Merge tag 'nfs-for-3.9-1' of git://git.linux-nfs.org/projects/trondmy/linux-nfs

Pull NFS client bugfixes from Trond Myklebust:

 - Fix an Oops in the pNFS layoutget code

 - Fix a number of NFSv4 and v4.1 state recovery deadlocks and hangs due
   to the interaction of the session drain lock and state management
   locks.

 - Remove task->tk_xprt, which was hiding a lot of RCU dereferencing
   bugs

 - Fix a long standing NFSv3 posix lock recovery bug.

 - Revert commit 324d003b0c ("NFS: add nfs_sb_deactive_async to avoid
   deadlock").  It turned out that the root cause of the deadlock was
   due to interactions with the workqueues that have now been resolved.

* tag 'nfs-for-3.9-1' of git://git.linux-nfs.org/projects/trondmy/linux-nfs: (22 commits)
  NLM: Ensure that we resend all pending blocking locks after a reclaim
  umount oops when remove blocklayoutdriver first
  sunrpc: silence build warning in gss_fill_context
  nfs: remove kfree() redundant null checks
  NFSv4.1: Don't decode skipped layoutgets
  NFSv4.1: Fix bulk recall and destroy of layouts
  NFSv4.1: Fix an ABBA locking issue with session and state serialisation
  NFSv4: Fix a reboot recovery race when opening a file
  NFSv4: Ensure delegation recall and byte range lock removal don't conflict
  NFSv4: Fix up the return values of nfs4_open_delegation_recall
  NFSv4.1: Don't lose locks when a server reboots during delegation return
  NFSv4.1: Prevent deadlocks between state recovery and file locking
  NFSv4: Allow the state manager to mark an open_owner as being recovered
  SUNRPC: Add missing static declaration to _gss_mech_get_by_name
  Revert "NFS: add nfs_sb_deactive_async to avoid deadlock"
  SUNRPC: Nuke the tk_xprt macro
  SUNRPC: Avoid RCU dereferences in the transport bind and connect code
  SUNRPC: Fix an RCU dereference in xprt_reserve
  SUNRPC: Pass pointers to struct rpc_xprt to the congestion window
  SUNRPC: Fix an RCU dereference in xs_local_rpcbind
  ...

fs/lockd/clntproc.c

@@ -550,6 +550,9 @@ again:
 		status = nlmclnt_block(block, req, NLMCLNT_POLL_TIMEOUT);
 		if (status < 0)
 			break;
+		/* Resend the blocking lock request after a server reboot */
+		if (resp->status ==  nlm_lck_denied_grace_period)
+			continue;
 		if (resp->status != nlm_lck_blocked)
 			break;
 	}

fs/nfs/blocklayout/blocklayout.c

@@ -1273,6 +1273,7 @@ static const struct nfs_pageio_ops bl_pg_write_ops = {
 static struct pnfs_layoutdriver_type blocklayout_type = {
 	.id				= LAYOUT_BLOCK_VOLUME,
 	.name				= "LAYOUT_BLOCK_VOLUME",
+	.owner				= THIS_MODULE,
 	.read_pagelist			= bl_read_pagelist,
 	.write_pagelist			= bl_write_pagelist,
 	.alloc_layout_hdr		= bl_alloc_layout_hdr,

fs/nfs/callback_proc.c

@@ -183,60 +183,15 @@ static u32 initiate_file_draining(struct nfs_client *clp,
 static u32 initiate_bulk_draining(struct nfs_client *clp,
 				  struct cb_layoutrecallargs *args)
 {
-	struct nfs_server *server;
-	struct pnfs_layout_hdr *lo;
-	struct inode *ino;
-	u32 rv = NFS4ERR_NOMATCHING_LAYOUT;
-	struct pnfs_layout_hdr *tmp;
-	LIST_HEAD(recall_list);
-	LIST_HEAD(free_me_list);
-	struct pnfs_layout_range range = {
-		.iomode = IOMODE_ANY,
-		.offset = 0,
-		.length = NFS4_MAX_UINT64,
-	};
+	int stat;
 
-	spin_lock(&clp->cl_lock);
-	rcu_read_lock();
-	list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link) {
-		if ((args->cbl_recall_type == RETURN_FSID) &&
-		    memcmp(&server->fsid, &args->cbl_fsid,
-			   sizeof(struct nfs_fsid)))
-			continue;
-
-		list_for_each_entry(lo, &server->layouts, plh_layouts) {
-			ino = igrab(lo->plh_inode);
-			if (!ino)
-				continue;
-			spin_lock(&ino->i_lock);
-			/* Is this layout in the process of being freed? */
-			if (NFS_I(ino)->layout != lo) {
-				spin_unlock(&ino->i_lock);
-				iput(ino);
-				continue;
-			}
-			pnfs_get_layout_hdr(lo);
-			spin_unlock(&ino->i_lock);
-			list_add(&lo->plh_bulk_recall, &recall_list);
-		}
-	}
-	rcu_read_unlock();
-	spin_unlock(&clp->cl_lock);
-
-	list_for_each_entry_safe(lo, tmp,
-				 &recall_list, plh_bulk_recall) {
-		ino = lo->plh_inode;
-		spin_lock(&ino->i_lock);
-		set_bit(NFS_LAYOUT_BULK_RECALL, &lo->plh_flags);
-		if (pnfs_mark_matching_lsegs_invalid(lo, &free_me_list, &range))
-			rv = NFS4ERR_DELAY;
-		list_del_init(&lo->plh_bulk_recall);
-		spin_unlock(&ino->i_lock);
-		pnfs_free_lseg_list(&free_me_list);
-		pnfs_put_layout_hdr(lo);
-		iput(ino);
-	}
-	return rv;
+	if (args->cbl_recall_type == RETURN_FSID)
+		stat = pnfs_destroy_layouts_byfsid(clp, &args->cbl_fsid, true);
+	else
+		stat = pnfs_destroy_layouts_byclid(clp, true);
+	if (stat != 0)
+		return NFS4ERR_DELAY;
+	return NFS4ERR_NOMATCHING_LAYOUT;
 }
 
 static u32 do_callback_layoutrecall(struct nfs_client *clp,

fs/nfs/delegation.c

@@ -55,7 +55,8 @@ int nfs4_have_delegation(struct inode *inode, fmode_t flags)
 	flags &= FMODE_READ|FMODE_WRITE;
 	rcu_read_lock();
 	delegation = rcu_dereference(NFS_I(inode)->delegation);
-	if (delegation != NULL && (delegation->type & flags) == flags) {
+	if (delegation != NULL && (delegation->type & flags) == flags &&
+	    !test_bit(NFS_DELEGATION_RETURNING, &delegation->flags)) {
 		nfs_mark_delegation_referenced(delegation);
 		ret = 1;
 	}
@@ -70,8 +71,10 @@ static int nfs_delegation_claim_locks(struct nfs_open_context *ctx, struct nfs4_
 	int status = 0;
 
 	if (inode->i_flock == NULL)
-		goto out;
+		return 0;
 
+	if (inode->i_flock == NULL)
+		goto out;
 	/* Protect inode->i_flock using the file locks lock */
 	lock_flocks();
 	for (fl = inode->i_flock; fl != NULL; fl = fl->fl_next) {
@@ -94,7 +97,9 @@ static int nfs_delegation_claim_opens(struct inode *inode, const nfs4_stateid *s
 {
 	struct nfs_inode *nfsi = NFS_I(inode);
 	struct nfs_open_context *ctx;
+	struct nfs4_state_owner *sp;
 	struct nfs4_state *state;
+	unsigned int seq;
 	int err;
 
 again:
@@ -109,9 +114,16 @@ again:
 			continue;
 		get_nfs_open_context(ctx);
 		spin_unlock(&inode->i_lock);
+		sp = state->owner;
+		/* Block nfs4_proc_unlck */
+		mutex_lock(&sp->so_delegreturn_mutex);
+		seq = raw_seqcount_begin(&sp->so_reclaim_seqcount);
 		err = nfs4_open_delegation_recall(ctx, state, stateid);
-		if (err >= 0)
+		if (!err)
 			err = nfs_delegation_claim_locks(ctx, state);
+		if (!err && read_seqcount_retry(&sp->so_reclaim_seqcount, seq))
+			err = -EAGAIN;
+		mutex_unlock(&sp->so_delegreturn_mutex);
 		put_nfs_open_context(ctx);
 		if (err != 0)
 			return err;
@@ -182,39 +194,91 @@ static struct inode *nfs_delegation_grab_inode(struct nfs_delegation *delegation
 }
 
 static struct nfs_delegation *
-nfs_detach_delegation_locked(struct nfs_inode *nfsi,
-			     struct nfs_server *server)
+nfs_start_delegation_return_locked(struct nfs_inode *nfsi)
 {
-	struct nfs_delegation *delegation =
-		rcu_dereference_protected(nfsi->delegation,
-				lockdep_is_held(&server->nfs_client->cl_lock));
+	struct nfs_delegation *ret = NULL;
+	struct nfs_delegation *delegation = rcu_dereference(nfsi->delegation);
 
 	if (delegation == NULL)
-		goto nomatch;
+		goto out;
+	spin_lock(&delegation->lock);
+	if (!test_and_set_bit(NFS_DELEGATION_RETURNING, &delegation->flags))
+		ret = delegation;
+	spin_unlock(&delegation->lock);
+out:
+	return ret;
+}
+
+static struct nfs_delegation *
+nfs_start_delegation_return(struct nfs_inode *nfsi)
+{
+	struct nfs_delegation *delegation;
+
+	rcu_read_lock();
+	delegation = nfs_start_delegation_return_locked(nfsi);
+	rcu_read_unlock();
+	return delegation;
+}
+
+static void
+nfs_abort_delegation_return(struct nfs_delegation *delegation,
+		struct nfs_client *clp)
+{
 
 	spin_lock(&delegation->lock);
+	clear_bit(NFS_DELEGATION_RETURNING, &delegation->flags);
+	set_bit(NFS_DELEGATION_RETURN, &delegation->flags);
+	spin_unlock(&delegation->lock);
+	set_bit(NFS4CLNT_DELEGRETURN, &clp->cl_state);
+}
+
+static struct nfs_delegation *
+nfs_detach_delegation_locked(struct nfs_inode *nfsi,
+		struct nfs_delegation *delegation,
+		struct nfs_client *clp)
+{
+	struct nfs_delegation *deleg_cur =
+		rcu_dereference_protected(nfsi->delegation,
+				lockdep_is_held(&clp->cl_lock));
+
+	if (deleg_cur == NULL || delegation != deleg_cur)
+		return NULL;
+
+	spin_lock(&delegation->lock);
+	set_bit(NFS_DELEGATION_RETURNING, &delegation->flags);
 	list_del_rcu(&delegation->super_list);
 	delegation->inode = NULL;
 	nfsi->delegation_state = 0;
 	rcu_assign_pointer(nfsi->delegation, NULL);
 	spin_unlock(&delegation->lock);
 	return delegation;
-nomatch:
-	return NULL;
 }
 
 static struct nfs_delegation *nfs_detach_delegation(struct nfs_inode *nfsi,
-						    struct nfs_server *server)
+		struct nfs_delegation *delegation,
+		struct nfs_server *server)
 {
 	struct nfs_client *clp = server->nfs_client;
-	struct nfs_delegation *delegation;
 
 	spin_lock(&clp->cl_lock);
-	delegation = nfs_detach_delegation_locked(nfsi, server);
+	delegation = nfs_detach_delegation_locked(nfsi, delegation, clp);
 	spin_unlock(&clp->cl_lock);
 	return delegation;
 }
 
+static struct nfs_delegation *
+nfs_inode_detach_delegation(struct inode *inode)
+{
+	struct nfs_inode *nfsi = NFS_I(inode);
+	struct nfs_server *server = NFS_SERVER(inode);
+	struct nfs_delegation *delegation;
+
+	delegation = nfs_start_delegation_return(nfsi);
+	if (delegation == NULL)
+		return NULL;
+	return nfs_detach_delegation(nfsi, delegation, server);
+}
+
 /**
  * nfs_inode_set_delegation - set up a delegation on an inode
  * @inode: inode to which delegation applies
@@ -268,7 +332,10 @@ int nfs_inode_set_delegation(struct inode *inode, struct rpc_cred *cred, struct
 			delegation = NULL;
 			goto out;
 		}
-		freeme = nfs_detach_delegation_locked(nfsi, server);
+		freeme = nfs_detach_delegation_locked(nfsi, 
+				old_delegation, clp);
+		if (freeme == NULL)
+			goto out;
 	}
 	list_add_rcu(&delegation->super_list, &server->delegations);
 	nfsi->delegation_state = delegation->type;
@@ -292,19 +359,29 @@ out:
 /*
  * Basic procedure for returning a delegation to the server
  */
-static int __nfs_inode_return_delegation(struct inode *inode, struct nfs_delegation *delegation, int issync)
+static int nfs_end_delegation_return(struct inode *inode, struct nfs_delegation *delegation, int issync)
 {
+	struct nfs_client *clp = NFS_SERVER(inode)->nfs_client;
 	struct nfs_inode *nfsi = NFS_I(inode);
 	int err;
 
-	/*
-	 * Guard against new delegated open/lock/unlock calls and against
-	 * state recovery
-	 */
-	down_write(&nfsi->rwsem);
-	err = nfs_delegation_claim_opens(inode, &delegation->stateid);
-	up_write(&nfsi->rwsem);
-	if (err)
+	if (delegation == NULL)
+		return 0;
+	do {
+		err = nfs_delegation_claim_opens(inode, &delegation->stateid);
+		if (!issync || err != -EAGAIN)
+			break;
+		/*
+		 * Guard against state recovery
+		 */
+		err = nfs4_wait_clnt_recover(clp);
+	} while (err == 0);
+
+	if (err) {
+		nfs_abort_delegation_return(delegation, clp);
+		goto out;
+	}
+	if (!nfs_detach_delegation(nfsi, delegation, NFS_SERVER(inode)))
 		goto out;
 
 	err = nfs_do_return_delegation(inode, delegation, issync);
@@ -340,13 +417,10 @@ restart:
 			inode = nfs_delegation_grab_inode(delegation);
 			if (inode == NULL)
 				continue;
-			delegation = nfs_detach_delegation(NFS_I(inode),
-								server);
+			delegation = nfs_start_delegation_return_locked(NFS_I(inode));
 			rcu_read_unlock();
 
-			if (delegation != NULL)
-				err = __nfs_inode_return_delegation(inode,
-								delegation, 0);
+			err = nfs_end_delegation_return(inode, delegation, 0);
 			iput(inode);
 			if (!err)
 				goto restart;
@@ -367,15 +441,11 @@ restart:
  */
 void nfs_inode_return_delegation_noreclaim(struct inode *inode)
 {
-	struct nfs_server *server = NFS_SERVER(inode);
-	struct nfs_inode *nfsi = NFS_I(inode);
 	struct nfs_delegation *delegation;
 
-	if (rcu_access_pointer(nfsi->delegation) != NULL) {
-		delegation = nfs_detach_delegation(nfsi, server);
-		if (delegation != NULL)
-			nfs_do_return_delegation(inode, delegation, 0);
-	}
+	delegation = nfs_inode_detach_delegation(inode);
+	if (delegation != NULL)
+		nfs_do_return_delegation(inode, delegation, 0);
 }
 
 /**
@@ -390,18 +460,14 @@ void nfs_inode_return_delegation_noreclaim(struct inode *inode)
  */
 int nfs4_inode_return_delegation(struct inode *inode)
 {
-	struct nfs_server *server = NFS_SERVER(inode);
 	struct nfs_inode *nfsi = NFS_I(inode);
 	struct nfs_delegation *delegation;
 	int err = 0;
 
 	nfs_wb_all(inode);
-	if (rcu_access_pointer(nfsi->delegation) != NULL) {
-		delegation = nfs_detach_delegation(nfsi, server);
-		if (delegation != NULL) {
-			err = __nfs_inode_return_delegation(inode, delegation, 1);
-		}
-	}
+	delegation = nfs_start_delegation_return(nfsi);
+	if (delegation != NULL)
+		err = nfs_end_delegation_return(inode, delegation, 1);
 	return err;
 }
 
@@ -471,7 +537,7 @@ void nfs_remove_bad_delegation(struct inode *inode)
 {
 	struct nfs_delegation *delegation;
 
-	delegation = nfs_detach_delegation(NFS_I(inode), NFS_SERVER(inode));
+	delegation = nfs_inode_detach_delegation(inode);
 	if (delegation) {
 		nfs_inode_find_state_and_recover(inode, &delegation->stateid);
 		nfs_free_delegation(delegation);
@@ -649,7 +715,7 @@ restart:
 			if (inode == NULL)
 				continue;
 			delegation = nfs_detach_delegation(NFS_I(inode),
-								server);
+					delegation, server);
 			rcu_read_unlock();
 
 			if (delegation != NULL)

fs/nfs/delegation.h

@@ -29,6 +29,7 @@ enum {
 	NFS_DELEGATION_NEED_RECLAIM = 0,
 	NFS_DELEGATION_RETURN,
 	NFS_DELEGATION_REFERENCED,
+	NFS_DELEGATION_RETURNING,
 };
 
 int nfs_inode_set_delegation(struct inode *inode, struct rpc_cred *cred, struct nfs_openres *res);

fs/nfs/getroot.c

@@ -126,8 +126,7 @@ struct dentry *nfs_get_root(struct super_block *sb, struct nfs_fh *mntfh,
 	}
 	spin_unlock(&ret->d_lock);
 out:
-	if (name)
-		kfree(name);
+	kfree(name);
 	nfs_free_fattr(fsinfo.fattr);
 	return ret;
 }

fs/nfs/inode.c

@@ -694,10 +694,7 @@ static void __put_nfs_open_context(struct nfs_open_context *ctx, int is_sync)
 	if (ctx->cred != NULL)
 		put_rpccred(ctx->cred);
 	dput(ctx->dentry);
-	if (is_sync)
-		nfs_sb_deactive(sb);
-	else
-		nfs_sb_deactive_async(sb);
+	nfs_sb_deactive(sb);
 	kfree(ctx->mdsthreshold);
 	kfree(ctx);
 }

fs/nfs/internal.h

@@ -329,7 +329,6 @@ extern int __init register_nfs_fs(void);
 extern void __exit unregister_nfs_fs(void);
 extern void nfs_sb_active(struct super_block *sb);
 extern void nfs_sb_deactive(struct super_block *sb);
-extern void nfs_sb_deactive_async(struct super_block *sb);
 
 /* namespace.c */
 #define NFS_PATH_CANONICAL 1

fs/nfs/nfs4_fs.h

@@ -13,6 +13,8 @@
 
 #define NFS4_MAX_LOOP_ON_RECOVER (10)
 
+#include <linux/seqlock.h>
+
 struct idmap;
 
 enum nfs4_client_state {
@@ -90,6 +92,8 @@ struct nfs4_state_owner {
 	unsigned long	     so_flags;
 	struct list_head     so_states;
 	struct nfs_seqid_counter so_seqid;
+	seqcount_t	     so_reclaim_seqcount;
+	struct mutex	     so_delegreturn_mutex;
 };
 
 enum {

fs/nfs/nfs4proc.c

@@ -896,6 +896,8 @@ static int can_open_delegated(struct nfs_delegation *delegation, fmode_t fmode)
 		return 0;
 	if (test_bit(NFS_DELEGATION_NEED_RECLAIM, &delegation->flags))
 		return 0;
+	if (test_bit(NFS_DELEGATION_RETURNING, &delegation->flags))
+		return 0;
 	nfs_mark_delegation_referenced(delegation);
 	return 1;
 }
@@ -973,6 +975,7 @@ static int update_open_stateid(struct nfs4_state *state, nfs4_stateid *open_stat
 
 	spin_lock(&deleg_cur->lock);
 	if (nfsi->delegation != deleg_cur ||
+	   test_bit(NFS_DELEGATION_RETURNING, &deleg_cur->flags) ||
 	    (deleg_cur->type & fmode) != fmode)
 		goto no_delegation_unlock;
 
@@ -1352,19 +1355,18 @@ int nfs4_open_delegation_recall(struct nfs_open_context *ctx, struct nfs4_state
 			case -NFS4ERR_BAD_HIGH_SLOT:
 			case -NFS4ERR_CONN_NOT_BOUND_TO_SESSION:
 			case -NFS4ERR_DEADSESSION:
+				set_bit(NFS_DELEGATED_STATE, &state->flags);
 				nfs4_schedule_session_recovery(server->nfs_client->cl_session, err);
+				err = -EAGAIN;
 				goto out;
 			case -NFS4ERR_STALE_CLIENTID:
 			case -NFS4ERR_STALE_STATEID:
+				set_bit(NFS_DELEGATED_STATE, &state->flags);
 			case -NFS4ERR_EXPIRED:
 				/* Don't recall a delegation if it was lost */
 				nfs4_schedule_lease_recovery(server->nfs_client);
+				err = -EAGAIN;
 				goto out;
-			case -ERESTARTSYS:
-				/*
-				 * The show must go on: exit, but mark the
-				 * stateid as needing recovery.
-				 */
 			case -NFS4ERR_DELEG_REVOKED:
 			case -NFS4ERR_ADMIN_REVOKED:
 			case -NFS4ERR_BAD_STATEID:
@@ -1375,6 +1377,7 @@ int nfs4_open_delegation_recall(struct nfs_open_context *ctx, struct nfs4_state
 				err = 0;
 				goto out;
 		}
+		set_bit(NFS_DELEGATED_STATE, &state->flags);
 		err = nfs4_handle_exception(server, err, &exception);
 	} while (exception.retry);
 out:
@@ -1463,7 +1466,7 @@ static void nfs4_open_prepare(struct rpc_task *task, void *calldata)
 	struct nfs4_state_owner *sp = data->owner;
 
 	if (nfs_wait_on_sequence(data->o_arg.seqid, task) != 0)
-		return;
+		goto out_wait;
 	/*
 	 * Check if we still need to send an OPEN call, or if we can use
 	 * a delegation instead.
@@ -1498,6 +1501,7 @@ unlock_no_action:
 	rcu_read_unlock();
 out_no_action:
 	task->tk_action = NULL;
+out_wait:
 	nfs4_sequence_done(task, &data->o_res.seq_res);
 }
 
@@ -1845,6 +1849,43 @@ static inline void nfs4_exclusive_attrset(struct nfs4_opendata *opendata, struct
 		sattr->ia_valid |= ATTR_MTIME;
 }
 
+static int _nfs4_open_and_get_state(struct nfs4_opendata *opendata,
+		fmode_t fmode,
+		int flags,
+		struct nfs4_state **res)
+{
+	struct nfs4_state_owner *sp = opendata->owner;
+	struct nfs_server *server = sp->so_server;
+	struct nfs4_state *state;
+	unsigned int seq;
+	int ret;
+
+	seq = raw_seqcount_begin(&sp->so_reclaim_seqcount);
+
+	ret = _nfs4_proc_open(opendata);
+	if (ret != 0)
+		goto out;
+
+	state = nfs4_opendata_to_nfs4_state(opendata);
+	ret = PTR_ERR(state);
+	if (IS_ERR(state))
+		goto out;
+	if (server->caps & NFS_CAP_POSIX_LOCK)
+		set_bit(NFS_STATE_POSIX_LOCKS, &state->flags);
+
+	ret = nfs4_opendata_access(sp->so_cred, opendata, state, fmode, flags);
+	if (ret != 0)
+		goto out;
+
+	if (read_seqcount_retry(&sp->so_reclaim_seqcount, seq)) {
+		nfs4_schedule_stateid_recovery(server, state);
+		nfs4_wait_clnt_recover(server->nfs_client);
+	}
+	*res = state;
+out:
+	return ret;
+}
+
 /*
  * Returns a referenced nfs4_state
  */
@@ -1889,18 +1930,7 @@ static int _nfs4_do_open(struct inode *dir,
 	if (dentry->d_inode != NULL)
 		opendata->state = nfs4_get_open_state(dentry->d_inode, sp);
 
-	status = _nfs4_proc_open(opendata);
-	if (status != 0)
-		goto err_opendata_put;
-
-	state = nfs4_opendata_to_nfs4_state(opendata);
-	status = PTR_ERR(state);
-	if (IS_ERR(state))
-		goto err_opendata_put;
-	if (server->caps & NFS_CAP_POSIX_LOCK)
-		set_bit(NFS_STATE_POSIX_LOCKS, &state->flags);
-
-	status = nfs4_opendata_access(cred, opendata, state, fmode, flags);
+	status = _nfs4_open_and_get_state(opendata, fmode, flags, &state);
 	if (status != 0)
 		goto err_opendata_put;
 
@@ -2088,7 +2118,7 @@ static void nfs4_free_closedata(void *data)
 	nfs4_put_open_state(calldata->state);
 	nfs_free_seqid(calldata->arg.seqid);
 	nfs4_put_state_owner(sp);
-	nfs_sb_deactive_async(sb);
+	nfs_sb_deactive(sb);
 	kfree(calldata);
 }
 
@@ -2150,7 +2180,7 @@ static void nfs4_close_prepare(struct rpc_task *task, void *data)
 
 	dprintk("%s: begin!\n", __func__);
 	if (nfs_wait_on_sequence(calldata->arg.seqid, task) != 0)
-		return;
+		goto out_wait;
 
 	task->tk_msg.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_OPEN_DOWNGRADE];
 	calldata->arg.fmode = FMODE_READ|FMODE_WRITE;
@@ -2172,16 +2202,14 @@ static void nfs4_close_prepare(struct rpc_task *task, void *data)
 
 	if (!call_close) {
 		/* Note: exit _without_ calling nfs4_close_done */
-		task->tk_action = NULL;
-		nfs4_sequence_done(task, &calldata->res.seq_res);
-		goto out;
+		goto out_no_action;
 	}
 
 	if (calldata->arg.fmode == 0) {
 		task->tk_msg.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_CLOSE];
 		if (calldata->roc &&
 		    pnfs_roc_drain(inode, &calldata->roc_barrier, task))
-			goto out;
+			goto out_wait;
 	}
 
 	nfs_fattr_init(calldata->res.fattr);
@@ -2191,8 +2219,12 @@ static void nfs4_close_prepare(struct rpc_task *task, void *data)
 				&calldata->res.seq_res,
 				task) != 0)
 		nfs_release_seqid(calldata->arg.seqid);
-out:
 	dprintk("%s: done!\n", __func__);
+	return;
+out_no_action:
+	task->tk_action = NULL;
+out_wait:
+	nfs4_sequence_done(task, &calldata->res.seq_res);
 }
 
 static const struct rpc_call_ops nfs4_close_ops = {
@@ -4423,12 +4455,10 @@ static void nfs4_locku_prepare(struct rpc_task *task, void *data)
 	struct nfs4_unlockdata *calldata = data;
 
 	if (nfs_wait_on_sequence(calldata->arg.seqid, task) != 0)
-		return;
+		goto out_wait;
 	if (test_bit(NFS_LOCK_INITIALIZED, &calldata->lsp->ls_flags) == 0) {
 		/* Note: exit _without_ running nfs4_locku_done */
-		task->tk_action = NULL;
-		nfs4_sequence_done(task, &calldata->res.seq_res);
-		return;
+		goto out_no_action;
 	}
 	calldata->timestamp = jiffies;
 	if (nfs4_setup_sequence(calldata->server,
@@ -4436,6 +4466,11 @@ static void nfs4_locku_prepare(struct rpc_task *task, void *data)
 				&calldata->res.seq_res,
 				task) != 0)
 		nfs_release_seqid(calldata->arg.seqid);
+	return;
+out_no_action:
+	task->tk_action = NULL;
+out_wait:
+	nfs4_sequence_done(task, &calldata->res.seq_res);
 }
 
 static const struct rpc_call_ops nfs4_locku_ops = {
@@ -4482,7 +4517,9 @@ static struct rpc_task *nfs4_do_unlck(struct file_lock *fl,
 
 static int nfs4_proc_unlck(struct nfs4_state *state, int cmd, struct file_lock *request)
 {
-	struct nfs_inode *nfsi = NFS_I(state->inode);
+	struct inode *inode = state->inode;
+	struct nfs4_state_owner *sp = state->owner;
+	struct nfs_inode *nfsi = NFS_I(inode);
 	struct nfs_seqid *seqid;
 	struct nfs4_lock_state *lsp;
 	struct rpc_task *task;
@@ -4492,12 +4529,17 @@ static int nfs4_proc_unlck(struct nfs4_state *state, int cmd, struct file_lock *
 	status = nfs4_set_lock_state(state, request);
 	/* Unlock _before_ we do the RPC call */
 	request->fl_flags |= FL_EXISTS;
+	/* Exclude nfs_delegation_claim_locks() */
+	mutex_lock(&sp->so_delegreturn_mutex);
+	/* Exclude nfs4_reclaim_open_stateid() - note nesting! */
 	down_read(&nfsi->rwsem);
 	if (do_vfs_lock(request->fl_file, request) == -ENOENT) {
 		up_read(&nfsi->rwsem);
+		mutex_unlock(&sp->so_delegreturn_mutex);
 		goto out;
 	}
 	up_read(&nfsi->rwsem);
+	mutex_unlock(&sp->so_delegreturn_mutex);
 	if (status != 0)
 		goto out;
 	/* Is this a delegated lock? */
@@ -4576,7 +4618,7 @@ static void nfs4_lock_prepare(struct rpc_task *task, void *calldata)
 
 	dprintk("%s: begin!\n", __func__);
 	if (nfs_wait_on_sequence(data->arg.lock_seqid, task) != 0)
-		return;
+		goto out_wait;
 	/* Do we need to do an open_to_lock_owner? */
 	if (!(data->arg.lock_seqid->sequence->flags & NFS_SEQID_CONFIRMED)) {
 		if (nfs_wait_on_sequence(data->arg.open_seqid, task) != 0) {
@@ -4596,6 +4638,8 @@ static void nfs4_lock_prepare(struct rpc_task *task, void *calldata)
 	nfs_release_seqid(data->arg.open_seqid);
 out_release_lock_seqid:
 	nfs_release_seqid(data->arg.lock_seqid);
+out_wait:
+	nfs4_sequence_done(task, &data->res.seq_res);
 	dprintk("%s: done!, ret = %d\n", __func__, data->rpc_status);
 }
 
@@ -4813,8 +4857,10 @@ static int nfs41_lock_expired(struct nfs4_state *state, struct file_lock *reques
 
 static int _nfs4_proc_setlk(struct nfs4_state *state, int cmd, struct file_lock *request)
 {
+	struct nfs4_state_owner *sp = state->owner;
 	struct nfs_inode *nfsi = NFS_I(state->inode);
 	unsigned char fl_flags = request->fl_flags;
+	unsigned int seq;
 	int status = -ENOLCK;
 
 	if ((fl_flags & FL_POSIX) &&
@@ -4836,9 +4882,16 @@ static int _nfs4_proc_setlk(struct nfs4_state *state, int cmd, struct file_lock
 		status = do_vfs_lock(request->fl_file, request);
 		goto out_unlock;
 	}
+	seq = raw_seqcount_begin(&sp->so_reclaim_seqcount);
+	up_read(&nfsi->rwsem);
 	status = _nfs4_do_setlk(state, cmd, request, NFS_LOCK_NEW);
 	if (status != 0)
+		goto out;
+	down_read(&nfsi->rwsem);
+	if (read_seqcount_retry(&sp->so_reclaim_seqcount, seq)) {
+		status = -NFS4ERR_DELAY;
 		goto out_unlock;
+	}
 	/* Note: we always want to sleep here! */
 	request->fl_flags = fl_flags | FL_SLEEP;
 	if (do_vfs_lock(request->fl_file, request) < 0)
@@ -4945,24 +4998,22 @@ int nfs4_lock_delegation_recall(struct nfs4_state *state, struct file_lock *fl)
 			case 0:
 			case -ESTALE:
 				goto out;
-			case -NFS4ERR_EXPIRED:
-				nfs4_schedule_stateid_recovery(server, state);
 			case -NFS4ERR_STALE_CLIENTID:
 			case -NFS4ERR_STALE_STATEID:
+				set_bit(NFS_DELEGATED_STATE, &state->flags);
+			case -NFS4ERR_EXPIRED:
 				nfs4_schedule_lease_recovery(server->nfs_client);
+				err = -EAGAIN;
 				goto out;
 			case -NFS4ERR_BADSESSION:
 			case -NFS4ERR_BADSLOT:
 			case -NFS4ERR_BAD_HIGH_SLOT:
 			case -NFS4ERR_CONN_NOT_BOUND_TO_SESSION:
 			case -NFS4ERR_DEADSESSION:
+				set_bit(NFS_DELEGATED_STATE, &state->flags);
 				nfs4_schedule_session_recovery(server->nfs_client->cl_session, err);
+				err = -EAGAIN;
 				goto out;
-			case -ERESTARTSYS:
-				/*
-				 * The show must go on: exit, but mark the
-				 * stateid as needing recovery.
-				 */
 			case -NFS4ERR_DELEG_REVOKED:
 			case -NFS4ERR_ADMIN_REVOKED:
 			case -NFS4ERR_BAD_STATEID:
@@ -4975,9 +5026,8 @@ int nfs4_lock_delegation_recall(struct nfs4_state *state, struct file_lock *fl)
 				/* kill_proc(fl->fl_pid, SIGLOST, 1); */
 				err = 0;
 				goto out;
-			case -NFS4ERR_DELAY:
-				break;
 		}
+		set_bit(NFS_DELEGATED_STATE, &state->flags);
 		err = nfs4_handle_exception(server, err, &exception);
 	} while (exception.retry);
 out:
@@ -6134,7 +6184,8 @@ nfs4_proc_layoutget(struct nfs4_layoutget *lgp, gfp_t gfp_flags)
 	status = nfs4_wait_for_completion_rpc_task(task);
 	if (status == 0)
 		status = task->tk_status;
-	if (status == 0)
+	/* if layoutp->len is 0, nfs4_layoutget_prepare called rpc_exit */
+	if (status == 0 && lgp->res.layoutp->len)
 		lseg = pnfs_layout_process(lgp);
 	rpc_put_task(task);
 	dprintk("<-- %s status=%d\n", __func__, status);

fs/nfs/nfs4state.c

@@ -518,6 +518,8 @@ nfs4_alloc_state_owner(struct nfs_server *server,
 	nfs4_init_seqid_counter(&sp->so_seqid);
 	atomic_set(&sp->so_count, 1);
 	INIT_LIST_HEAD(&sp->so_lru);
+	seqcount_init(&sp->so_reclaim_seqcount);
+	mutex_init(&sp->so_delegreturn_mutex);
 	return sp;
 }
 
@@ -1390,8 +1392,9 @@ static int nfs4_reclaim_open_state(struct nfs4_state_owner *sp, const struct nfs
 	 * recovering after a network partition or a reboot from a
 	 * server that doesn't support a grace period.
 	 */
-restart:
 	spin_lock(&sp->so_lock);
+	write_seqcount_begin(&sp->so_reclaim_seqcount);
+restart:
 	list_for_each_entry(state, &sp->so_states, open_states) {
 		if (!test_and_clear_bit(ops->state_flag_bit, &state->flags))
 			continue;
@@ -1412,6 +1415,7 @@ restart:
 				}
 				spin_unlock(&state->state_lock);
 				nfs4_put_open_state(state);
+				spin_lock(&sp->so_lock);
 				goto restart;
 			}
 		}
@@ -1449,12 +1453,17 @@ restart:
 				goto out_err;
 		}
 		nfs4_put_open_state(state);
+		spin_lock(&sp->so_lock);
 		goto restart;
 	}
+	write_seqcount_end(&sp->so_reclaim_seqcount);
 	spin_unlock(&sp->so_lock);
 	return 0;
 out_err:
 	nfs4_put_open_state(state);
+	spin_lock(&sp->so_lock);
+	write_seqcount_end(&sp->so_reclaim_seqcount);
+	spin_unlock(&sp->so_lock);
 	return status;
 }
 

fs/nfs/objlayout/objio_osd.c

@@ -647,6 +647,7 @@ static struct pnfs_layoutdriver_type objlayout_type = {
 	.flags                   = PNFS_LAYOUTRET_ON_SETATTR |
 				   PNFS_LAYOUTRET_ON_ERROR,
 
+	.owner		       	 = THIS_MODULE,
 	.alloc_layout_hdr        = objlayout_alloc_layout_hdr,
 	.free_layout_hdr         = objlayout_free_layout_hdr,
 

fs/nfs/pnfs.c

@@ -505,6 +505,136 @@ pnfs_destroy_layout(struct nfs_inode *nfsi)
 }
 EXPORT_SYMBOL_GPL(pnfs_destroy_layout);
 
+static bool
+pnfs_layout_add_bulk_destroy_list(struct inode *inode,
+		struct list_head *layout_list)
+{
+	struct pnfs_layout_hdr *lo;
+	bool ret = false;
+
+	spin_lock(&inode->i_lock);
+	lo = NFS_I(inode)->layout;
+	if (lo != NULL && list_empty(&lo->plh_bulk_destroy)) {
+		pnfs_get_layout_hdr(lo);
+		list_add(&lo->plh_bulk_destroy, layout_list);
+		ret = true;
+	}
+	spin_unlock(&inode->i_lock);
+	return ret;
+}
+
+/* Caller must hold rcu_read_lock and clp->cl_lock */
+static int
+pnfs_layout_bulk_destroy_byserver_locked(struct nfs_client *clp,
+		struct nfs_server *server,
+		struct list_head *layout_list)
+{
+	struct pnfs_layout_hdr *lo, *next;
+	struct inode *inode;
+
+	list_for_each_entry_safe(lo, next, &server->layouts, plh_layouts) {
+		inode = igrab(lo->plh_inode);
+		if (inode == NULL)
+			continue;
+		list_del_init(&lo->plh_layouts);
+		if (pnfs_layout_add_bulk_destroy_list(inode, layout_list))
+			continue;
+		rcu_read_unlock();
+		spin_unlock(&clp->cl_lock);
+		iput(inode);
+		spin_lock(&clp->cl_lock);
+		rcu_read_lock();
+		return -EAGAIN;
+	}
+	return 0;
+}
+
+static int
+pnfs_layout_free_bulk_destroy_list(struct list_head *layout_list,
+		bool is_bulk_recall)
+{
+	struct pnfs_layout_hdr *lo;
+	struct inode *inode;
+	struct pnfs_layout_range range = {
+		.iomode = IOMODE_ANY,
+		.offset = 0,
+		.length = NFS4_MAX_UINT64,
+	};
+	LIST_HEAD(lseg_list);
+	int ret = 0;
+
+	while (!list_empty(layout_list)) {
+		lo = list_entry(layout_list->next, struct pnfs_layout_hdr,
+				plh_bulk_destroy);
+		dprintk("%s freeing layout for inode %lu\n", __func__,
+			lo->plh_inode->i_ino);
+		inode = lo->plh_inode;
+		spin_lock(&inode->i_lock);
+		list_del_init(&lo->plh_bulk_destroy);
+		lo->plh_block_lgets++; /* permanently block new LAYOUTGETs */
+		if (is_bulk_recall)
+			set_bit(NFS_LAYOUT_BULK_RECALL, &lo->plh_flags);
+		if (pnfs_mark_matching_lsegs_invalid(lo, &lseg_list, &range))
+			ret = -EAGAIN;
+		spin_unlock(&inode->i_lock);
+		pnfs_free_lseg_list(&lseg_list);
+		pnfs_put_layout_hdr(lo);
+		iput(inode);
+	}
+	return ret;
+}
+
+int
+pnfs_destroy_layouts_byfsid(struct nfs_client *clp,
+		struct nfs_fsid *fsid,
+		bool is_recall)
+{
+	struct nfs_server *server;
+	LIST_HEAD(layout_list);
+
+	spin_lock(&clp->cl_lock);
+	rcu_read_lock();
+restart:
+	list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link) {
+		if (memcmp(&server->fsid, fsid, sizeof(*fsid)) != 0)
+			continue;
+		if (pnfs_layout_bulk_destroy_byserver_locked(clp,
+				server,
+				&layout_list) != 0)
+			goto restart;
+	}
+	rcu_read_unlock();
+	spin_unlock(&clp->cl_lock);
+
+	if (list_empty(&layout_list))
+		return 0;
+	return pnfs_layout_free_bulk_destroy_list(&layout_list, is_recall);
+}
+
+int
+pnfs_destroy_layouts_byclid(struct nfs_client *clp,
+		bool is_recall)
+{
+	struct nfs_server *server;
+	LIST_HEAD(layout_list);
+
+	spin_lock(&clp->cl_lock);
+	rcu_read_lock();
+restart:
+	list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link) {
+		if (pnfs_layout_bulk_destroy_byserver_locked(clp,
+					server,
+					&layout_list) != 0)
+			goto restart;
+	}
+	rcu_read_unlock();
+	spin_unlock(&clp->cl_lock);
+
+	if (list_empty(&layout_list))
+		return 0;
+	return pnfs_layout_free_bulk_destroy_list(&layout_list, is_recall);
+}
+
 /*
  * Called by the state manger to remove all layouts established under an
  * expired lease.
@@ -512,30 +642,10 @@ EXPORT_SYMBOL_GPL(pnfs_destroy_layout);
 void
 pnfs_destroy_all_layouts(struct nfs_client *clp)
 {
-	struct nfs_server *server;
-	struct pnfs_layout_hdr *lo;
-	LIST_HEAD(tmp_list);
-
 	nfs4_deviceid_mark_client_invalid(clp);
 	nfs4_deviceid_purge_client(clp);
 
-	spin_lock(&clp->cl_lock);
-	rcu_read_lock();
-	list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link) {
-		if (!list_empty(&server->layouts))
-			list_splice_init(&server->layouts, &tmp_list);
-	}
-	rcu_read_unlock();
-	spin_unlock(&clp->cl_lock);
-
-	while (!list_empty(&tmp_list)) {
-		lo = list_entry(tmp_list.next, struct pnfs_layout_hdr,
-				plh_layouts);
-		dprintk("%s freeing layout for inode %lu\n", __func__,
-			lo->plh_inode->i_ino);
-		list_del_init(&lo->plh_layouts);
-		pnfs_destroy_layout(NFS_I(lo->plh_inode));
-	}
+	pnfs_destroy_layouts_byclid(clp, false);
 }
 
 /*
@@ -888,7 +998,7 @@ alloc_init_layout_hdr(struct inode *ino,
 	atomic_set(&lo->plh_refcount, 1);
 	INIT_LIST_HEAD(&lo->plh_layouts);
 	INIT_LIST_HEAD(&lo->plh_segs);
-	INIT_LIST_HEAD(&lo->plh_bulk_recall);
+	INIT_LIST_HEAD(&lo->plh_bulk_destroy);
 	lo->plh_inode = ino;
 	lo->plh_lc_cred = get_rpccred(ctx->state->owner->so_cred);
 	return lo;

fs/nfs/pnfs.h

@@ -132,7 +132,7 @@ struct pnfs_layoutdriver_type {
 struct pnfs_layout_hdr {
 	atomic_t		plh_refcount;
 	struct list_head	plh_layouts;   /* other client layouts */
-	struct list_head	plh_bulk_recall; /* clnt list of bulk recalls */
+	struct list_head	plh_bulk_destroy;
 	struct list_head	plh_segs;      /* layout segments list */
 	nfs4_stateid		plh_stateid;
 	atomic_t		plh_outstanding; /* number of RPCs out */
@@ -196,6 +196,11 @@ struct pnfs_layout_segment *pnfs_layout_process(struct nfs4_layoutget *lgp);
 void pnfs_free_lseg_list(struct list_head *tmp_list);
 void pnfs_destroy_layout(struct nfs_inode *);
 void pnfs_destroy_all_layouts(struct nfs_client *);
+int pnfs_destroy_layouts_byfsid(struct nfs_client *clp,
+		struct nfs_fsid *fsid,
+		bool is_recall);
+int pnfs_destroy_layouts_byclid(struct nfs_client *clp,
+		bool is_recall);
 void pnfs_put_layout_hdr(struct pnfs_layout_hdr *lo);
 void pnfs_set_layout_stateid(struct pnfs_layout_hdr *lo,
 			     const nfs4_stateid *new,

fs/nfs/super.c

@@ -54,7 +54,6 @@
 #include <linux/parser.h>
 #include <linux/nsproxy.h>
 #include <linux/rcupdate.h>
-#include <linux/kthread.h>
 
 #include <asm/uaccess.h>
 
@@ -418,54 +417,6 @@ void nfs_sb_deactive(struct super_block *sb)
 }
 EXPORT_SYMBOL_GPL(nfs_sb_deactive);
 
-static int nfs_deactivate_super_async_work(void *ptr)
-{
-	struct super_block *sb = ptr;
-
-	deactivate_super(sb);
-	module_put_and_exit(0);
-	return 0;
-}
-
-/*
- * same effect as deactivate_super, but will do final unmount in kthread
- * context
- */
-static void nfs_deactivate_super_async(struct super_block *sb)
-{
-	struct task_struct *task;
-	char buf[INET6_ADDRSTRLEN + 1];
-	struct nfs_server *server = NFS_SB(sb);
-	struct nfs_client *clp = server->nfs_client;
-
-	if (!atomic_add_unless(&sb->s_active, -1, 1)) {
-		rcu_read_lock();
-		snprintf(buf, sizeof(buf),
-			rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_ADDR));
-		rcu_read_unlock();
-
-		__module_get(THIS_MODULE);
-		task = kthread_run(nfs_deactivate_super_async_work, sb,
-				"%s-deactivate-super", buf);
-		if (IS_ERR(task)) {
-			pr_err("%s: kthread_run: %ld\n",
-				__func__, PTR_ERR(task));
-			/* make synchronous call and hope for the best */
-			deactivate_super(sb);
-			module_put(THIS_MODULE);
-		}
-	}
-}
-
-void nfs_sb_deactive_async(struct super_block *sb)
-{
-	struct nfs_server *server = NFS_SB(sb);
-
-	if (atomic_dec_and_test(&server->active))
-		nfs_deactivate_super_async(sb);
-}
-EXPORT_SYMBOL_GPL(nfs_sb_deactive_async);
-
 /*
  * Deliver file system statistics to userspace
  */

fs/nfs/unlink.c

@@ -95,7 +95,7 @@ static void nfs_async_unlink_release(void *calldata)
 
 	nfs_dec_sillycount(data->dir);
 	nfs_free_unlinkdata(data);
-	nfs_sb_deactive_async(sb);
+	nfs_sb_deactive(sb);
 }
 
 static void nfs_unlink_prepare(struct rpc_task *task, void *calldata)
@@ -268,8 +268,7 @@ nfs_async_unlink(struct inode *dir, struct dentry *dentry)
 	 * point dentry is definitely not a root, so we won't need
 	 * that anymore.
 	 */
-	if (devname_garbage)
-		kfree(devname_garbage);
+	kfree(devname_garbage);
 	return 0;
 out_unlock:
 	spin_unlock(&dentry->d_lock);