 	if (!clp->cl_principal && (clp->cl_flavor >= RPC_AUTH_GSS_KRB5))
 		return -EINVAL;
 	if (cb->cb_minorversion) {
-		args.bc_xprt = clp->cl_cb_xprt;
+		args.bc_xprt = cb->cb_xprt;
 		args.protocol = XPRT_TRANSPORT_BC_TCP;
 	}
 	/* Create RPC client */
@@ -496,7 +497,7 @@ int setup_callback_client(struct nfs4_client *clp)
 			PTR_ERR(client));
 		return PTR_ERR(client);
 	}
-	cb->cb_client = client;
+	nfsd4_set_callback_client(clp, client);
 	return 0;
 
 }
@@ -514,8 +515,7 @@ static void nfsd4_cb_probe_done(struct rpc_task *task, void *calldata)
 	if (task->tk_status)
 		warn_no_callback_path(clp, task->tk_status);
 	else
-		atomic_set(&clp->cl_cb_conn.cb_set, 1);
-	put_nfs4_client(clp);
+		atomic_set(&clp->cl_cb_set, 1);
 }
 
 static const struct rpc_call_ops nfsd4_cb_probe_ops = {
@@ -537,7 +537,6 @@ int set_callback_cred(void)
 
 void do_probe_callback(struct nfs4_client *clp)
 {
-	struct nfs4_cb_conn *cb = &clp->cl_cb_conn;
 	struct rpc_message msg = {
 		.rpc_proc       = &nfs4_cb_procedures[NFSPROC4_CLNT_CB_NULL],
 		.rpc_argp       = clp,
@@ -545,34 +544,27 @@ void do_probe_callback(struct nfs4_client *clp)
 	};
 	int status;
 
-	status = rpc_call_async(cb->cb_client, &msg,
+	status = rpc_call_async(clp->cl_cb_client, &msg,
 				RPC_TASK_SOFT | RPC_TASK_SOFTCONN,
 				&nfsd4_cb_probe_ops, (void *)clp);
-	if (status) {
+	if (status)
 		warn_no_callback_path(clp, status);
-		put_nfs4_client(clp);
-	}
 }
 
 /*
  * Set up the callback client and put a NFSPROC4_CB_NULL on the wire...
  */
-void
-nfsd4_probe_callback(struct nfs4_client *clp)
+void nfsd4_probe_callback(struct nfs4_client *clp, struct nfs4_cb_conn *cb)
 {
 	int status;
 
-	BUG_ON(atomic_read(&clp->cl_cb_conn.cb_set));
+	BUG_ON(atomic_read(&clp->cl_cb_set));
 
-	status = setup_callback_client(clp);
+	status = setup_callback_client(clp, cb);
 	if (status) {
 		warn_no_callback_path(clp, status);
 		return;
 	}
-
-	/* the task holds a reference to the nfs4_client struct */
-	atomic_inc(&clp->cl_count);
-
 	do_probe_callback(clp);
 }
 
@@ -658,18 +650,32 @@ 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 nfs4_client *clp = dp->dl_client;
+	struct rpc_clnt *current_rpc_client = clp->cl_cb_client;
 
 	nfsd4_cb_done(task, calldata);
 
+	if (current_rpc_client == NULL) {
+		/* We're shutting down; give up. */
+		/* XXX: err, or is it ok just to fall through
+		 * and rpc_restart_call? */
+		return;
+	}
+
 	switch (task->tk_status) {
 	case -EIO:
 		/* Network partition? */
-		atomic_set(&clp->cl_cb_conn.cb_set, 0);
+		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: */
+			nfsd4_cb_recall(dp);
+			return;
+		}
 	case -EBADHANDLE:
 	case -NFS4ERR_BAD_STATEID:
 		/* Race: client probably got cb_recall
@@ -677,7 +683,7 @@ static void nfsd4_cb_recall_done(struct rpc_task *task, void *calldata)
 		break;
 	default:
 		/* success, or error we can't handle */
-		goto done;
+		return;
 	}
 	if (dp->dl_retries--) {
 		rpc_delay(task, 2*HZ);
@@ -685,20 +691,16 @@ static void nfsd4_cb_recall_done(struct rpc_task *task, void *calldata)
 		rpc_restart_call(task);
 		return;
 	} else {
-		atomic_set(&clp->cl_cb_conn.cb_set, 0);
+		atomic_set(&clp->cl_cb_set, 0);
 		warn_no_callback_path(clp, task->tk_status);
 	}
-done:
-	kfree(task->tk_msg.rpc_argp);
 }
 
 static void nfsd4_cb_recall_release(void *calldata)
 {
 	struct nfs4_delegation *dp = calldata;
-	struct nfs4_client *clp = dp->dl_client;
 
 	nfs4_put_delegation(dp);
-	put_nfs4_client(clp);
 }
 
 static const struct rpc_call_ops nfsd4_cb_recall_ops = {
@@ -707,33 +709,75 @@ 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");
+	if (!callback_wq)
+		return -ENOMEM;
+	return 0;
+}
+
+void nfsd4_destroy_callback_queue(void)
+{
+	destroy_workqueue(callback_wq);
+}
+
+/* must be called under the state lock */
+void nfsd4_set_callback_client(struct nfs4_client *clp, struct rpc_clnt *new)
+{
+	struct rpc_clnt *old = clp->cl_cb_client;
+
+	clp->cl_cb_client = new;
+	/*
+	 * After this, any work that saw the old value of cl_cb_client will
+	 * be gone:
+	 */
+	flush_workqueue(callback_wq);
+	/* So we can safely shut it down: */
+	if (old)
+		rpc_shutdown_client(old);
+}
+
 /*
  * called with dp->dl_count inc'ed.
  */
-void
-nfsd4_cb_recall(struct nfs4_delegation *dp)
+static void _nfsd4_cb_recall(struct nfs4_delegation *dp)
 {
 	struct nfs4_client *clp = dp->dl_client;
-	struct rpc_clnt *clnt = clp->cl_cb_conn.cb_client;
-	struct nfs4_rpc_args *args;
+	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 = -ENOMEM;
+	int status;
+
+	if (clnt == NULL)
+		return; /* Client is shutting down; give up. */
 
-	args = kzalloc(sizeof(*args), GFP_KERNEL);
-	if (!args)
-		goto out;
 	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);
-out:
-	if (status) {
-		kfree(args);
-		put_nfs4_client(clp);
+	if (status)
 		nfs4_put_delegation(dp);
-	}
+}
+
+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);
+
+	_nfsd4_cb_recall(dp);
+}
+
+
+void nfsd4_cb_recall(struct nfs4_delegation *dp)
+{
+	queue_work(callback_wq, &dp->dl_recall.cb_work);
 }

fs/nfsd/nfs4proc.c

@@ -969,20 +969,36 @@ static struct nfsd4_operation nfsd4_ops[];
 static const char *nfsd4_op_name(unsigned opnum);
 
 /*
- * Enforce NFSv4.1 COMPOUND ordering rules.
+ * Enforce NFSv4.1 COMPOUND ordering rules:
  *
- * TODO:
- * - enforce NFS4ERR_NOT_ONLY_OP,
- * - DESTROY_SESSION MUST be the final operation in the COMPOUND request.
+ * Also note, enforced elsewhere:
+ *	- SEQUENCE other than as first op results in
+ *	  NFS4ERR_SEQUENCE_POS. (Enforced in nfsd4_sequence().)
+ *	- BIND_CONN_TO_SESSION must be the only op in its compound
+ *	  (Will be enforced in nfsd4_bind_conn_to_session().)
+ *	- DESTROY_SESSION must be the final operation in a compound, if
+ *	  sessionid's in SEQUENCE and DESTROY_SESSION are the same.
+ *	  (Enforced in nfsd4_destroy_session().)
  */
-static bool nfs41_op_ordering_ok(struct nfsd4_compoundargs *args)
+static __be32 nfs41_check_op_ordering(struct nfsd4_compoundargs *args)
 {
-	if (args->minorversion && args->opcnt > 0) {
-		struct nfsd4_op *op = &args->ops[0];
-		return (op->status == nfserr_op_illegal) ||
-		       (nfsd4_ops[op->opnum].op_flags & ALLOWED_AS_FIRST_OP);
-	}
-	return true;
+	struct nfsd4_op *op = &args->ops[0];
+
+	/* These ordering requirements don't apply to NFSv4.0: */
+	if (args->minorversion == 0)
+		return nfs_ok;
+	/* This is weird, but OK, not our problem: */
+	if (args->opcnt == 0)
+		return nfs_ok;
+	if (op->status == nfserr_op_illegal)
+		return nfs_ok;
+	if (!(nfsd4_ops[op->opnum].op_flags & ALLOWED_AS_FIRST_OP))
+		return nfserr_op_not_in_session;
+	if (op->opnum == OP_SEQUENCE)
+		return nfs_ok;
+	if (args->opcnt != 1)
+		return nfserr_not_only_op;
+	return nfs_ok;
 }
 
 /*
@@ -1012,6 +1028,7 @@ nfsd4_proc_compound(struct svc_rqst *rqstp,
 	resp->rqstp = rqstp;
 	resp->cstate.minorversion = args->minorversion;
 	resp->cstate.replay_owner = NULL;
+	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 */
@@ -1024,13 +1041,13 @@ nfsd4_proc_compound(struct svc_rqst *rqstp,
 	if (args->minorversion > nfsd_supported_minorversion)
 		goto out;
 
-	if (!nfs41_op_ordering_ok(args)) {
+	status = nfs41_check_op_ordering(args);
+	if (status) {
 		op = &args->ops[0];
-		op->status = nfserr_sequence_pos;
+		op->status = status;
 		goto encode_op;
 	}
 
-	status = nfs_ok;
 	while (!status && resp->opcnt < args->opcnt) {
 		op = &args->ops[resp->opcnt++];
 
@@ -1295,6 +1312,11 @@ static struct nfsd4_operation nfsd4_ops[] = {
 		.op_flags = ALLOWED_WITHOUT_FH | ALLOWED_AS_FIRST_OP,
 		.op_name = "OP_SEQUENCE",
 	},
+	[OP_RECLAIM_COMPLETE] = {
+		.op_func = (nfsd4op_func)nfsd4_reclaim_complete,
+		.op_flags = ALLOWED_WITHOUT_FH,
+		.op_name = "OP_RECLAIM_COMPLETE",
+	},
 };
 
 static const char *nfsd4_op_name(unsigned opnum)

fs/nfsd/nfs4state.c

@@ -45,8 +45,8 @@
 #define NFSDDBG_FACILITY                NFSDDBG_PROC
 
 /* Globals */
-static time_t lease_time = 90;     /* default lease time */
-static time_t user_lease_time = 90;
+time_t nfsd4_lease = 90;     /* default lease time */
+time_t nfsd4_grace = 90;
 static time_t boot_time;
 static u32 current_ownerid = 1;
 static u32 current_fileid = 1;
@@ -190,7 +190,7 @@ alloc_init_deleg(struct nfs4_client *clp, struct nfs4_stateid *stp, struct svc_f
 	dp->dl_vfs_file = stp->st_vfs_file;
 	dp->dl_type = type;
 	dp->dl_ident = cb->cb_ident;
-	dp->dl_stateid.si_boot = get_seconds();
+	dp->dl_stateid.si_boot = boot_time;
 	dp->dl_stateid.si_stateownerid = current_delegid++;
 	dp->dl_stateid.si_fileid = 0;
 	dp->dl_stateid.si_generation = 0;
@@ -199,6 +199,7 @@ alloc_init_deleg(struct nfs4_client *clp, struct nfs4_stateid *stp, struct svc_f
 	atomic_set(&dp->dl_count, 1);
 	list_add(&dp->dl_perfile, &fp->fi_delegations);
 	list_add(&dp->dl_perclnt, &clp->cl_delegations);
+	INIT_WORK(&dp->dl_recall.cb_work, nfsd4_do_callback_rpc);
 	return dp;
 }
 
@@ -249,6 +250,9 @@ unhash_delegation(struct nfs4_delegation *dp)
  * SETCLIENTID state 
  */
 
+/* client_lock protects the client lru list and session hash table */
+static DEFINE_SPINLOCK(client_lock);
+
 /* Hash tables for nfs4_clientid state */
 #define CLIENT_HASH_BITS                 4
 #define CLIENT_HASH_SIZE                (1 << CLIENT_HASH_BITS)
@@ -367,7 +371,6 @@ static void release_openowner(struct nfs4_stateowner *sop)
 	nfs4_put_stateowner(sop);
 }
 
-static DEFINE_SPINLOCK(sessionid_lock);
 #define SESSION_HASH_SIZE	512
 static struct list_head sessionid_hashtbl[SESSION_HASH_SIZE];
 
@@ -565,10 +568,10 @@ alloc_init_session(struct svc_rqst *rqstp, struct nfs4_client *clp,
 
 	new->se_flags = cses->flags;
 	kref_init(&new->se_ref);
-	spin_lock(&sessionid_lock);
+	spin_lock(&client_lock);
 	list_add(&new->se_hash, &sessionid_hashtbl[idx]);
 	list_add(&new->se_perclnt, &clp->cl_sessions);
-	spin_unlock(&sessionid_lock);
+	spin_unlock(&client_lock);
 
 	status = nfs_ok;
 out:
@@ -579,7 +582,7 @@ out_free:
 	goto out;
 }
 
-/* caller must hold sessionid_lock */
+/* caller must hold client_lock */
 static struct nfsd4_session *
 find_in_sessionid_hashtbl(struct nfs4_sessionid *sessionid)
 {
@@ -602,7 +605,7 @@ find_in_sessionid_hashtbl(struct nfs4_sessionid *sessionid)
 	return NULL;
 }
 
-/* caller must hold sessionid_lock */
+/* caller must hold client_lock */
 static void
 unhash_session(struct nfsd4_session *ses)
 {
@@ -610,15 +613,6 @@ unhash_session(struct nfsd4_session *ses)
 	list_del(&ses->se_perclnt);
 }
 
-static void
-release_session(struct nfsd4_session *ses)
-{
-	spin_lock(&sessionid_lock);
-	unhash_session(ses);
-	spin_unlock(&sessionid_lock);
-	nfsd4_put_session(ses);
-}
-
 void
 free_session(struct kref *kref)
 {
@@ -634,9 +628,18 @@ free_session(struct kref *kref)
 	kfree(ses);
 }
 
+/* must be called under the client_lock */
 static inline void
-renew_client(struct nfs4_client *clp)
+renew_client_locked(struct nfs4_client *clp)
 {
+	if (is_client_expired(clp)) {
+		dprintk("%s: client (clientid %08x/%08x) already expired\n",
+			__func__,
+			clp->cl_clientid.cl_boot,
+			clp->cl_clientid.cl_id);
+		return;
+	}
+
 	/*
 	* Move client to the end to the LRU list.
 	*/
@@ -647,6 +650,14 @@ renew_client(struct nfs4_client *clp)
 	clp->cl_time = get_seconds();
 }
 
+static inline void
+renew_client(struct nfs4_client *clp)
+{
+	spin_lock(&client_lock);
+	renew_client_locked(clp);
+	spin_unlock(&client_lock);
+}
+
 /* SETCLIENTID and SETCLIENTID_CONFIRM Helper functions */
 static int
 STALE_CLIENTID(clientid_t *clid)
@@ -680,27 +691,9 @@ static struct nfs4_client *alloc_client(struct xdr_netobj name)
 	return clp;
 }
 
-static void
-shutdown_callback_client(struct nfs4_client *clp)
-{
-	struct rpc_clnt *clnt = clp->cl_cb_conn.cb_client;
-
-	if (clnt) {
-		/*
-		 * Callback threads take a reference on the client, so there
-		 * should be no outstanding callbacks at this point.
-		 */
-		clp->cl_cb_conn.cb_client = NULL;
-		rpc_shutdown_client(clnt);
-	}
-}
-
 static inline void
 free_client(struct nfs4_client *clp)
 {
-	shutdown_callback_client(clp);
-	if (clp->cl_cb_xprt)
-		svc_xprt_put(clp->cl_cb_xprt);
 	if (clp->cl_cred.cr_group_info)
 		put_group_info(clp->cl_cred.cr_group_info);
 	kfree(clp->cl_principal);
@@ -709,10 +702,34 @@ free_client(struct nfs4_client *clp)
 }
 
 void
-put_nfs4_client(struct nfs4_client *clp)
+release_session_client(struct nfsd4_session *session)
 {
-	if (atomic_dec_and_test(&clp->cl_count))
+	struct nfs4_client *clp = session->se_client;
+
+	if (!atomic_dec_and_lock(&clp->cl_refcount, &client_lock))
+		return;
+	if (is_client_expired(clp)) {
 		free_client(clp);
+		session->se_client = NULL;
+	} 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)
+{
+	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);
+	}
 }
 
 static void
@@ -722,9 +739,6 @@ expire_client(struct nfs4_client *clp)
 	struct nfs4_delegation *dp;
 	struct list_head reaplist;
 
-	dprintk("NFSD: expire_client cl_count %d\n",
-	                    atomic_read(&clp->cl_count));
-
 	INIT_LIST_HEAD(&reaplist);
 	spin_lock(&recall_lock);
 	while (!list_empty(&clp->cl_delegations)) {
@@ -740,20 +754,20 @@ expire_client(struct nfs4_client *clp)
 		list_del_init(&dp->dl_recall_lru);
 		unhash_delegation(dp);
 	}
-	list_del(&clp->cl_idhash);
-	list_del(&clp->cl_strhash);
-	list_del(&clp->cl_lru);
 	while (!list_empty(&clp->cl_openowners)) {
 		sop = list_entry(clp->cl_openowners.next, struct nfs4_stateowner, so_perclient);
 		release_openowner(sop);
 	}
-	while (!list_empty(&clp->cl_sessions)) {
-		struct nfsd4_session  *ses;
-		ses = list_entry(clp->cl_sessions.next, struct nfsd4_session,
-				 se_perclnt);
-		release_session(ses);
-	}
-	put_nfs4_client(clp);
+	nfsd4_set_callback_client(clp, NULL);
+	if (clp->cl_cb_conn.cb_xprt)
+		svc_xprt_put(clp->cl_cb_conn.cb_xprt);
+	list_del(&clp->cl_idhash);
+	list_del(&clp->cl_strhash);
+	spin_lock(&client_lock);
+	unhash_client_locked(clp);
+	if (atomic_read(&clp->cl_refcount) == 0)
+		free_client(clp);
+	spin_unlock(&client_lock);
 }
 
 static void copy_verf(struct nfs4_client *target, nfs4_verifier *source)
@@ -839,14 +853,15 @@ static struct nfs4_client *create_client(struct xdr_netobj name, char *recdir,
 	}
 
 	memcpy(clp->cl_recdir, recdir, HEXDIR_LEN);
-	atomic_set(&clp->cl_count, 1);
-	atomic_set(&clp->cl_cb_conn.cb_set, 0);
+	atomic_set(&clp->cl_refcount, 0);
+	atomic_set(&clp->cl_cb_set, 0);
 	INIT_LIST_HEAD(&clp->cl_idhash);
 	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);
+	clp->cl_time = get_seconds();
 	clear_bit(0, &clp->cl_cb_slot_busy);
 	rpc_init_wait_queue(&clp->cl_cb_waitq, "Backchannel slot table");
 	copy_verf(clp, verf);
@@ -877,8 +892,7 @@ add_to_unconfirmed(struct nfs4_client *clp, unsigned int strhashval)
 	list_add(&clp->cl_strhash, &unconf_str_hashtbl[strhashval]);
 	idhashval = clientid_hashval(clp->cl_clientid.cl_id);
 	list_add(&clp->cl_idhash, &unconf_id_hashtbl[idhashval]);
-	list_add_tail(&clp->cl_lru, &client_lru);
-	clp->cl_time = get_seconds();
+	renew_client(clp);
 }
 
 static void
@@ -888,10 +902,9 @@ move_to_confirmed(struct nfs4_client *clp)
 	unsigned int strhashval;
 
 	dprintk("NFSD: move_to_confirm nfs4_client %p\n", clp);
-	list_del_init(&clp->cl_strhash);
 	list_move(&clp->cl_idhash, &conf_id_hashtbl[idhashval]);
 	strhashval = clientstr_hashval(clp->cl_recdir);
-	list_add(&clp->cl_strhash, &conf_str_hashtbl[strhashval]);
+	list_move(&clp->cl_strhash, &conf_str_hashtbl[strhashval]);
 	renew_client(clp);
 }
 
@@ -1327,15 +1340,9 @@ nfsd4_create_session(struct svc_rqst *rqstp,
 		cs_slot->sl_seqid++; /* from 0 to 1 */
 		move_to_confirmed(unconf);
 
-		/*
-		 * We do not support RDMA or persistent sessions
-		 */
-		cr_ses->flags &= ~SESSION4_PERSIST;
-		cr_ses->flags &= ~SESSION4_RDMA;
-
 		if (cr_ses->flags & SESSION4_BACK_CHAN) {
-			unconf->cl_cb_xprt = rqstp->rq_xprt;
-			svc_xprt_get(unconf->cl_cb_xprt);
+			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);
@@ -1344,7 +1351,7 @@ nfsd4_create_session(struct svc_rqst *rqstp,
 				cstate->minorversion;
 			unconf->cl_cb_conn.cb_prog = cr_ses->callback_prog;
 			unconf->cl_cb_seq_nr = 1;
-			nfsd4_probe_callback(unconf);
+			nfsd4_probe_callback(unconf, &unconf->cl_cb_conn);
 		}
 		conf = unconf;
 	} else {
@@ -1352,6 +1359,12 @@ nfsd4_create_session(struct svc_rqst *rqstp,
 		goto out;
 	}
 
+	/*
+	 * 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)
 		goto out;
@@ -1369,6 +1382,21 @@ out:
 	return status;
 }
 
+static bool nfsd4_last_compound_op(struct svc_rqst *rqstp)
+{
+	struct nfsd4_compoundres *resp = rqstp->rq_resp;
+	struct nfsd4_compoundargs *argp = rqstp->rq_argp;
+
+	return argp->opcnt == resp->opcnt;
+}
+
+static bool nfsd4_compound_in_session(struct nfsd4_session *session, struct nfs4_sessionid *sid)
+{
+	if (!session)
+		return 0;
+	return !memcmp(sid, &session->se_sessionid, sizeof(*sid));
+}
+
 __be32
 nfsd4_destroy_session(struct svc_rqst *r,
 		      struct nfsd4_compound_state *cstate,
@@ -1384,19 +1412,25 @@ nfsd4_destroy_session(struct svc_rqst *r,
 	 * - Do we need to clear any callback info from previous session?
 	 */
 
+	if (nfsd4_compound_in_session(cstate->session, &sessionid->sessionid)) {
+		if (!nfsd4_last_compound_op(r))
+			return nfserr_not_only_op;
+	}
 	dump_sessionid(__func__, &sessionid->sessionid);
-	spin_lock(&sessionid_lock);
+	spin_lock(&client_lock);
 	ses = find_in_sessionid_hashtbl(&sessionid->sessionid);
 	if (!ses) {
-		spin_unlock(&sessionid_lock);
+		spin_unlock(&client_lock);
 		goto out;
 	}
 
 	unhash_session(ses);
-	spin_unlock(&sessionid_lock);
+	spin_unlock(&client_lock);
 
+	nfs4_lock_state();
 	/* wait for callbacks */
-	shutdown_callback_client(ses->se_client);
+	nfsd4_set_callback_client(ses->se_client, NULL);
+	nfs4_unlock_state();
 	nfsd4_put_session(ses);
 	status = nfs_ok;
 out:
@@ -1417,7 +1451,7 @@ nfsd4_sequence(struct svc_rqst *rqstp,
 	if (resp->opcnt != 1)
 		return nfserr_sequence_pos;
 
-	spin_lock(&sessionid_lock);
+	spin_lock(&client_lock);
 	status = nfserr_badsession;
 	session = find_in_sessionid_hashtbl(&seq->sessionid);
 	if (!session)
@@ -1456,22 +1490,46 @@ nfsd4_sequence(struct svc_rqst *rqstp,
 	cstate->slot = slot;
 	cstate->session = session;
 
-	/* Hold a session reference until done processing the compound:
-	 * nfsd4_put_session called only if the cstate slot is set.
-	 */
-	nfsd4_get_session(session);
 out:
-	spin_unlock(&sessionid_lock);
-	/* Renew the clientid on success and on replay */
+	/* Hold a session reference until done processing the compound. */
 	if (cstate->session) {
-		nfs4_lock_state();
-		renew_client(session->se_client);
-		nfs4_unlock_state();
+		nfsd4_get_session(cstate->session);
+		atomic_inc(&session->se_client->cl_refcount);
 	}
+	spin_unlock(&client_lock);
 	dprintk("%s: return %d\n", __func__, ntohl(status));
 	return status;
 }
 
+__be32
+nfsd4_reclaim_complete(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, struct nfsd4_reclaim_complete *rc)
+{
+	if (rc->rca_one_fs) {
+		if (!cstate->current_fh.fh_dentry)
+			return nfserr_nofilehandle;
+		/*
+		 * We don't take advantage of the rca_one_fs case.
+		 * That's OK, it's optional, we can safely ignore it.
+		 */
+		 return nfs_ok;
+	}
+	nfs4_lock_state();
+	if (is_client_expired(cstate->session->se_client)) {
+		nfs4_unlock_state();
+		/*
+		 * The following error isn't really legal.
+		 * But we only get here if the client just explicitly
+		 * destroyed the client.  Surely it no longer cares what
+		 * error it gets back on an operation for the dead
+		 * client.
+		 */
+		return nfserr_stale_clientid;
+	}
+	nfsd4_create_clid_dir(cstate->session->se_client);
+	nfs4_unlock_state();
+	return nfs_ok;
+}
+
 __be32
 nfsd4_setclientid(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 		  struct nfsd4_setclientid *setclid)
@@ -1631,9 +1689,8 @@ nfsd4_setclientid_confirm(struct svc_rqst *rqstp,
 		if (!same_creds(&conf->cl_cred, &unconf->cl_cred))
 			status = nfserr_clid_inuse;
 		else {
-			/* XXX: We just turn off callbacks until we can handle
-			  * change request correctly. */
-			atomic_set(&conf->cl_cb_conn.cb_set, 0);
+			atomic_set(&conf->cl_cb_set, 0);
+			nfsd4_probe_callback(conf, &unconf->cl_cb_conn);
 			expire_client(unconf);
 			status = nfs_ok;
 
@@ -1667,7 +1724,7 @@ nfsd4_setclientid_confirm(struct svc_rqst *rqstp,
 			}
 			move_to_confirmed(unconf);
 			conf = unconf;
-			nfsd4_probe_callback(conf);
+			nfsd4_probe_callback(conf, &conf->cl_cb_conn);
 			status = nfs_ok;
 		}
 	} else if ((!conf || (conf && !same_verf(&conf->cl_confirm, &confirm)))
@@ -1700,12 +1757,12 @@ alloc_init_file(struct inode *ino)
 		INIT_LIST_HEAD(&fp->fi_hash);
 		INIT_LIST_HEAD(&fp->fi_stateids);
 		INIT_LIST_HEAD(&fp->fi_delegations);
-		spin_lock(&recall_lock);
-		list_add(&fp->fi_hash, &file_hashtbl[hashval]);
-		spin_unlock(&recall_lock);
 		fp->fi_inode = igrab(ino);
 		fp->fi_id = current_fileid++;
 		fp->fi_had_conflict = false;
+		spin_lock(&recall_lock);
+		list_add(&fp->fi_hash, &file_hashtbl[hashval]);
+		spin_unlock(&recall_lock);
 		return fp;
 	}
 	return NULL;
@@ -1827,7 +1884,7 @@ init_stateid(struct nfs4_stateid *stp, struct nfs4_file *fp, struct nfsd4_open *
 	stp->st_stateowner = sop;
 	get_nfs4_file(fp);
 	stp->st_file = fp;
-	stp->st_stateid.si_boot = get_seconds();
+	stp->st_stateid.si_boot = boot_time;
 	stp->st_stateid.si_stateownerid = sop->so_id;
 	stp->st_stateid.si_fileid = fp->fi_id;
 	stp->st_stateid.si_generation = 0;
@@ -2028,7 +2085,6 @@ void nfsd_break_deleg_cb(struct file_lock *fl)
 	 * lock) we know the server hasn't removed the lease yet, we know
 	 * it's safe to take a reference: */
 	atomic_inc(&dp->dl_count);
-	atomic_inc(&dp->dl_client->cl_count);
 
 	spin_lock(&recall_lock);
 	list_add_tail(&dp->dl_recall_lru, &del_recall_lru);
@@ -2347,7 +2403,7 @@ 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;
-	struct nfs4_cb_conn *cb = &sop->so_client->cl_cb_conn;
+	int cb_up = atomic_read(&sop->so_client->cl_cb_set);
 	struct file_lock fl, *flp = &fl;
 	int status, flag = 0;
 
@@ -2355,7 +2411,7 @@ nfs4_open_delegation(struct svc_fh *fh, struct nfsd4_open *open, struct nfs4_sta
 	open->op_recall = 0;
 	switch (open->op_claim_type) {
 		case NFS4_OPEN_CLAIM_PREVIOUS:
-			if (!atomic_read(&cb->cb_set))
+			if (!cb_up)
 				open->op_recall = 1;
 			flag = open->op_delegate_type;
 			if (flag == NFS4_OPEN_DELEGATE_NONE)
@@ -2366,7 +2422,7 @@ nfs4_open_delegation(struct svc_fh *fh, struct nfsd4_open *open, struct nfs4_sta
 			 * had the chance to reclaim theirs.... */
 			if (locks_in_grace())
 				goto out;
-			if (!atomic_read(&cb->cb_set) || !sop->so_confirmed)
+			if (!cb_up || !sop->so_confirmed)
 				goto out;
 			if (open->op_share_access & NFS4_SHARE_ACCESS_WRITE)
 				flag = NFS4_OPEN_DELEGATE_WRITE;
@@ -2483,10 +2539,8 @@ nfsd4_process_open2(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct nf
 	}
 	memcpy(&open->op_stateid, &stp->st_stateid, sizeof(stateid_t));
 
-	if (nfsd4_has_session(&resp->cstate)) {
+	if (nfsd4_has_session(&resp->cstate))
 		open->op_stateowner->so_confirmed = 1;
-		nfsd4_create_clid_dir(open->op_stateowner->so_client);
-	}
 
 	/*
 	* Attempt to hand out a delegation. No error return, because the
@@ -2537,7 +2591,7 @@ nfsd4_renew(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	renew_client(clp);
 	status = nfserr_cb_path_down;
 	if (!list_empty(&clp->cl_delegations)
-			&& !atomic_read(&clp->cl_cb_conn.cb_set))
+			&& !atomic_read(&clp->cl_cb_set))
 		goto out;
 	status = nfs_ok;
 out:
@@ -2554,6 +2608,12 @@ nfsd4_end_grace(void)
 	dprintk("NFSD: end of grace period\n");
 	nfsd4_recdir_purge_old();
 	locks_end_grace(&nfsd4_manager);
+	/*
+	 * Now that every NFSv4 client has had the chance to recover and
+	 * to see the (possibly new, possibly shorter) lease time, we
+	 * can safely set the next grace time to the current lease time:
+	 */
+	nfsd4_grace = nfsd4_lease;
 }
 
 static time_t
@@ -2563,15 +2623,17 @@ nfs4_laundromat(void)
 	struct nfs4_stateowner *sop;
 	struct nfs4_delegation *dp;
 	struct list_head *pos, *next, reaplist;
-	time_t cutoff = get_seconds() - NFSD_LEASE_TIME;
-	time_t t, clientid_val = NFSD_LEASE_TIME;
-	time_t u, test_val = NFSD_LEASE_TIME;
+	time_t cutoff = get_seconds() - nfsd4_lease;
+	time_t t, clientid_val = nfsd4_lease;
+	time_t u, test_val = nfsd4_lease;
 
 	nfs4_lock_state();
 
 	dprintk("NFSD: laundromat service - starting\n");
 	if (locks_in_grace())
 		nfsd4_end_grace();
+	INIT_LIST_HEAD(&reaplist);
+	spin_lock(&client_lock);
 	list_for_each_safe(pos, next, &client_lru) {
 		clp = list_entry(pos, struct nfs4_client, cl_lru);
 		if (time_after((unsigned long)clp->cl_time, (unsigned long)cutoff)) {
@@ -2580,12 +2642,22 @@ nfs4_laundromat(void)
 				clientid_val = t;
 			break;
 		}
+		if (atomic_read(&clp->cl_refcount)) {
+			dprintk("NFSD: client in use (clientid %08x)\n",
+				clp->cl_clientid.cl_id);
+			continue;
+		}
+		unhash_client_locked(clp);
+		list_add(&clp->cl_lru, &reaplist);
+	}
+	spin_unlock(&client_lock);
+	list_for_each_safe(pos, next, &reaplist) {
+		clp = list_entry(pos, struct nfs4_client, cl_lru);
 		dprintk("NFSD: purging unused client (clientid %08x)\n",
 			clp->cl_clientid.cl_id);
 		nfsd4_remove_clid_dir(clp);
 		expire_client(clp);
 	}
-	INIT_LIST_HEAD(&reaplist);
 	spin_lock(&recall_lock);
 	list_for_each_safe(pos, next, &del_recall_lru) {
 		dp = list_entry (pos, struct nfs4_delegation, dl_recall_lru);
@@ -2605,7 +2677,7 @@ nfs4_laundromat(void)
 		list_del_init(&dp->dl_recall_lru);
 		unhash_delegation(dp);
 	}
-	test_val = NFSD_LEASE_TIME;
+	test_val = nfsd4_lease;
 	list_for_each_safe(pos, next, &close_lru) {
 		sop = list_entry(pos, struct nfs4_stateowner, so_close_lru);
 		if (time_after((unsigned long)sop->so_time, (unsigned long)cutoff)) {
@@ -2661,39 +2733,11 @@ nfs4_check_fh(struct svc_fh *fhp, struct nfs4_stateid *stp)
 static int
 STALE_STATEID(stateid_t *stateid)
 {
-	if (time_after((unsigned long)boot_time,
-			(unsigned long)stateid->si_boot)) {
-		dprintk("NFSD: stale stateid " STATEID_FMT "!\n",
-			STATEID_VAL(stateid));
-		return 1;
-	}
-	return 0;
-}
-
-static int
-EXPIRED_STATEID(stateid_t *stateid)
-{
-	if (time_before((unsigned long)boot_time,
-			((unsigned long)stateid->si_boot)) &&
-	    time_before((unsigned long)(stateid->si_boot + lease_time), get_seconds())) {
-		dprintk("NFSD: expired stateid " STATEID_FMT "!\n",
-			STATEID_VAL(stateid));
-		return 1;
-	}
-	return 0;
-}
-
-static __be32
-stateid_error_map(stateid_t *stateid)
-{
-	if (STALE_STATEID(stateid))
-		return nfserr_stale_stateid;
-	if (EXPIRED_STATEID(stateid))
-		return nfserr_expired;
-
-	dprintk("NFSD: bad stateid " STATEID_FMT "!\n",
+	if (stateid->si_boot == boot_time)
+		return 0;
+	dprintk("NFSD: stale stateid " STATEID_FMT "!\n",
 		STATEID_VAL(stateid));
-	return nfserr_bad_stateid;
+	return 1;
 }
 
 static inline int
@@ -2817,10 +2861,8 @@ nfs4_preprocess_stateid_op(struct nfsd4_compound_state *cstate,
 	status = nfserr_bad_stateid;
 	if (is_delegation_stateid(stateid)) {
 		dp = find_delegation_stateid(ino, stateid);
-		if (!dp) {
-			status = stateid_error_map(stateid);
+		if (!dp)
 			goto out;
-		}
 		status = check_stateid_generation(stateid, &dp->dl_stateid,
 						  flags);
 		if (status)
@@ -2833,10 +2875,8 @@ nfs4_preprocess_stateid_op(struct nfsd4_compound_state *cstate,
 			*filpp = dp->dl_vfs_file;
 	} else { /* open or lock stateid */
 		stp = find_stateid(stateid, flags);
-		if (!stp) {
-			status = stateid_error_map(stateid);
+		if (!stp)
 			goto out;
-		}
 		if (nfs4_check_fh(current_fh, stp))
 			goto out;
 		if (!stp->st_stateowner->so_confirmed)
@@ -2908,7 +2948,7 @@ nfs4_preprocess_seqid_op(struct nfsd4_compound_state *cstate, u32 seqid,
 		 */
 		sop = search_close_lru(stateid->si_stateownerid, flags);
 		if (sop == NULL)
-			return stateid_error_map(stateid);
+			return nfserr_bad_stateid;
 		*sopp = sop;
 		goto check_replay;
 	}
@@ -3175,10 +3215,8 @@ nfsd4_delegreturn(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	if (!is_delegation_stateid(stateid))
 		goto out;
 	dp = find_delegation_stateid(inode, stateid);
-	if (!dp) {
-		status = stateid_error_map(stateid);
+	if (!dp)
 		goto out;
-	}
 	status = check_stateid_generation(stateid, &dp->dl_stateid, flags);
 	if (status)
 		goto out;
@@ -3404,7 +3442,7 @@ alloc_init_lock_stateid(struct nfs4_stateowner *sop, struct nfs4_file *fp, struc
 	stp->st_stateowner = sop;
 	get_nfs4_file(fp);
 	stp->st_file = fp;
-	stp->st_stateid.si_boot = get_seconds();
+	stp->st_stateid.si_boot = boot_time;
 	stp->st_stateid.si_stateownerid = sop->so_id;
 	stp->st_stateid.si_fileid = fp->fi_id;
 	stp->st_stateid.si_generation = 0;
@@ -3976,12 +4014,6 @@ nfsd4_load_reboot_recovery_data(void)
 		printk("NFSD: Failure reading reboot recovery data\n");
 }
 
-unsigned long
-get_nfs4_grace_period(void)
-{
-	return max(user_lease_time, lease_time) * HZ;
-}
-
 /*
  * Since the lifetime of a delegation isn't limited to that of an open, a
  * client may quite reasonably hang on to a delegation as long as it has
@@ -4008,20 +4040,27 @@ set_max_delegations(void)
 static int
 __nfs4_state_start(void)
 {
-	unsigned long grace_time;
+	int ret;
 
 	boot_time = get_seconds();
-	grace_time = get_nfs4_grace_period();
-	lease_time = user_lease_time;
 	locks_start_grace(&nfsd4_manager);
 	printk(KERN_INFO "NFSD: starting %ld-second grace period\n",
-	       grace_time/HZ);
+	       nfsd4_grace);
+	ret = set_callback_cred();
+	if (ret)
+		return -ENOMEM;
 	laundry_wq = create_singlethread_workqueue("nfsd4");
 	if (laundry_wq == NULL)
 		return -ENOMEM;
-	queue_delayed_work(laundry_wq, &laundromat_work, grace_time);
+	ret = nfsd4_create_callback_queue();
+	if (ret)
+		goto out_free_laundry;
+	queue_delayed_work(laundry_wq, &laundromat_work, nfsd4_grace * HZ);
 	set_max_delegations();
-	return set_callback_cred();
+	return 0;
+out_free_laundry:
+	destroy_workqueue(laundry_wq);
+	return ret;
 }
 
 int
@@ -4039,12 +4078,6 @@ nfs4_state_start(void)
 	return 0;
 }
 
-time_t
-nfs4_lease_time(void)
-{
-	return lease_time;
-}
-
 static void
 __nfs4_state_shutdown(void)
 {
@@ -4089,6 +4122,7 @@ nfs4_state_shutdown(void)
 	nfs4_lock_state();
 	nfs4_release_reclaim();
 	__nfs4_state_shutdown();
+	nfsd4_destroy_callback_queue();
 	nfs4_unlock_state();
 }
 
@@ -4128,21 +4162,3 @@ nfs4_recoverydir(void)
 {
 	return user_recovery_dirname;
 }
-
-/*
- * Called when leasetime is changed.
- *
- * The only way the protocol gives us to handle on-the-fly lease changes is to
- * simulate a reboot.  Instead of doing that, we just wait till the next time
- * we start to register any changes in lease time.  If the administrator
- * really wants to change the lease time *now*, they can go ahead and bring
- * nfsd down and then back up again after changing the lease time.
- *
- * user_lease_time is protected by nfsd_mutex since it's only really accessed
- * when nfsd is starting
- */
-void
-nfs4_reset_lease(time_t leasetime)
-{
-	user_lease_time = leasetime;
-}

fs/nfsd/nfs4xdr.c

@@ -1234,6 +1234,16 @@ nfsd4_decode_sequence(struct nfsd4_compoundargs *argp,
 	DECODE_TAIL;
 }
 
+static __be32 nfsd4_decode_reclaim_complete(struct nfsd4_compoundargs *argp, struct nfsd4_reclaim_complete *rc)
+{
+	DECODE_HEAD;
+
+	READ_BUF(4);
+	READ32(rc->rca_one_fs);
+
+	DECODE_TAIL;
+}
+
 static __be32
 nfsd4_decode_noop(struct nfsd4_compoundargs *argp, void *p)
 {
@@ -1346,7 +1356,7 @@ static nfsd4_dec nfsd41_dec_ops[] = {
 	[OP_TEST_STATEID]	= (nfsd4_dec)nfsd4_decode_notsupp,
 	[OP_WANT_DELEGATION]	= (nfsd4_dec)nfsd4_decode_notsupp,
 	[OP_DESTROY_CLIENTID]	= (nfsd4_dec)nfsd4_decode_notsupp,
-	[OP_RECLAIM_COMPLETE]	= (nfsd4_dec)nfsd4_decode_notsupp,
+	[OP_RECLAIM_COMPLETE]	= (nfsd4_dec)nfsd4_decode_reclaim_complete,
 };
 
 struct nfsd4_minorversion_ops {
@@ -1900,7 +1910,7 @@ nfsd4_encode_fattr(struct svc_fh *fhp, struct svc_export *exp,
 	if (bmval0 & FATTR4_WORD0_LEASE_TIME) {
 		if ((buflen -= 4) < 0)
 			goto out_resource;
-		WRITE32(NFSD_LEASE_TIME);
+		WRITE32(nfsd4_lease);
 	}
 	if (bmval0 & FATTR4_WORD0_RDATTR_ERROR) {
 		if ((buflen -= 4) < 0)
@@ -3307,11 +3317,14 @@ nfs4svc_encode_compoundres(struct svc_rqst *rqstp, __be32 *p, struct nfsd4_compo
 		iov = &rqstp->rq_res.head[0];
 	iov->iov_len = ((char*)resp->p) - (char*)iov->iov_base;
 	BUG_ON(iov->iov_len > PAGE_SIZE);
-	if (nfsd4_has_session(cs) && cs->status != nfserr_replay_cache) {
-		nfsd4_store_cache_entry(resp);
-		dprintk("%s: SET SLOT STATE TO AVAILABLE\n", __func__);
-		resp->cstate.slot->sl_inuse = false;
-		nfsd4_put_session(resp->cstate.session);
+	if (nfsd4_has_session(cs)) {
+		if (cs->status != nfserr_replay_cache) {
+			nfsd4_store_cache_entry(resp);
+			dprintk("%s: SET SLOT STATE TO AVAILABLE\n", __func__);
+			cs->slot->sl_inuse = false;
+		}
+		/* Renew the clientid on success and on replay */
+		release_session_client(cs->session);
 	}
 	return 1;
 }

fs/nfsd/nfsctl.c

@@ -46,6 +46,7 @@ enum {
 	 */
 #ifdef CONFIG_NFSD_V4
 	NFSD_Leasetime,
+	NFSD_Gracetime,
 	NFSD_RecoveryDir,
 #endif
 };
@@ -70,6 +71,7 @@ static ssize_t write_ports(struct file *file, char *buf, size_t size);
 static ssize_t write_maxblksize(struct file *file, char *buf, size_t size);
 #ifdef CONFIG_NFSD_V4
 static ssize_t write_leasetime(struct file *file, char *buf, size_t size);
+static ssize_t write_gracetime(struct file *file, char *buf, size_t size);
 static ssize_t write_recoverydir(struct file *file, char *buf, size_t size);
 #endif
 
@@ -91,6 +93,7 @@ static ssize_t (*write_op[])(struct file *, char *, size_t) = {
 	[NFSD_MaxBlkSize] = write_maxblksize,
 #ifdef CONFIG_NFSD_V4
 	[NFSD_Leasetime] = write_leasetime,
+	[NFSD_Gracetime] = write_gracetime,
 	[NFSD_RecoveryDir] = write_recoverydir,
 #endif
 };
@@ -1204,29 +1207,45 @@ static ssize_t write_maxblksize(struct file *file, char *buf, size_t size)
 }
 
 #ifdef CONFIG_NFSD_V4
-extern time_t nfs4_leasetime(void);
-
-static ssize_t __write_leasetime(struct file *file, char *buf, size_t size)
+static ssize_t __nfsd4_write_time(struct file *file, char *buf, size_t size, time_t *time)
 {
-	/* if size > 10 seconds, call
-	 * nfs4_reset_lease() then write out the new lease (seconds) as reply
-	 */
 	char *mesg = buf;
-	int rv, lease;
+	int rv, i;
 
 	if (size > 0) {
 		if (nfsd_serv)
 			return -EBUSY;
-		rv = get_int(&mesg, &lease);
+		rv = get_int(&mesg, &i);
 		if (rv)
 			return rv;
-		if (lease < 10 || lease > 3600)
+		/*
+		 * Some sanity checking.  We don't have a reason for
+		 * these particular numbers, but problems with the
+		 * extremes are:
+		 *	- Too short: the briefest network outage may
+		 *	  cause clients to lose all their locks.  Also,
+		 *	  the frequent polling may be wasteful.
+		 *	- Too long: do you really want reboot recovery
+		 *	  to take more than an hour?  Or to make other
+		 *	  clients wait an hour before being able to
+		 *	  revoke a dead client's locks?
+		 */
+		if (i < 10 || i > 3600)
 			return -EINVAL;
-		nfs4_reset_lease(lease);
+		*time = i;
 	}
 
-	return scnprintf(buf, SIMPLE_TRANSACTION_LIMIT, "%ld\n",
-							nfs4_lease_time());
+	return scnprintf(buf, SIMPLE_TRANSACTION_LIMIT, "%ld\n", *time);
+}
+
+static ssize_t nfsd4_write_time(struct file *file, char *buf, size_t size, time_t *time)
+{
+	ssize_t rv;
+
+	mutex_lock(&nfsd_mutex);
+	rv = __nfsd4_write_time(file, buf, size, time);
+	mutex_unlock(&nfsd_mutex);
+	return rv;
 }
 
 /**
@@ -1252,12 +1271,22 @@ static ssize_t __write_leasetime(struct file *file, char *buf, size_t size)
  */
 static ssize_t write_leasetime(struct file *file, char *buf, size_t size)
 {
-	ssize_t rv;
+	return nfsd4_write_time(file, buf, size, &nfsd4_lease);
+}
 
-	mutex_lock(&nfsd_mutex);
-	rv = __write_leasetime(file, buf, size);
-	mutex_unlock(&nfsd_mutex);
-	return rv;
+/**
+ * write_gracetime - Set or report current NFSv4 grace period time
+ *
+ * As above, but sets the time of the NFSv4 grace period.
+ *
+ * Note this should never be set to less than the *previous*
+ * lease-period time, but we don't try to enforce this.  (In the common
+ * case (a new boot), we don't know what the previous lease time was
+ * anyway.)
+ */
+static ssize_t write_gracetime(struct file *file, char *buf, size_t size)
+{
+	return nfsd4_write_time(file, buf, size, &nfsd4_grace);
 }
 
 extern char *nfs4_recoverydir(void);
@@ -1351,6 +1380,7 @@ static int nfsd_fill_super(struct super_block * sb, void * data, int silent)
 		[NFSD_MaxBlkSize] = {"max_block_size", &transaction_ops, S_IWUSR|S_IRUGO},
 #ifdef CONFIG_NFSD_V4
 		[NFSD_Leasetime] = {"nfsv4leasetime", &transaction_ops, S_IWUSR|S_IRUSR},
+		[NFSD_Gracetime] = {"nfsv4gracetime", &transaction_ops, S_IWUSR|S_IRUSR},
 		[NFSD_RecoveryDir] = {"nfsv4recoverydir", &transaction_ops, S_IWUSR|S_IRUSR},
 #endif
 		/* last one */ {""}

fs/nfsd/nfsd.h

@@ -82,7 +82,6 @@ int nfs4_state_init(void);
 void nfsd4_free_slabs(void);
 int nfs4_state_start(void);
 void nfs4_state_shutdown(void);
-time_t nfs4_lease_time(void);
 void nfs4_reset_lease(time_t leasetime);
 int nfs4_reset_recoverydir(char *recdir);
 #else
@@ -90,7 +89,6 @@ static inline int nfs4_state_init(void) { return 0; }
 static inline void nfsd4_free_slabs(void) { }
 static inline int nfs4_state_start(void) { return 0; }
 static inline void nfs4_state_shutdown(void) { }
-static inline time_t nfs4_lease_time(void) { return 0; }
 static inline void nfs4_reset_lease(time_t leasetime) { }
 static inline int nfs4_reset_recoverydir(char *recdir) { return 0; }
 #endif
@@ -229,6 +227,9 @@ extern struct timeval	nfssvc_boot;
 
 #ifdef CONFIG_NFSD_V4
 
+extern time_t nfsd4_lease;
+extern time_t nfsd4_grace;
+
 /* before processing a COMPOUND operation, we have to check that there
  * is enough space in the buffer for XDR encode to succeed.  otherwise,
  * we might process an operation with side effects, and be unable to
@@ -247,7 +248,6 @@ extern struct timeval	nfssvc_boot;
 #define	COMPOUND_SLACK_SPACE		140    /* OP_GETFH */
 #define COMPOUND_ERR_SLACK_SPACE	12     /* OP_SETATTR */
 
-#define NFSD_LEASE_TIME                 (nfs4_lease_time())
 #define NFSD_LAUNDROMAT_MINTIMEOUT      10   /* seconds */
 
 /*

fs/nfsd/nfssvc.c

@@ -120,7 +120,7 @@ u32 nfsd_supported_minorversion;
 int nfsd_vers(int vers, enum vers_op change)
 {
 	if (vers < NFSD_MINVERS || vers >= NFSD_NRVERS)
-		return -1;
+		return 0;
 	switch(change) {
 	case NFSD_SET:
 		nfsd_versions[vers] = nfsd_version[vers];

fs/nfsd/state.h

@@ -70,6 +70,16 @@ struct nfsd4_cb_sequence {
 	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;
+	struct work_struct cb_work;
+};
+
 struct nfs4_delegation {
 	struct list_head	dl_perfile;
 	struct list_head	dl_perclnt;
@@ -86,6 +96,7 @@ struct nfs4_delegation {
 	stateid_t		dl_stateid;
 	struct knfsd_fh		dl_fh;
 	int			dl_retries;
+	struct nfsd4_callback	dl_recall;
 };
 
 /* client delegation callback info */
@@ -96,9 +107,7 @@ struct nfs4_cb_conn {
 	u32                     cb_prog;
 	u32			cb_minorversion;
 	u32                     cb_ident;	/* minorversion 0 only */
-	/* RPC client info */
-	atomic_t		cb_set;     /* successful CB_NULL call */
-	struct rpc_clnt *       cb_client;
+	struct svc_xprt		*cb_xprt;	/* minorversion 1 only */
 };
 
 /* Maximum number of slots per session. 160 is useful for long haul TCP */
@@ -157,7 +166,7 @@ struct nfsd4_session {
 	struct list_head	se_hash;	/* hash by sessionid */
 	struct list_head	se_perclnt;
 	u32			se_flags;
-	struct nfs4_client	*se_client;	/* for expire_client */
+	struct nfs4_client	*se_client;
 	struct nfs4_sessionid	se_sessionid;
 	struct nfsd4_channel_attrs se_fchannel;
 	struct nfsd4_channel_attrs se_bchannel;
@@ -212,25 +221,41 @@ struct nfs4_client {
 	struct svc_cred		cl_cred; 	/* setclientid principal */
 	clientid_t		cl_clientid;	/* generated by server */
 	nfs4_verifier		cl_confirm;	/* generated by server */
-	struct nfs4_cb_conn	cl_cb_conn;     /* callback info */
-	atomic_t		cl_count;	/* ref count */
 	u32			cl_firststate;	/* recovery dir creation */
 
+	/* for v4.0 and v4.1 callbacks: */
+	struct nfs4_cb_conn	cl_cb_conn;
+	struct rpc_clnt		*cl_cb_client;
+	atomic_t		cl_cb_set;
+
 	/* 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 svc_xprt		*cl_cb_xprt;	/* 4.1 callback transport */
 	struct rpc_wait_queue	cl_cb_waitq;	/* backchannel callers may */
 						/* wait here for slots */
 };
 
+static inline void
+mark_client_expired(struct nfs4_client *clp)
+{
+	clp->cl_time = 0;
+}
+
+static inline bool
+is_client_expired(struct nfs4_client *clp)
+{
+	return clp->cl_time == 0;
+}
+
 /* struct nfs4_client_reset
  * one per old client. Populates reset_str_hashtbl. Filled from conf_id_hashtbl
  * upon lease reset, or from upcall to state_daemon (to read in state
@@ -377,11 +402,14 @@ extern void nfs4_lock_state(void);
 extern void nfs4_unlock_state(void);
 extern int nfs4_in_grace(void);
 extern __be32 nfs4_check_open_reclaim(clientid_t *clid);
-extern void put_nfs4_client(struct nfs4_client *clp);
 extern void nfs4_free_stateowner(struct kref *kref);
 extern int set_callback_cred(void);
-extern void nfsd4_probe_callback(struct nfs4_client *clp);
+extern void nfsd4_probe_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 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);
@@ -392,6 +420,7 @@ extern int nfs4_has_reclaimed_state(const char *name, bool use_exchange_id);
 extern void nfsd4_recdir_purge_old(void);
 extern int nfsd4_create_clid_dir(struct nfs4_client *clp);
 extern void nfsd4_remove_clid_dir(struct nfs4_client *clp);
+extern void release_session_client(struct nfsd4_session *);
 
 static inline void
 nfs4_put_stateowner(struct nfs4_stateowner *so)

fs/nfsd/vfs.c

@@ -724,7 +724,7 @@ nfsd_open(struct svc_rqst *rqstp, struct svc_fh *fhp, int type,
 	struct inode	*inode;
 	int		flags = O_RDONLY|O_LARGEFILE;
 	__be32		err;
-	int		host_err;
+	int		host_err = 0;
 
 	validate_process_creds();
 
@@ -761,7 +761,8 @@ nfsd_open(struct svc_rqst *rqstp, struct svc_fh *fhp, int type,
 	 * Check to see if there are any leases on this file.
 	 * This may block while leases are broken.
 	 */
-	host_err = break_lease(inode, O_NONBLOCK | ((access & NFSD_MAY_WRITE) ? O_WRONLY : 0));
+	if (!(access & NFSD_MAY_NOT_BREAK_LEASE))
+		host_err = break_lease(inode, O_NONBLOCK | ((access & NFSD_MAY_WRITE) ? O_WRONLY : 0));
 	if (host_err == -EWOULDBLOCK)
 		host_err = -ETIMEDOUT;
 	if (host_err) /* NOMEM or WOULDBLOCK */
@@ -1169,7 +1170,8 @@ nfsd_commit(struct svc_rqst *rqstp, struct svc_fh *fhp,
 			goto out;
 	}
 
-	err = nfsd_open(rqstp, fhp, S_IFREG, NFSD_MAY_WRITE, &file);
+	err = nfsd_open(rqstp, fhp, S_IFREG,
+			NFSD_MAY_WRITE|NFSD_MAY_NOT_BREAK_LEASE, &file);
 	if (err)
 		goto out;
 	if (EX_ISSYNC(fhp->fh_export)) {

fs/nfsd/vfs.h

@@ -20,6 +20,7 @@
 #define NFSD_MAY_OWNER_OVERRIDE	64
 #define NFSD_MAY_LOCAL_ACCESS	128 /* IRIX doing local access check on device special file*/
 #define NFSD_MAY_BYPASS_GSS_ON_ROOT 256
+#define NFSD_MAY_NOT_BREAK_LEASE 512
 
 #define NFSD_MAY_CREATE		(NFSD_MAY_EXEC|NFSD_MAY_WRITE)
 #define NFSD_MAY_REMOVE		(NFSD_MAY_EXEC|NFSD_MAY_WRITE|NFSD_MAY_TRUNC)

fs/nfsd/xdr4.h

@@ -381,6 +381,10 @@ struct nfsd4_destroy_session {
 	struct nfs4_sessionid	sessionid;
 };
 
+struct nfsd4_reclaim_complete {
+	u32 rca_one_fs;
+};
+
 struct nfsd4_op {
 	int					opnum;
 	__be32					status;
@@ -421,6 +425,7 @@ struct nfsd4_op {
 		struct nfsd4_create_session	create_session;
 		struct nfsd4_destroy_session	destroy_session;
 		struct nfsd4_sequence		sequence;
+		struct nfsd4_reclaim_complete	reclaim_complete;
 	} u;
 	struct nfs4_replay *			replay;
 };
@@ -513,9 +518,8 @@ extern void nfsd4_store_cache_entry(struct nfsd4_compoundres *resp);
 extern __be32 nfsd4_replay_cache_entry(struct nfsd4_compoundres *resp,
 		struct nfsd4_sequence *seq);
 extern __be32 nfsd4_exchange_id(struct svc_rqst *rqstp,
-		struct nfsd4_compound_state *,
-struct nfsd4_exchange_id *);
-		extern __be32 nfsd4_create_session(struct svc_rqst *,
+		struct nfsd4_compound_state *, struct nfsd4_exchange_id *);
+extern __be32 nfsd4_create_session(struct svc_rqst *,
 		struct nfsd4_compound_state *,
 		struct nfsd4_create_session *);
 extern __be32 nfsd4_sequence(struct svc_rqst *,
@@ -524,6 +528,7 @@ extern __be32 nfsd4_sequence(struct svc_rqst *,
 extern __be32 nfsd4_destroy_session(struct svc_rqst *,
 		struct nfsd4_compound_state *,
 		struct nfsd4_destroy_session *);
+__be32 nfsd4_reclaim_complete(struct svc_rqst *, struct nfsd4_compound_state *, struct nfsd4_reclaim_complete *);
 extern __be32 nfsd4_process_open1(struct nfsd4_compound_state *,
 		struct nfsd4_open *open);
 extern __be32 nfsd4_process_open2(struct svc_rqst *rqstp,

include/linux/nfsd/nfsfh.h

@@ -40,12 +40,12 @@ struct nfs_fhbase_old {
  * This is the new flexible, extensible style NFSv2/v3 file handle.
  * by Neil Brown <neilb@cse.unsw.edu.au> - March 2000
  *
- * The file handle is seens as a list of 4byte words.
- * The first word contains a version number (1) and four descriptor bytes
+ * The file handle starts with a sequence of four-byte words.
+ * The first word contains a version number (1) and three descriptor bytes
  * that tell how the remaining 3 variable length fields should be handled.
  * These three bytes are auth_type, fsid_type and fileid_type.
  *
- * All 4byte values are in host-byte-order.
+ * All four-byte values are in host-byte-order.
  *
  * The auth_type field specifies how the filehandle can be authenticated
  * This might allow a file to be confirmed to be in a writable part of a

net/sunrpc/cache.c

@@ -50,11 +50,17 @@ static void cache_init(struct cache_head *h)
 	h->last_refresh = now;
 }
 
+static inline int cache_is_expired(struct cache_detail *detail, struct cache_head *h)
+{
+	return  (h->expiry_time < get_seconds()) ||
+		(detail->flush_time > h->last_refresh);
+}
+
 struct cache_head *sunrpc_cache_lookup(struct cache_detail *detail,
 				       struct cache_head *key, int hash)
 {
 	struct cache_head **head,  **hp;
-	struct cache_head *new = NULL;
+	struct cache_head *new = NULL, *freeme = NULL;
 
 	head = &detail->hash_table[hash];
 
@@ -63,6 +69,9 @@ struct cache_head *sunrpc_cache_lookup(struct cache_detail *detail,
 	for (hp=head; *hp != NULL ; hp = &(*hp)->next) {
 		struct cache_head *tmp = *hp;
 		if (detail->match(tmp, key)) {
+			if (cache_is_expired(detail, tmp))
+				/* This entry is expired, we will discard it. */
+				break;
 			cache_get(tmp);
 			read_unlock(&detail->hash_lock);
 			return tmp;
@@ -87,6 +96,13 @@ struct cache_head *sunrpc_cache_lookup(struct cache_detail *detail,
 	for (hp=head; *hp != NULL ; hp = &(*hp)->next) {
 		struct cache_head *tmp = *hp;
 		if (detail->match(tmp, key)) {
+			if (cache_is_expired(detail, tmp)) {
+				*hp = tmp->next;
+				tmp->next = NULL;
+				detail->entries --;
+				freeme = tmp;
+				break;
+			}
 			cache_get(tmp);
 			write_unlock(&detail->hash_lock);
 			cache_put(new, detail);
@@ -99,6 +115,8 @@ struct cache_head *sunrpc_cache_lookup(struct cache_detail *detail,
 	cache_get(new);
 	write_unlock(&detail->hash_lock);
 
+	if (freeme)
+		cache_put(freeme, detail);
 	return new;
 }
 EXPORT_SYMBOL_GPL(sunrpc_cache_lookup);
@@ -184,10 +202,7 @@ static int cache_make_upcall(struct cache_detail *cd, struct cache_head *h)
 
 static inline int cache_is_valid(struct cache_detail *detail, struct cache_head *h)
 {
-	if (!test_bit(CACHE_VALID, &h->flags) ||
-	    h->expiry_time < get_seconds())
-		return -EAGAIN;
-	else if (detail->flush_time > h->last_refresh)
+	if (!test_bit(CACHE_VALID, &h->flags))
 		return -EAGAIN;
 	else {
 		/* entry is valid */
@@ -398,31 +413,27 @@ static int cache_clean(void)
 		/* Ok, now to clean this strand */
 
 		cp = & current_detail->hash_table[current_index];
-		ch = *cp;
-		for (; ch; cp= & ch->next, ch= *cp) {
+		for (ch = *cp ; ch ; cp = & ch->next, ch = *cp) {
 			if (current_detail->nextcheck > ch->expiry_time)
 				current_detail->nextcheck = ch->expiry_time+1;
-			if (ch->expiry_time >= get_seconds() &&
-			    ch->last_refresh >= current_detail->flush_time)
+			if (!cache_is_expired(current_detail, ch))
 				continue;
-			if (test_and_clear_bit(CACHE_PENDING, &ch->flags))
-				cache_dequeue(current_detail, ch);
 
-			if (atomic_read(&ch->ref.refcount) == 1)
-				break;
-		}
-		if (ch) {
 			*cp = ch->next;
 			ch->next = NULL;
 			current_detail->entries--;
 			rv = 1;
+			break;
 		}
+
 		write_unlock(&current_detail->hash_lock);
 		d = current_detail;
 		if (!ch)
 			current_index ++;
 		spin_unlock(&cache_list_lock);
 		if (ch) {
+			if (test_and_clear_bit(CACHE_PENDING, &ch->flags))
+				cache_dequeue(current_detail, ch);
 			cache_revisit_request(ch);
 			cache_put(ch, d);
 		}
@@ -1234,8 +1245,10 @@ static int content_open(struct inode *inode, struct file *file,
 	if (!cd || !try_module_get(cd->owner))
 		return -EACCES;
 	han = __seq_open_private(file, &cache_content_op, sizeof(*han));
-	if (han == NULL)
+	if (han == NULL) {
+		module_put(cd->owner);
 		return -ENOMEM;
+	}
 
 	han->cd = cd;
 	return 0;

net/sunrpc/svc_xprt.c

@@ -744,8 +744,10 @@ int svc_recv(struct svc_rqst *rqstp, long timeout)
 		if (rqstp->rq_deferred) {
 			svc_xprt_received(xprt);
 			len = svc_deferred_recv(rqstp);
-		} else
+		} else {
 			len = xprt->xpt_ops->xpo_recvfrom(rqstp);
+			svc_xprt_received(xprt);
+		}
 		dprintk("svc: got len=%d\n", len);
 	}
 
@@ -893,12 +895,12 @@ void svc_delete_xprt(struct svc_xprt *xprt)
 	 */
 	if (test_bit(XPT_TEMP, &xprt->xpt_flags))
 		serv->sv_tmpcnt--;
+	spin_unlock_bh(&serv->sv_lock);
 
 	while ((dr = svc_deferred_dequeue(xprt)) != NULL)
 		kfree(dr);
 
 	svc_xprt_put(xprt);
-	spin_unlock_bh(&serv->sv_lock);
 }
 
 void svc_close_xprt(struct svc_xprt *xprt)

net/sunrpc/svcsock.c

@@ -547,7 +547,6 @@ static int svc_udp_recvfrom(struct svc_rqst *rqstp)
 			dprintk("svc: recvfrom returned error %d\n", -err);
 			set_bit(XPT_DATA, &svsk->sk_xprt.xpt_flags);
 		}
-		svc_xprt_received(&svsk->sk_xprt);
 		return -EAGAIN;
 	}
 	len = svc_addr_len(svc_addr(rqstp));
@@ -562,11 +561,6 @@ static int svc_udp_recvfrom(struct svc_rqst *rqstp)
 	svsk->sk_sk->sk_stamp = skb->tstamp;
 	set_bit(XPT_DATA, &svsk->sk_xprt.xpt_flags); /* there may be more data... */
 
-	/*
-	 * Maybe more packets - kick another thread ASAP.
-	 */
-	svc_xprt_received(&svsk->sk_xprt);
-
 	len  = skb->len - sizeof(struct udphdr);
 	rqstp->rq_arg.len = len;
 
@@ -917,7 +911,6 @@ static int svc_tcp_recv_record(struct svc_sock *svsk, struct svc_rqst *rqstp)
 		if (len < want) {
 			dprintk("svc: short recvfrom while reading record "
 				"length (%d of %d)\n", len, want);
-			svc_xprt_received(&svsk->sk_xprt);
 			goto err_again; /* record header not complete */
 		}
 
@@ -953,7 +946,6 @@ static int svc_tcp_recv_record(struct svc_sock *svsk, struct svc_rqst *rqstp)
 	if (len < svsk->sk_reclen) {
 		dprintk("svc: incomplete TCP record (%d of %d)\n",
 			len, svsk->sk_reclen);
-		svc_xprt_received(&svsk->sk_xprt);
 		goto err_again;	/* record not complete */
 	}
 	len = svsk->sk_reclen;
@@ -961,14 +953,11 @@ static int svc_tcp_recv_record(struct svc_sock *svsk, struct svc_rqst *rqstp)
 
 	return len;
  error:
-	if (len == -EAGAIN) {
+	if (len == -EAGAIN)
 		dprintk("RPC: TCP recv_record got EAGAIN\n");
-		svc_xprt_received(&svsk->sk_xprt);
-	}
 	return len;
  err_delete:
 	set_bit(XPT_CLOSE, &svsk->sk_xprt.xpt_flags);
-	svc_xprt_received(&svsk->sk_xprt);
  err_again:
 	return -EAGAIN;
 }
@@ -1110,7 +1099,6 @@ out:
 	svsk->sk_tcplen = 0;
 
 	svc_xprt_copy_addrs(rqstp, &svsk->sk_xprt);
-	svc_xprt_received(&svsk->sk_xprt);
 	if (serv->sv_stats)
 		serv->sv_stats->nettcpcnt++;
 
@@ -1119,7 +1107,6 @@ out:
 err_again:
 	if (len == -EAGAIN) {
 		dprintk("RPC: TCP recvfrom got EAGAIN\n");
-		svc_xprt_received(&svsk->sk_xprt);
 		return len;
 	}
 error:

net/sunrpc/xprtrdma/svc_rdma_recvfrom.c

@@ -566,7 +566,6 @@ static int rdma_read_complete(struct svc_rqst *rqstp,
 		ret, rqstp->rq_arg.len,	rqstp->rq_arg.head[0].iov_base,
 		rqstp->rq_arg.head[0].iov_len);
 
-	svc_xprt_received(rqstp->rq_xprt);
 	return ret;
 }
 
@@ -665,7 +664,6 @@ int svc_rdma_recvfrom(struct svc_rqst *rqstp)
 		rqstp->rq_arg.head[0].iov_len);
 	rqstp->rq_prot = IPPROTO_MAX;
 	svc_xprt_copy_addrs(rqstp, xprt);
-	svc_xprt_received(xprt);
 	return ret;
 
  close_out:
@@ -678,6 +676,5 @@ int svc_rdma_recvfrom(struct svc_rqst *rqstp)
 	 */
 	set_bit(XPT_CLOSE, &xprt->xpt_flags);
 defer:
-	svc_xprt_received(xprt);
 	return 0;
 }