Merge branch 'nfs-for-3.1' of git://git.linux-nfs.org/projects/trondmy/linux-nfs

* 'nfs-for-3.1' of git://git.linux-nfs.org/projects/trondmy/linux-nfs: (44 commits)
  NFSv4: Don't use the delegation->inode in nfs_mark_return_delegation()
  nfs: don't use d_move in nfs_async_rename_done
  RDMA: Increasing RPCRDMA_MAX_DATA_SEGS
  SUNRPC: Replace xprt->resend and xprt->sending with a priority queue
  SUNRPC: Allow caller of rpc_sleep_on() to select priority levels
  SUNRPC: Support dynamic slot allocation for TCP connections
  SUNRPC: Clean up the slot table allocation
  SUNRPC: Initalise the struct xprt upon allocation
  SUNRPC: Ensure that we grab the XPRT_LOCK before calling xprt_alloc_slot
  pnfs: simplify pnfs files module autoloading
  nfs: document nfsv4 sillyrename issues
  NFS: Convert nfs4_set_ds_client to EXPORT_SYMBOL_GPL
  SUNRPC: Convert the backchannel exports to EXPORT_SYMBOL_GPL
  SUNRPC: sunrpc should not explicitly depend on NFS config options
  NFS: Clean up - simplify the switch to read/write-through-MDS
  NFS: Move the pnfs write code into pnfs.c
  NFS: Move the pnfs read code into pnfs.c
  NFS: Allow the nfs_pageio_descriptor to signal that a re-coalesce is needed
  NFS: Use the nfs_pageio_descriptor->pg_bsize in the read/write request
  NFS: Cache rpc_ops in struct nfs_pageio_descriptor
  ...

fs/lockd/clntproc.c

@@ -302,7 +302,8 @@ nlmclnt_call(struct rpc_cred *cred, struct nlm_rqst *req, u32 proc)
 				/* We appear to be out of the grace period */
 				wake_up_all(&host->h_gracewait);
 			}
-			dprintk("lockd: server returns status %d\n", resp->status);
+			dprintk("lockd: server returns status %d\n",
+				ntohl(resp->status));
 			return 0;	/* Okay, call complete */
 		}
 
@@ -690,7 +691,8 @@ nlmclnt_unlock(struct nlm_rqst *req, struct file_lock *fl)
 		goto out;
 
 	if (resp->status != nlm_lck_denied_nolocks)
-		printk("lockd: unexpected unlock status: %d\n", resp->status);
+		printk("lockd: unexpected unlock status: %d\n",
+			ntohl(resp->status));
 	/* What to do now? I'm out of my depth... */
 	status = -ENOLCK;
 out:
@@ -843,6 +845,7 @@ nlm_stat_to_errno(__be32 status)
 		return -ENOLCK;
 #endif
 	}
-	printk(KERN_NOTICE "lockd: unexpected server status %d\n", status);
+	printk(KERN_NOTICE "lockd: unexpected server status %d\n",
+		 ntohl(status));
 	return -ENOLCK;
 }

fs/nfs/Kconfig

@@ -77,6 +77,7 @@ config NFS_V4
 config NFS_V4_1
 	bool "NFS client support for NFSv4.1 (EXPERIMENTAL)"
 	depends on NFS_FS && NFS_V4 && EXPERIMENTAL
+	select SUNRPC_BACKCHANNEL
 	select PNFS_FILE_LAYOUT
 	help
 	  This option enables support for minor version 1 of the NFSv4 protocol

fs/nfs/callback_proc.c

@@ -111,6 +111,7 @@ int nfs4_validate_delegation_stateid(struct nfs_delegation *delegation, const nf
 static u32 initiate_file_draining(struct nfs_client *clp,
 				  struct cb_layoutrecallargs *args)
 {
+	struct nfs_server *server;
 	struct pnfs_layout_hdr *lo;
 	struct inode *ino;
 	bool found = false;
@@ -118,21 +119,28 @@ static u32 initiate_file_draining(struct nfs_client *clp,
 	LIST_HEAD(free_me_list);
 
 	spin_lock(&clp->cl_lock);
-	list_for_each_entry(lo, &clp->cl_layouts, plh_layouts) {
-		if (nfs_compare_fh(&args->cbl_fh,
-				   &NFS_I(lo->plh_inode)->fh))
-			continue;
-		ino = igrab(lo->plh_inode);
-		if (!ino)
-			continue;
-		found = true;
-		/* Without this, layout can be freed as soon
-		 * as we release cl_lock.
-		 */
-		get_layout_hdr(lo);
-		break;
+	rcu_read_lock();
+	list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link) {
+		list_for_each_entry(lo, &server->layouts, plh_layouts) {
+			if (nfs_compare_fh(&args->cbl_fh,
+					   &NFS_I(lo->plh_inode)->fh))
+				continue;
+			ino = igrab(lo->plh_inode);
+			if (!ino)
+				continue;
+			found = true;
+			/* Without this, layout can be freed as soon
+			 * as we release cl_lock.
+			 */
+			get_layout_hdr(lo);
+			break;
+		}
+		if (found)
+			break;
 	}
+	rcu_read_unlock();
 	spin_unlock(&clp->cl_lock);
+
 	if (!found)
 		return NFS4ERR_NOMATCHING_LAYOUT;
 
@@ -154,6 +162,7 @@ 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;
@@ -167,18 +176,24 @@ static u32 initiate_bulk_draining(struct nfs_client *clp,
 	};
 
 	spin_lock(&clp->cl_lock);
-	list_for_each_entry(lo, &clp->cl_layouts, plh_layouts) {
+	rcu_read_lock();
+	list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link) {
 		if ((args->cbl_recall_type == RETURN_FSID) &&
-		    memcmp(&NFS_SERVER(lo->plh_inode)->fsid,
-			   &args->cbl_fsid, sizeof(struct nfs_fsid)))
-			continue;
-		if (!igrab(lo->plh_inode))
+		    memcmp(&server->fsid, &args->cbl_fsid,
+			   sizeof(struct nfs_fsid)))
 			continue;
-		get_layout_hdr(lo);
-		BUG_ON(!list_empty(&lo->plh_bulk_recall));
-		list_add(&lo->plh_bulk_recall, &recall_list);
+
+		list_for_each_entry(lo, &server->layouts, plh_layouts) {
+			if (!igrab(lo->plh_inode))
+				continue;
+			get_layout_hdr(lo);
+			BUG_ON(!list_empty(&lo->plh_bulk_recall));
+			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;

fs/nfs/client.c

@@ -188,9 +188,6 @@ static struct nfs_client *nfs_alloc_client(const struct nfs_client_initdata *cl_
 	cred = rpc_lookup_machine_cred();
 	if (!IS_ERR(cred))
 		clp->cl_machine_cred = cred;
-#if defined(CONFIG_NFS_V4_1)
-	INIT_LIST_HEAD(&clp->cl_layouts);
-#endif
 	nfs_fscache_get_client_cookie(clp);
 
 	return clp;
@@ -293,6 +290,7 @@ static void nfs_free_client(struct nfs_client *clp)
 	nfs4_deviceid_purge_client(clp);
 
 	kfree(clp->cl_hostname);
+	kfree(clp->server_scope);
 	kfree(clp);
 
 	dprintk("<-- nfs_free_client()\n");
@@ -1062,6 +1060,7 @@ static struct nfs_server *nfs_alloc_server(void)
 	INIT_LIST_HEAD(&server->client_link);
 	INIT_LIST_HEAD(&server->master_link);
 	INIT_LIST_HEAD(&server->delegations);
+	INIT_LIST_HEAD(&server->layouts);
 
 	atomic_set(&server->active, 0);
 
@@ -1464,7 +1463,7 @@ struct nfs_client *nfs4_set_ds_client(struct nfs_client* mds_clp,
 	dprintk("<-- %s %p\n", __func__, clp);
 	return clp;
 }
-EXPORT_SYMBOL(nfs4_set_ds_client);
+EXPORT_SYMBOL_GPL(nfs4_set_ds_client);
 
 /*
  * Session has been established, and the client marked ready.

fs/nfs/delegation.c

@@ -398,12 +398,11 @@ int nfs_inode_return_delegation(struct inode *inode)
 	return err;
 }
 
-static void nfs_mark_return_delegation(struct nfs_delegation *delegation)
+static void nfs_mark_return_delegation(struct nfs_server *server,
+		struct nfs_delegation *delegation)
 {
-	struct nfs_client *clp = NFS_SERVER(delegation->inode)->nfs_client;
-
 	set_bit(NFS_DELEGATION_RETURN, &delegation->flags);
-	set_bit(NFS4CLNT_DELEGRETURN, &clp->cl_state);
+	set_bit(NFS4CLNT_DELEGRETURN, &server->nfs_client->cl_state);
 }
 
 /**
@@ -441,7 +440,7 @@ static void nfs_mark_return_all_delegation_types(struct nfs_server *server,
 		if ((delegation->type == (FMODE_READ|FMODE_WRITE)) && !(flags & FMODE_WRITE))
 			continue;
 		if (delegation->type & flags)
-			nfs_mark_return_delegation(delegation);
+			nfs_mark_return_delegation(server, delegation);
 	}
 }
 
@@ -508,7 +507,7 @@ static void nfs_mark_return_unreferenced_delegations(struct nfs_server *server)
 	list_for_each_entry_rcu(delegation, &server->delegations, super_list) {
 		if (test_and_clear_bit(NFS_DELEGATION_REFERENCED, &delegation->flags))
 			continue;
-		nfs_mark_return_delegation(delegation);
+		nfs_mark_return_delegation(server, delegation);
 	}
 }
 
@@ -539,7 +538,8 @@ void nfs_expire_unreferenced_delegations(struct nfs_client *clp)
 int nfs_async_inode_return_delegation(struct inode *inode,
 				      const nfs4_stateid *stateid)
 {
-	struct nfs_client *clp = NFS_SERVER(inode)->nfs_client;
+	struct nfs_server *server = NFS_SERVER(inode);
+	struct nfs_client *clp = server->nfs_client;
 	struct nfs_delegation *delegation;
 
 	rcu_read_lock();
@@ -549,7 +549,7 @@ int nfs_async_inode_return_delegation(struct inode *inode,
 		rcu_read_unlock();
 		return -ENOENT;
 	}
-	nfs_mark_return_delegation(delegation);
+	nfs_mark_return_delegation(server, delegation);
 	rcu_read_unlock();
 
 	nfs_delegation_run_state_manager(clp);

fs/nfs/internal.h

@@ -277,6 +277,9 @@ extern void nfs_sb_deactive(struct super_block *sb);
 extern char *nfs_path(char **p, struct dentry *dentry,
 		      char *buffer, ssize_t buflen);
 extern struct vfsmount *nfs_d_automount(struct path *path);
+#ifdef CONFIG_NFS_V4
+rpc_authflavor_t nfs_find_best_sec(struct nfs4_secinfo_flavors *);
+#endif
 
 /* getroot.c */
 extern struct dentry *nfs_get_root(struct super_block *, struct nfs_fh *,
@@ -288,12 +291,22 @@ extern struct dentry *nfs4_get_root(struct super_block *, struct nfs_fh *,
 extern int nfs4_get_rootfh(struct nfs_server *server, struct nfs_fh *mntfh);
 #endif
 
+struct nfs_pageio_descriptor;
 /* read.c */
 extern int nfs_initiate_read(struct nfs_read_data *data, struct rpc_clnt *clnt,
 			     const struct rpc_call_ops *call_ops);
 extern void nfs_read_prepare(struct rpc_task *task, void *calldata);
+extern int nfs_generic_pagein(struct nfs_pageio_descriptor *desc,
+		struct list_head *head);
+
+extern void nfs_pageio_reset_read_mds(struct nfs_pageio_descriptor *pgio);
+extern void nfs_readdata_release(struct nfs_read_data *rdata);
 
 /* write.c */
+extern int nfs_generic_flush(struct nfs_pageio_descriptor *desc,
+		struct list_head *head);
+extern void nfs_pageio_reset_write_mds(struct nfs_pageio_descriptor *pgio);
+extern void nfs_writedata_release(struct nfs_write_data *wdata);
 extern void nfs_commit_free(struct nfs_write_data *p);
 extern int nfs_initiate_write(struct nfs_write_data *data,
 			      struct rpc_clnt *clnt,

fs/nfs/namespace.c

@@ -119,7 +119,7 @@ Elong:
 }
 
 #ifdef CONFIG_NFS_V4
-static rpc_authflavor_t nfs_find_best_sec(struct nfs4_secinfo_flavors *flavors)
+rpc_authflavor_t nfs_find_best_sec(struct nfs4_secinfo_flavors *flavors)
 {
 	struct gss_api_mech *mech;
 	struct xdr_netobj oid;

fs/nfs/nfs4_fs.h

@@ -48,6 +48,7 @@ enum nfs4_client_state {
 	NFS4CLNT_SESSION_RESET,
 	NFS4CLNT_RECALL_SLOT,
 	NFS4CLNT_LEASE_CONFIRM,
+	NFS4CLNT_SERVER_SCOPE_MISMATCH,
 };
 
 enum nfs4_session_state {
@@ -66,6 +67,8 @@ struct nfs4_minor_version_ops {
 			int cache_reply);
 	int	(*validate_stateid)(struct nfs_delegation *,
 			const nfs4_stateid *);
+	int	(*find_root_sec)(struct nfs_server *, struct nfs_fh *,
+			struct nfs_fsinfo *);
 	const struct nfs4_state_recovery_ops *reboot_recovery_ops;
 	const struct nfs4_state_recovery_ops *nograce_recovery_ops;
 	const struct nfs4_state_maintenance_ops *state_renewal_ops;
@@ -349,6 +352,8 @@ extern void nfs4_schedule_state_manager(struct nfs_client *);
 extern void nfs4_schedule_stateid_recovery(const struct nfs_server *, struct nfs4_state *);
 extern void nfs41_handle_sequence_flag_errors(struct nfs_client *clp, u32 flags);
 extern void nfs41_handle_recall_slot(struct nfs_client *clp);
+extern void nfs41_handle_server_scope(struct nfs_client *,
+				      struct server_scope **);
 extern void nfs4_put_lock_state(struct nfs4_lock_state *lsp);
 extern int nfs4_set_lock_state(struct nfs4_state *state, struct file_lock *fl);
 extern void nfs4_copy_stateid(nfs4_stateid *, struct nfs4_state *, fl_owner_t, pid_t);

fs/nfs/nfs4filelayout.c

@@ -334,6 +334,9 @@ filelayout_read_pagelist(struct nfs_read_data *data)
 		__func__, data->inode->i_ino,
 		data->args.pgbase, (size_t)data->args.count, offset);
 
+	if (test_bit(NFS_DEVICEID_INVALID, &FILELAYOUT_DEVID_NODE(lseg)->flags))
+		return PNFS_NOT_ATTEMPTED;
+
 	/* Retrieve the correct rpc_client for the byte range */
 	j = nfs4_fl_calc_j_index(lseg, offset);
 	idx = nfs4_fl_calc_ds_index(lseg, j);
@@ -344,8 +347,7 @@ filelayout_read_pagelist(struct nfs_read_data *data)
 		set_bit(lo_fail_bit(IOMODE_READ), &lseg->pls_layout->plh_flags);
 		return PNFS_NOT_ATTEMPTED;
 	}
-	dprintk("%s USE DS:ip %x %hu\n", __func__,
-		ntohl(ds->ds_ip_addr), ntohs(ds->ds_port));
+	dprintk("%s USE DS: %s\n", __func__, ds->ds_remotestr);
 
 	/* No multipath support. Use first DS */
 	data->ds_clp = ds->ds_clp;
@@ -374,6 +376,9 @@ filelayout_write_pagelist(struct nfs_write_data *data, int sync)
 	struct nfs_fh *fh;
 	int status;
 
+	if (test_bit(NFS_DEVICEID_INVALID, &FILELAYOUT_DEVID_NODE(lseg)->flags))
+		return PNFS_NOT_ATTEMPTED;
+
 	/* Retrieve the correct rpc_client for the byte range */
 	j = nfs4_fl_calc_j_index(lseg, offset);
 	idx = nfs4_fl_calc_ds_index(lseg, j);
@@ -384,9 +389,9 @@ filelayout_write_pagelist(struct nfs_write_data *data, int sync)
 		set_bit(lo_fail_bit(IOMODE_READ), &lseg->pls_layout->plh_flags);
 		return PNFS_NOT_ATTEMPTED;
 	}
-	dprintk("%s ino %lu sync %d req %Zu@%llu DS:%x:%hu\n", __func__,
+	dprintk("%s ino %lu sync %d req %Zu@%llu DS: %s\n", __func__,
 		data->inode->i_ino, sync, (size_t) data->args.count, offset,
-		ntohl(ds->ds_ip_addr), ntohs(ds->ds_port));
+		ds->ds_remotestr);
 
 	data->write_done_cb = filelayout_write_done_cb;
 	data->ds_clp = ds->ds_clp;
@@ -428,6 +433,14 @@ filelayout_check_layout(struct pnfs_layout_hdr *lo,
 
 	dprintk("--> %s\n", __func__);
 
+	/* FIXME: remove this check when layout segment support is added */
+	if (lgr->range.offset != 0 ||
+	    lgr->range.length != NFS4_MAX_UINT64) {
+		dprintk("%s Only whole file layouts supported. Use MDS i/o\n",
+			__func__);
+		goto out;
+	}
+
 	if (fl->pattern_offset > lgr->range.offset) {
 		dprintk("%s pattern_offset %lld too large\n",
 				__func__, fl->pattern_offset);
@@ -449,6 +462,10 @@ filelayout_check_layout(struct pnfs_layout_hdr *lo,
 			goto out;
 	} else
 		dsaddr = container_of(d, struct nfs4_file_layout_dsaddr, id_node);
+	/* Found deviceid is being reaped */
+	if (test_bit(NFS_DEVICEID_INVALID, &dsaddr->id_node.flags))
+			goto out_put;
+
 	fl->dsaddr = dsaddr;
 
 	if (fl->first_stripe_index < 0 ||
@@ -659,7 +676,7 @@ filelayout_alloc_lseg(struct pnfs_layout_hdr *layoutid,
  * return true  : coalesce page
  * return false : don't coalesce page
  */
-bool
+static bool
 filelayout_pg_test(struct nfs_pageio_descriptor *pgio, struct nfs_page *prev,
 		   struct nfs_page *req)
 {
@@ -670,8 +687,6 @@ filelayout_pg_test(struct nfs_pageio_descriptor *pgio, struct nfs_page *prev,
 	    !nfs_generic_pg_test(pgio, prev, req))
 		return false;
 
-	if (!pgio->pg_lseg)
-		return 1;
 	p_stripe = (u64)prev->wb_index << PAGE_CACHE_SHIFT;
 	r_stripe = (u64)req->wb_index << PAGE_CACHE_SHIFT;
 	stripe_unit = FILELAYOUT_LSEG(pgio->pg_lseg)->stripe_unit;
@@ -682,6 +697,52 @@ filelayout_pg_test(struct nfs_pageio_descriptor *pgio, struct nfs_page *prev,
 	return (p_stripe == r_stripe);
 }
 
+void
+filelayout_pg_init_read(struct nfs_pageio_descriptor *pgio,
+			struct nfs_page *req)
+{
+	BUG_ON(pgio->pg_lseg != NULL);
+
+	pgio->pg_lseg = pnfs_update_layout(pgio->pg_inode,
+					   req->wb_context,
+					   0,
+					   NFS4_MAX_UINT64,
+					   IOMODE_READ,
+					   GFP_KERNEL);
+	/* If no lseg, fall back to read through mds */
+	if (pgio->pg_lseg == NULL)
+		nfs_pageio_reset_read_mds(pgio);
+}
+
+void
+filelayout_pg_init_write(struct nfs_pageio_descriptor *pgio,
+			 struct nfs_page *req)
+{
+	BUG_ON(pgio->pg_lseg != NULL);
+
+	pgio->pg_lseg = pnfs_update_layout(pgio->pg_inode,
+					   req->wb_context,
+					   0,
+					   NFS4_MAX_UINT64,
+					   IOMODE_RW,
+					   GFP_NOFS);
+	/* If no lseg, fall back to write through mds */
+	if (pgio->pg_lseg == NULL)
+		nfs_pageio_reset_write_mds(pgio);
+}
+
+static const struct nfs_pageio_ops filelayout_pg_read_ops = {
+	.pg_init = filelayout_pg_init_read,
+	.pg_test = filelayout_pg_test,
+	.pg_doio = pnfs_generic_pg_readpages,
+};
+
+static const struct nfs_pageio_ops filelayout_pg_write_ops = {
+	.pg_init = filelayout_pg_init_write,
+	.pg_test = filelayout_pg_test,
+	.pg_doio = pnfs_generic_pg_writepages,
+};
+
 static bool filelayout_mark_pnfs_commit(struct pnfs_layout_segment *lseg)
 {
 	return !FILELAYOUT_LSEG(lseg)->commit_through_mds;
@@ -879,7 +940,8 @@ static struct pnfs_layoutdriver_type filelayout_type = {
 	.owner			= THIS_MODULE,
 	.alloc_lseg		= filelayout_alloc_lseg,
 	.free_lseg		= filelayout_free_lseg,
-	.pg_test		= filelayout_pg_test,
+	.pg_read_ops		= &filelayout_pg_read_ops,
+	.pg_write_ops		= &filelayout_pg_write_ops,
 	.mark_pnfs_commit	= filelayout_mark_pnfs_commit,
 	.choose_commit_list	= filelayout_choose_commit_list,
 	.commit_pagelist	= filelayout_commit_pagelist,
@@ -902,5 +964,7 @@ static void __exit nfs4filelayout_exit(void)
 	pnfs_unregister_layoutdriver(&filelayout_type);
 }
 
+MODULE_ALIAS("nfs-layouttype4-1");
+
 module_init(nfs4filelayout_init);
 module_exit(nfs4filelayout_exit);

fs/nfs/nfs4filelayout.h

@@ -47,10 +47,17 @@ enum stripetype4 {
 };
 
 /* Individual ip address */
+struct nfs4_pnfs_ds_addr {
+	struct sockaddr_storage	da_addr;
+	size_t			da_addrlen;
+	struct list_head	da_node;  /* nfs4_pnfs_dev_hlist dev_dslist */
+	char			*da_remotestr;	/* human readable addr+port */
+};
+
 struct nfs4_pnfs_ds {
 	struct list_head	ds_node;  /* nfs4_pnfs_dev_hlist dev_dslist */
-	u32			ds_ip_addr;
-	u32			ds_port;
+	char			*ds_remotestr;	/* comma sep list of addrs */
+	struct list_head	ds_addrs;
 	struct nfs_client	*ds_clp;
 	atomic_t		ds_count;
 };
@@ -89,6 +96,12 @@ FILELAYOUT_LSEG(struct pnfs_layout_segment *lseg)
 			    generic_hdr);
 }
 
+static inline struct nfs4_deviceid_node *
+FILELAYOUT_DEVID_NODE(struct pnfs_layout_segment *lseg)
+{
+	return &FILELAYOUT_LSEG(lseg)->dsaddr->id_node;
+}
+
 extern struct nfs_fh *
 nfs4_fl_select_ds_fh(struct pnfs_layout_segment *lseg, u32 j);
 

fs/nfs/nfs4filelayoutdev.c

@@ -56,54 +56,139 @@ print_ds(struct nfs4_pnfs_ds *ds)
 		printk("%s NULL device\n", __func__);
 		return;
 	}
-	printk("        ip_addr %x port %hu\n"
+	printk("        ds %s\n"
 		"        ref count %d\n"
 		"        client %p\n"
 		"        cl_exchange_flags %x\n",
-		ntohl(ds->ds_ip_addr), ntohs(ds->ds_port),
+		ds->ds_remotestr,
 		atomic_read(&ds->ds_count), ds->ds_clp,
 		ds->ds_clp ? ds->ds_clp->cl_exchange_flags : 0);
 }
 
-/* nfs4_ds_cache_lock is held */
-static struct nfs4_pnfs_ds *
-_data_server_lookup_locked(u32 ip_addr, u32 port)
+static bool
+same_sockaddr(struct sockaddr *addr1, struct sockaddr *addr2)
 {
-	struct nfs4_pnfs_ds *ds;
+	struct sockaddr_in *a, *b;
+	struct sockaddr_in6 *a6, *b6;
+
+	if (addr1->sa_family != addr2->sa_family)
+		return false;
+
+	switch (addr1->sa_family) {
+	case AF_INET:
+		a = (struct sockaddr_in *)addr1;
+		b = (struct sockaddr_in *)addr2;
+
+		if (a->sin_addr.s_addr == b->sin_addr.s_addr &&
+		    a->sin_port == b->sin_port)
+			return true;
+		break;
+
+	case AF_INET6:
+		a6 = (struct sockaddr_in6 *)addr1;
+		b6 = (struct sockaddr_in6 *)addr2;
+
+		/* LINKLOCAL addresses must have matching scope_id */
+		if (ipv6_addr_scope(&a6->sin6_addr) ==
+		    IPV6_ADDR_SCOPE_LINKLOCAL &&
+		    a6->sin6_scope_id != b6->sin6_scope_id)
+			return false;
+
+		if (ipv6_addr_equal(&a6->sin6_addr, &b6->sin6_addr) &&
+		    a6->sin6_port == b6->sin6_port)
+			return true;
+		break;
+
+	default:
+		dprintk("%s: unhandled address family: %u\n",
+			__func__, addr1->sa_family);
+		return false;
+	}
 
-	dprintk("_data_server_lookup: ip_addr=%x port=%hu\n",
-			ntohl(ip_addr), ntohs(port));
+	return false;
+}
 
-	list_for_each_entry(ds, &nfs4_data_server_cache, ds_node) {
-		if (ds->ds_ip_addr == ip_addr &&
-		    ds->ds_port == port) {
-			return ds;
+/*
+ * Lookup DS by addresses.  The first matching address returns true.
+ * nfs4_ds_cache_lock is held
+ */
+static struct nfs4_pnfs_ds *
+_data_server_lookup_locked(struct list_head *dsaddrs)
+{
+	struct nfs4_pnfs_ds *ds;
+	struct nfs4_pnfs_ds_addr *da1, *da2;
+
+	list_for_each_entry(da1, dsaddrs, da_node) {
+		list_for_each_entry(ds, &nfs4_data_server_cache, ds_node) {
+			list_for_each_entry(da2, &ds->ds_addrs, da_node) {
+				if (same_sockaddr(
+					(struct sockaddr *)&da1->da_addr,
+					(struct sockaddr *)&da2->da_addr))
+					return ds;
+			}
 		}
 	}
 	return NULL;
 }
 
+/*
+ * Compare two lists of addresses.
+ */
+static bool
+_data_server_match_all_addrs_locked(struct list_head *dsaddrs1,
+				    struct list_head *dsaddrs2)
+{
+	struct nfs4_pnfs_ds_addr *da1, *da2;
+	size_t count1 = 0,
+	       count2 = 0;
+
+	list_for_each_entry(da1, dsaddrs1, da_node)
+		count1++;
+
+	list_for_each_entry(da2, dsaddrs2, da_node) {
+		bool found = false;
+		count2++;
+		list_for_each_entry(da1, dsaddrs1, da_node) {
+			if (same_sockaddr((struct sockaddr *)&da1->da_addr,
+				(struct sockaddr *)&da2->da_addr)) {
+				found = true;
+				break;
+			}
+		}
+		if (!found)
+			return false;
+	}
+
+	return (count1 == count2);
+}
+
 /*
  * Create an rpc connection to the nfs4_pnfs_ds data server
- * Currently only support IPv4
+ * Currently only supports IPv4 and IPv6 addresses
  */
 static int
 nfs4_ds_connect(struct nfs_server *mds_srv, struct nfs4_pnfs_ds *ds)
 {
-	struct nfs_client *clp;
-	struct sockaddr_in sin;
+	struct nfs_client *clp = ERR_PTR(-EIO);
+	struct nfs4_pnfs_ds_addr *da;
 	int status = 0;
 
-	dprintk("--> %s ip:port %x:%hu au_flavor %d\n", __func__,
-		ntohl(ds->ds_ip_addr), ntohs(ds->ds_port),
+	dprintk("--> %s DS %s au_flavor %d\n", __func__, ds->ds_remotestr,
 		mds_srv->nfs_client->cl_rpcclient->cl_auth->au_flavor);
 
-	sin.sin_family = AF_INET;
-	sin.sin_addr.s_addr = ds->ds_ip_addr;
-	sin.sin_port = ds->ds_port;
+	BUG_ON(list_empty(&ds->ds_addrs));
+
+	list_for_each_entry(da, &ds->ds_addrs, da_node) {
+		dprintk("%s: DS %s: trying address %s\n",
+			__func__, ds->ds_remotestr, da->da_remotestr);
+
+		clp = nfs4_set_ds_client(mds_srv->nfs_client,
+				 (struct sockaddr *)&da->da_addr,
+				 da->da_addrlen, IPPROTO_TCP);
+		if (!IS_ERR(clp))
+			break;
+	}
 
-	clp = nfs4_set_ds_client(mds_srv->nfs_client, (struct sockaddr *)&sin,
-				 sizeof(sin), IPPROTO_TCP);
 	if (IS_ERR(clp)) {
 		status = PTR_ERR(clp);
 		goto out;
@@ -115,8 +200,8 @@ nfs4_ds_connect(struct nfs_server *mds_srv, struct nfs4_pnfs_ds *ds)
 			goto out_put;
 		}
 		ds->ds_clp = clp;
-		dprintk("%s [existing] ip=%x, port=%hu\n", __func__,
-			ntohl(ds->ds_ip_addr), ntohs(ds->ds_port));
+		dprintk("%s [existing] server=%s\n", __func__,
+			ds->ds_remotestr);
 		goto out;
 	}
 
@@ -135,8 +220,7 @@ nfs4_ds_connect(struct nfs_server *mds_srv, struct nfs4_pnfs_ds *ds)
 		goto out_put;
 
 	ds->ds_clp = clp;
-	dprintk("%s [new] ip=%x, port=%hu\n", __func__, ntohl(ds->ds_ip_addr),
-		ntohs(ds->ds_port));
+	dprintk("%s [new] addr: %s\n", __func__, ds->ds_remotestr);
 out:
 	return status;
 out_put:
@@ -147,12 +231,25 @@ out_put:
 static void
 destroy_ds(struct nfs4_pnfs_ds *ds)
 {
+	struct nfs4_pnfs_ds_addr *da;
+
 	dprintk("--> %s\n", __func__);
 	ifdebug(FACILITY)
 		print_ds(ds);
 
 	if (ds->ds_clp)
 		nfs_put_client(ds->ds_clp);
+
+	while (!list_empty(&ds->ds_addrs)) {
+		da = list_first_entry(&ds->ds_addrs,
+				      struct nfs4_pnfs_ds_addr,
+				      da_node);
+		list_del_init(&da->da_node);
+		kfree(da->da_remotestr);
+		kfree(da);
+	}
+
+	kfree(ds->ds_remotestr);
 	kfree(ds);
 }
 
@@ -179,31 +276,96 @@ nfs4_fl_free_deviceid(struct nfs4_file_layout_dsaddr *dsaddr)
 	kfree(dsaddr);
 }
 
+/*
+ * Create a string with a human readable address and port to avoid
+ * complicated setup around many dprinks.
+ */
+static char *
+nfs4_pnfs_remotestr(struct list_head *dsaddrs, gfp_t gfp_flags)
+{
+	struct nfs4_pnfs_ds_addr *da;
+	char *remotestr;
+	size_t len;
+	char *p;
+
+	len = 3;        /* '{', '}' and eol */
+	list_for_each_entry(da, dsaddrs, da_node) {
+		len += strlen(da->da_remotestr) + 1;    /* string plus comma */
+	}
+
+	remotestr = kzalloc(len, gfp_flags);
+	if (!remotestr)
+		return NULL;
+
+	p = remotestr;
+	*(p++) = '{';
+	len--;
+	list_for_each_entry(da, dsaddrs, da_node) {
+		size_t ll = strlen(da->da_remotestr);
+
+		if (ll > len)
+			goto out_err;
+
+		memcpy(p, da->da_remotestr, ll);
+		p += ll;
+		len -= ll;
+
+		if (len < 1)
+			goto out_err;
+		(*p++) = ',';
+		len--;
+	}
+	if (len < 2)
+		goto out_err;
+	*(p++) = '}';
+	*p = '\0';
+	return remotestr;
+out_err:
+	kfree(remotestr);
+	return NULL;
+}
+
 static struct nfs4_pnfs_ds *
-nfs4_pnfs_ds_add(struct inode *inode, u32 ip_addr, u32 port, gfp_t gfp_flags)
+nfs4_pnfs_ds_add(struct list_head *dsaddrs, gfp_t gfp_flags)
 {
-	struct nfs4_pnfs_ds *tmp_ds, *ds;
+	struct nfs4_pnfs_ds *tmp_ds, *ds = NULL;
+	char *remotestr;
 
-	ds = kzalloc(sizeof(*tmp_ds), gfp_flags);
+	if (list_empty(dsaddrs)) {
+		dprintk("%s: no addresses defined\n", __func__);
+		goto out;
+	}
+
+	ds = kzalloc(sizeof(*ds), gfp_flags);
 	if (!ds)
 		goto out;
 
+	/* this is only used for debugging, so it's ok if its NULL */
+	remotestr = nfs4_pnfs_remotestr(dsaddrs, gfp_flags);
+
 	spin_lock(&nfs4_ds_cache_lock);
-	tmp_ds = _data_server_lookup_locked(ip_addr, port);
+	tmp_ds = _data_server_lookup_locked(dsaddrs);
 	if (tmp_ds == NULL) {
-		ds->ds_ip_addr = ip_addr;
-		ds->ds_port = port;
+		INIT_LIST_HEAD(&ds->ds_addrs);
+		list_splice_init(dsaddrs, &ds->ds_addrs);
+		ds->ds_remotestr = remotestr;
 		atomic_set(&ds->ds_count, 1);
 		INIT_LIST_HEAD(&ds->ds_node);
 		ds->ds_clp = NULL;
 		list_add(&ds->ds_node, &nfs4_data_server_cache);
-		dprintk("%s add new data server ip 0x%x\n", __func__,
-			ds->ds_ip_addr);
+		dprintk("%s add new data server %s\n", __func__,
+			ds->ds_remotestr);
 	} else {
+		if (!_data_server_match_all_addrs_locked(&tmp_ds->ds_addrs,
+							 dsaddrs)) {
+			dprintk("%s:  multipath address mismatch: %s != %s",
+				__func__, tmp_ds->ds_remotestr, remotestr);
+		}
+		kfree(remotestr);
 		kfree(ds);
 		atomic_inc(&tmp_ds->ds_count);
-		dprintk("%s data server found ip 0x%x, inc'ed ds_count to %d\n",
-			__func__, tmp_ds->ds_ip_addr,
+		dprintk("%s data server %s found, inc'ed ds_count to %d\n",
+			__func__, tmp_ds->ds_remotestr,
 			atomic_read(&tmp_ds->ds_count));
 		ds = tmp_ds;
 	}
@@ -213,18 +375,22 @@ out:
 }
 
 /*
- * Currently only support ipv4, and one multi-path address.
+ * Currently only supports ipv4, ipv6 and one multi-path address.
  */
-static struct nfs4_pnfs_ds *
-decode_and_add_ds(struct xdr_stream *streamp, struct inode *inode, gfp_t gfp_flags)
+static struct nfs4_pnfs_ds_addr *
+decode_ds_addr(struct xdr_stream *streamp, gfp_t gfp_flags)
 {
-	struct nfs4_pnfs_ds *ds = NULL;
-	char *buf;
-	const char *ipend, *pstr;
-	u32 ip_addr, port;
-	int nlen, rlen, i;
+	struct nfs4_pnfs_ds_addr *da = NULL;
+	char *buf, *portstr;
+	u32 port;
+	int nlen, rlen;
 	int tmp[2];
 	__be32 *p;
+	char *netid, *match_netid;
+	size_t len, match_netid_len;
+	char *startsep = "";
+	char *endsep = "";
+
 
 	/* r_netid */
 	p = xdr_inline_decode(streamp, 4);
@@ -236,64 +402,123 @@ decode_and_add_ds(struct xdr_stream *streamp, struct inode *inode, gfp_t gfp_fla
 	if (unlikely(!p))
 		goto out_err;
 
-	/* Check that netid is "tcp" */
-	if (nlen != 3 ||  memcmp((char *)p, "tcp", 3)) {
-		dprintk("%s: ERROR: non ipv4 TCP r_netid\n", __func__);
+	netid = kmalloc(nlen+1, gfp_flags);
+	if (unlikely(!netid))
 		goto out_err;
-	}
 
-	/* r_addr */
+	netid[nlen] = '\0';
+	memcpy(netid, p, nlen);
+
+	/* r_addr: ip/ip6addr with port in dec octets - see RFC 5665 */
 	p = xdr_inline_decode(streamp, 4);
 	if (unlikely(!p))
-		goto out_err;
+		goto out_free_netid;
 	rlen = be32_to_cpup(p);
 
 	p = xdr_inline_decode(streamp, rlen);
 	if (unlikely(!p))
-		goto out_err;
+		goto out_free_netid;
 
-	/* ipv6 length plus port is legal */
-	if (rlen > INET6_ADDRSTRLEN + 8) {
+	/* port is ".ABC.DEF", 8 chars max */
+	if (rlen > INET6_ADDRSTRLEN + IPV6_SCOPE_ID_LEN + 8) {
 		dprintk("%s: Invalid address, length %d\n", __func__,
 			rlen);
-		goto out_err;
+		goto out_free_netid;
 	}
 	buf = kmalloc(rlen + 1, gfp_flags);
 	if (!buf) {
 		dprintk("%s: Not enough memory\n", __func__);
-		goto out_err;
+		goto out_free_netid;
 	}
 	buf[rlen] = '\0';
 	memcpy(buf, p, rlen);
 
-	/* replace the port dots with dashes for the in4_pton() delimiter*/
-	for (i = 0; i < 2; i++) {
-		char *res = strrchr(buf, '.');
-		if (!res) {
-			dprintk("%s: Failed finding expected dots in port\n",
-				__func__);
-			goto out_free;
-		}
-		*res = '-';
+	/* replace port '.' with '-' */
+	portstr = strrchr(buf, '.');
+	if (!portstr) {
+		dprintk("%s: Failed finding expected dot in port\n",
+			__func__);
+		goto out_free_buf;
+	}
+	*portstr = '-';
+
+	/* find '.' between address and port */
+	portstr = strrchr(buf, '.');
+	if (!portstr) {
+		dprintk("%s: Failed finding expected dot between address and "
+			"port\n", __func__);
+		goto out_free_buf;
 	}
+	*portstr = '\0';
 
-	/* Currently only support ipv4 address */
-	if (in4_pton(buf, rlen, (u8 *)&ip_addr, '-', &ipend) == 0) {
-		dprintk("%s: Only ipv4 addresses supported\n", __func__);
-		goto out_free;
+	da = kzalloc(sizeof(*da), gfp_flags);
+	if (unlikely(!da))
+		goto out_free_buf;
+
+	INIT_LIST_HEAD(&da->da_node);
+
+	if (!rpc_pton(buf, portstr-buf, (struct sockaddr *)&da->da_addr,
+		      sizeof(da->da_addr))) {
+		dprintk("%s: error parsing address %s\n", __func__, buf);
+		goto out_free_da;
 	}
 
-	/* port */
-	pstr = ipend;
-	sscanf(pstr, "-%d-%d", &tmp[0], &tmp[1]);
+	portstr++;
+	sscanf(portstr, "%d-%d", &tmp[0], &tmp[1]);
 	port = htons((tmp[0] << 8) | (tmp[1]));
 
-	ds = nfs4_pnfs_ds_add(inode, ip_addr, port, gfp_flags);
-	dprintk("%s: Decoded address and port %s\n", __func__, buf);
-out_free:
+	switch (da->da_addr.ss_family) {
+	case AF_INET:
+		((struct sockaddr_in *)&da->da_addr)->sin_port = port;
+		da->da_addrlen = sizeof(struct sockaddr_in);
+		match_netid = "tcp";
+		match_netid_len = 3;
+		break;
+
+	case AF_INET6:
+		((struct sockaddr_in6 *)&da->da_addr)->sin6_port = port;
+		da->da_addrlen = sizeof(struct sockaddr_in6);
+		match_netid = "tcp6";
+		match_netid_len = 4;
+		startsep = "[";
+		endsep = "]";
+		break;
+
+	default:
+		dprintk("%s: unsupported address family: %u\n",
+			__func__, da->da_addr.ss_family);
+		goto out_free_da;
+	}
+
+	if (nlen != match_netid_len || strncmp(netid, match_netid, nlen)) {
+		dprintk("%s: ERROR: r_netid \"%s\" != \"%s\"\n",
+			__func__, netid, match_netid);
+		goto out_free_da;
+	}
+
+	/* save human readable address */
+	len = strlen(startsep) + strlen(buf) + strlen(endsep) + 7;
+	da->da_remotestr = kzalloc(len, gfp_flags);
+
+	/* NULL is ok, only used for dprintk */
+	if (da->da_remotestr)
+		snprintf(da->da_remotestr, len, "%s%s%s:%u", startsep,
+			 buf, endsep, ntohs(port));
+
+	dprintk("%s: Parsed DS addr %s\n", __func__, da->da_remotestr);
 	kfree(buf);
+	kfree(netid);
+	return da;
+
+out_free_da:
+	kfree(da);
+out_free_buf:
+	dprintk("%s: Error parsing DS addr: %s\n", __func__, buf);
+	kfree(buf);
+out_free_netid:
+	kfree(netid);
 out_err:
-	return ds;
+	return NULL;
 }
 
 /* Decode opaque device data and return the result */
@@ -310,6 +535,8 @@ decode_device(struct inode *ino, struct pnfs_device *pdev, gfp_t gfp_flags)
 	struct xdr_stream stream;
 	struct xdr_buf buf;
 	struct page *scratch;
+	struct list_head dsaddrs;
+	struct nfs4_pnfs_ds_addr *da;
 
 	/* set up xdr stream */
 	scratch = alloc_page(gfp_flags);
@@ -386,6 +613,8 @@ decode_device(struct inode *ino, struct pnfs_device *pdev, gfp_t gfp_flags)
 				NFS_SERVER(ino)->nfs_client,
 				&pdev->dev_id);
 
+	INIT_LIST_HEAD(&dsaddrs);
+
 	for (i = 0; i < dsaddr->ds_num; i++) {
 		int j;
 		u32 mp_count;
@@ -395,48 +624,43 @@ decode_device(struct inode *ino, struct pnfs_device *pdev, gfp_t gfp_flags)
 			goto out_err_free_deviceid;
 
 		mp_count = be32_to_cpup(p); /* multipath count */
-		if (mp_count > 1) {
-			printk(KERN_WARNING
-			       "%s: Multipath count %d not supported, "
-			       "skipping all greater than 1\n", __func__,
-				mp_count);
-		}
 		for (j = 0; j < mp_count; j++) {
-			if (j == 0) {
-				dsaddr->ds_list[i] = decode_and_add_ds(&stream,
-					ino, gfp_flags);
-				if (dsaddr->ds_list[i] == NULL)
-					goto out_err_free_deviceid;
-			} else {
-				u32 len;
-				/* skip extra multipath */
-
-				/* read len, skip */
-				p = xdr_inline_decode(&stream, 4);
-				if (unlikely(!p))
-					goto out_err_free_deviceid;
-				len = be32_to_cpup(p);
-
-				p = xdr_inline_decode(&stream, len);
-				if (unlikely(!p))
-					goto out_err_free_deviceid;
-
-				/* read len, skip */
-				p = xdr_inline_decode(&stream, 4);
-				if (unlikely(!p))
-					goto out_err_free_deviceid;
-				len = be32_to_cpup(p);
-
-				p = xdr_inline_decode(&stream, len);
-				if (unlikely(!p))
-					goto out_err_free_deviceid;
-			}
+			da = decode_ds_addr(&stream, gfp_flags);
+			if (da)
+				list_add_tail(&da->da_node, &dsaddrs);
+		}
+		if (list_empty(&dsaddrs)) {
+			dprintk("%s: no suitable DS addresses found\n",
+				__func__);
+			goto out_err_free_deviceid;
+		}
+
+		dsaddr->ds_list[i] = nfs4_pnfs_ds_add(&dsaddrs, gfp_flags);
+		if (!dsaddr->ds_list[i])
+			goto out_err_drain_dsaddrs;
+
+		/* If DS was already in cache, free ds addrs */
+		while (!list_empty(&dsaddrs)) {
+			da = list_first_entry(&dsaddrs,
+					      struct nfs4_pnfs_ds_addr,
+					      da_node);
+			list_del_init(&da->da_node);
+			kfree(da->da_remotestr);
+			kfree(da);
 		}
 	}
 
 	__free_page(scratch);
 	return dsaddr;
 
+out_err_drain_dsaddrs:
+	while (!list_empty(&dsaddrs)) {
+		da = list_first_entry(&dsaddrs, struct nfs4_pnfs_ds_addr,
+				      da_node);
+		list_del_init(&da->da_node);
+		kfree(da->da_remotestr);
+		kfree(da);
+	}
 out_err_free_deviceid:
 	nfs4_fl_free_deviceid(dsaddr);
 	/* stripe_indicies was part of dsaddr */
@@ -591,13 +815,13 @@ nfs4_fl_select_ds_fh(struct pnfs_layout_segment *lseg, u32 j)
 
 static void
 filelayout_mark_devid_negative(struct nfs4_file_layout_dsaddr *dsaddr,
-			       int err, u32 ds_addr)
+			       int err, const char *ds_remotestr)
 {
 	u32 *p = (u32 *)&dsaddr->id_node.deviceid;
 
-	printk(KERN_ERR "NFS: data server %x connection error %d."
+	printk(KERN_ERR "NFS: data server %s connection error %d."
 		" Deviceid [%x%x%x%x] marked out of use.\n",
-		ds_addr, err, p[0], p[1], p[2], p[3]);
+		ds_remotestr, err, p[0], p[1], p[2], p[3]);
 
 	spin_lock(&nfs4_ds_cache_lock);
 	dsaddr->flags |= NFS4_DEVICE_ID_NEG_ENTRY;
@@ -628,7 +852,7 @@ nfs4_fl_prepare_ds(struct pnfs_layout_segment *lseg, u32 ds_idx)
 		err = nfs4_ds_connect(s, ds);
 		if (err) {
 			filelayout_mark_devid_negative(dsaddr, err,
-						       ntohl(ds->ds_ip_addr));
+						       ds->ds_remotestr);
 			return NULL;
 		}
 	}

fs/nfs/nfs4proc.c

@@ -80,7 +80,10 @@ static int _nfs4_proc_getattr(struct nfs_server *server, struct nfs_fh *fhandle,
 static int nfs4_do_setattr(struct inode *inode, struct rpc_cred *cred,
 			    struct nfs_fattr *fattr, struct iattr *sattr,
 			    struct nfs4_state *state);
-
+#ifdef CONFIG_NFS_V4_1
+static int nfs41_test_stateid(struct nfs_server *, struct nfs4_state *);
+static int nfs41_free_stateid(struct nfs_server *, struct nfs4_state *);
+#endif
 /* Prevent leaks of NFSv4 errors into userland */
 static int nfs4_map_errors(int err)
 {
@@ -1689,6 +1692,20 @@ static int nfs4_open_expired(struct nfs4_state_owner *sp, struct nfs4_state *sta
 	return ret;
 }
 
+#if defined(CONFIG_NFS_V4_1)
+static int nfs41_open_expired(struct nfs4_state_owner *sp, struct nfs4_state *state)
+{
+	int status;
+	struct nfs_server *server = NFS_SERVER(state->inode);
+
+	status = nfs41_test_stateid(server, state);
+	if (status == NFS_OK)
+		return 0;
+	nfs41_free_stateid(server, state);
+	return nfs4_open_expired(sp, state);
+}
+#endif
+
 /*
  * on an EXCLUSIVE create, the server should send back a bitmask with FATTR4-*
  * fields corresponding to attributes that were used to store the verifier.
@@ -2252,13 +2269,14 @@ static int nfs4_find_root_sec(struct nfs_server *server, struct nfs_fh *fhandle,
 static int nfs4_proc_get_root(struct nfs_server *server, struct nfs_fh *fhandle,
 			      struct nfs_fsinfo *info)
 {
+	int minor_version = server->nfs_client->cl_minorversion;
 	int status = nfs4_lookup_root(server, fhandle, info);
 	if ((status == -NFS4ERR_WRONGSEC) && !(server->flags & NFS_MOUNT_SECFLAVOUR))
 		/*
 		 * A status of -NFS4ERR_WRONGSEC will be mapped to -EPERM
 		 * by nfs4_map_errors() as this function exits.
 		 */
-		status = nfs4_find_root_sec(server, fhandle, info);
+		status = nfs_v4_minor_ops[minor_version]->find_root_sec(server, fhandle, info);
 	if (status == 0)
 		status = nfs4_server_capabilities(server, fhandle);
 	if (status == 0)
@@ -4441,6 +4459,20 @@ out:
 	return err;
 }
 
+#if defined(CONFIG_NFS_V4_1)
+static int nfs41_lock_expired(struct nfs4_state *state, struct file_lock *request)
+{
+	int status;
+	struct nfs_server *server = NFS_SERVER(state->inode);
+
+	status = nfs41_test_stateid(server, state);
+	if (status == NFS_OK)
+		return 0;
+	nfs41_free_stateid(server, state);
+	return nfs4_lock_expired(state, request);
+}
+#endif
+
 static int _nfs4_proc_setlk(struct nfs4_state *state, int cmd, struct file_lock *request)
 {
 	struct nfs_inode *nfsi = NFS_I(state->inode);
@@ -4779,6 +4811,16 @@ out_inval:
 	return -NFS4ERR_INVAL;
 }
 
+static bool
+nfs41_same_server_scope(struct server_scope *a, struct server_scope *b)
+{
+	if (a->server_scope_sz == b->server_scope_sz &&
+	    memcmp(a->server_scope, b->server_scope, a->server_scope_sz) == 0)
+		return true;
+
+	return false;
+}
+
 /*
  * nfs4_proc_exchange_id()
  *
@@ -4821,9 +4863,31 @@ int nfs4_proc_exchange_id(struct nfs_client *clp, struct rpc_cred *cred)
 				init_utsname()->domainname,
 				clp->cl_rpcclient->cl_auth->au_flavor);
 
+	res.server_scope = kzalloc(sizeof(struct server_scope), GFP_KERNEL);
+	if (unlikely(!res.server_scope))
+		return -ENOMEM;
+
 	status = rpc_call_sync(clp->cl_rpcclient, &msg, RPC_TASK_TIMEOUT);
 	if (!status)
 		status = nfs4_check_cl_exchange_flags(clp->cl_exchange_flags);
+
+	if (!status) {
+		if (clp->server_scope &&
+		    !nfs41_same_server_scope(clp->server_scope,
+					     res.server_scope)) {
+			dprintk("%s: server_scope mismatch detected\n",
+				__func__);
+			set_bit(NFS4CLNT_SERVER_SCOPE_MISMATCH, &clp->cl_state);
+			kfree(clp->server_scope);
+			clp->server_scope = NULL;
+		}
+
+		if (!clp->server_scope)
+			clp->server_scope = res.server_scope;
+		else
+			kfree(res.server_scope);
+	}
+
 	dprintk("<-- %s status= %d\n", __func__, status);
 	return status;
 }
@@ -5704,7 +5768,7 @@ static void nfs4_layoutreturn_done(struct rpc_task *task, void *calldata)
 {
 	struct nfs4_layoutreturn *lrp = calldata;
 	struct nfs_server *server;
-	struct pnfs_layout_hdr *lo = NFS_I(lrp->args.inode)->layout;
+	struct pnfs_layout_hdr *lo = lrp->args.layout;
 
 	dprintk("--> %s\n", __func__);
 
@@ -5733,7 +5797,7 @@ static void nfs4_layoutreturn_release(void *calldata)
 	struct nfs4_layoutreturn *lrp = calldata;
 
 	dprintk("--> %s\n", __func__);
-	put_layout_hdr(NFS_I(lrp->args.inode)->layout);
+	put_layout_hdr(lrp->args.layout);
 	kfree(calldata);
 	dprintk("<-- %s\n", __func__);
 }
@@ -5901,6 +5965,143 @@ out:
 	rpc_put_task(task);
 	return status;
 }
+
+static int
+_nfs41_proc_secinfo_no_name(struct nfs_server *server, struct nfs_fh *fhandle,
+		    struct nfs_fsinfo *info, struct nfs4_secinfo_flavors *flavors)
+{
+	struct nfs41_secinfo_no_name_args args = {
+		.style = SECINFO_STYLE_CURRENT_FH,
+	};
+	struct nfs4_secinfo_res res = {
+		.flavors = flavors,
+	};
+	struct rpc_message msg = {
+		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_SECINFO_NO_NAME],
+		.rpc_argp = &args,
+		.rpc_resp = &res,
+	};
+	return nfs4_call_sync(server->client, server, &msg, &args.seq_args, &res.seq_res, 0);
+}
+
+static int
+nfs41_proc_secinfo_no_name(struct nfs_server *server, struct nfs_fh *fhandle,
+			   struct nfs_fsinfo *info, struct nfs4_secinfo_flavors *flavors)
+{
+	struct nfs4_exception exception = { };
+	int err;
+	do {
+		err = _nfs41_proc_secinfo_no_name(server, fhandle, info, flavors);
+		switch (err) {
+		case 0:
+		case -NFS4ERR_WRONGSEC:
+		case -NFS4ERR_NOTSUPP:
+			break;
+		default:
+			err = nfs4_handle_exception(server, err, &exception);
+		}
+	} while (exception.retry);
+	return err;
+}
+
+static int
+nfs41_find_root_sec(struct nfs_server *server, struct nfs_fh *fhandle,
+		    struct nfs_fsinfo *info)
+{
+	int err;
+	struct page *page;
+	rpc_authflavor_t flavor;
+	struct nfs4_secinfo_flavors *flavors;
+
+	page = alloc_page(GFP_KERNEL);
+	if (!page) {
+		err = -ENOMEM;
+		goto out;
+	}
+
+	flavors = page_address(page);
+	err = nfs41_proc_secinfo_no_name(server, fhandle, info, flavors);
+
+	/*
+	 * Fall back on "guess and check" method if
+	 * the server doesn't support SECINFO_NO_NAME
+	 */
+	if (err == -NFS4ERR_WRONGSEC || err == -NFS4ERR_NOTSUPP) {
+		err = nfs4_find_root_sec(server, fhandle, info);
+		goto out_freepage;
+	}
+	if (err)
+		goto out_freepage;
+
+	flavor = nfs_find_best_sec(flavors);
+	if (err == 0)
+		err = nfs4_lookup_root_sec(server, fhandle, info, flavor);
+
+out_freepage:
+	put_page(page);
+	if (err == -EACCES)
+		return -EPERM;
+out:
+	return err;
+}
+static int _nfs41_test_stateid(struct nfs_server *server, struct nfs4_state *state)
+{
+	int status;
+	struct nfs41_test_stateid_args args = {
+		.stateid = &state->stateid,
+	};
+	struct nfs41_test_stateid_res res;
+	struct rpc_message msg = {
+		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_TEST_STATEID],
+		.rpc_argp = &args,
+		.rpc_resp = &res,
+	};
+	args.seq_args.sa_session = res.seq_res.sr_session = NULL;
+	status = nfs4_call_sync_sequence(server->client, server, &msg, &args.seq_args, &res.seq_res, 0, 1);
+	return status;
+}
+
+static int nfs41_test_stateid(struct nfs_server *server, struct nfs4_state *state)
+{
+	struct nfs4_exception exception = { };
+	int err;
+	do {
+		err = nfs4_handle_exception(server,
+				_nfs41_test_stateid(server, state),
+				&exception);
+	} while (exception.retry);
+	return err;
+}
+
+static int _nfs4_free_stateid(struct nfs_server *server, struct nfs4_state *state)
+{
+	int status;
+	struct nfs41_free_stateid_args args = {
+		.stateid = &state->stateid,
+	};
+	struct nfs41_free_stateid_res res;
+	struct rpc_message msg = {
+		.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_FREE_STATEID],
+		.rpc_argp = &args,
+		.rpc_resp = &res,
+	};
+
+	args.seq_args.sa_session = res.seq_res.sr_session = NULL;
+	status = nfs4_call_sync_sequence(server->client, server, &msg, &args.seq_args, &res.seq_res, 0, 1);
+	return status;
+}
+
+static int nfs41_free_stateid(struct nfs_server *server, struct nfs4_state *state)
+{
+	struct nfs4_exception exception = { };
+	int err;
+	do {
+		err = nfs4_handle_exception(server,
+				_nfs4_free_stateid(server, state),
+				&exception);
+	} while (exception.retry);
+	return err;
+}
 #endif /* CONFIG_NFS_V4_1 */
 
 struct nfs4_state_recovery_ops nfs40_reboot_recovery_ops = {
@@ -5937,8 +6138,8 @@ struct nfs4_state_recovery_ops nfs40_nograce_recovery_ops = {
 struct nfs4_state_recovery_ops nfs41_nograce_recovery_ops = {
 	.owner_flag_bit = NFS_OWNER_RECLAIM_NOGRACE,
 	.state_flag_bit	= NFS_STATE_RECLAIM_NOGRACE,
-	.recover_open	= nfs4_open_expired,
-	.recover_lock	= nfs4_lock_expired,
+	.recover_open	= nfs41_open_expired,
+	.recover_lock	= nfs41_lock_expired,
 	.establish_clid = nfs41_init_clientid,
 	.get_clid_cred	= nfs4_get_exchange_id_cred,
 };
@@ -5962,6 +6163,7 @@ static const struct nfs4_minor_version_ops nfs_v4_0_minor_ops = {
 	.minor_version = 0,
 	.call_sync = _nfs4_call_sync,
 	.validate_stateid = nfs4_validate_delegation_stateid,
+	.find_root_sec = nfs4_find_root_sec,
 	.reboot_recovery_ops = &nfs40_reboot_recovery_ops,
 	.nograce_recovery_ops = &nfs40_nograce_recovery_ops,
 	.state_renewal_ops = &nfs40_state_renewal_ops,
@@ -5972,6 +6174,7 @@ static const struct nfs4_minor_version_ops nfs_v4_1_minor_ops = {
 	.minor_version = 1,
 	.call_sync = _nfs4_call_sync_session,
 	.validate_stateid = nfs41_validate_delegation_stateid,
+	.find_root_sec = nfs41_find_root_sec,
 	.reboot_recovery_ops = &nfs41_reboot_recovery_ops,
 	.nograce_recovery_ops = &nfs41_nograce_recovery_ops,
 	.state_renewal_ops = &nfs41_state_renewal_ops,

fs/nfs/nfs4state.c

@@ -1643,7 +1643,14 @@ static void nfs4_state_manager(struct nfs_client *clp)
 				goto out_error;
 			}
 			clear_bit(NFS4CLNT_CHECK_LEASE, &clp->cl_state);
-			set_bit(NFS4CLNT_RECLAIM_REBOOT, &clp->cl_state);
+
+			if (test_and_clear_bit(NFS4CLNT_SERVER_SCOPE_MISMATCH,
+					       &clp->cl_state))
+				nfs4_state_start_reclaim_nograce(clp);
+			else
+				set_bit(NFS4CLNT_RECLAIM_REBOOT,
+					&clp->cl_state);
+
 			pnfs_destroy_all_layouts(clp);
 		}
 

fs/nfs/nfs4xdr.c

@@ -343,6 +343,14 @@ static int nfs4_stat_to_errno(int);
 				1 /* FIXME: opaque lrf_body always empty at the moment */)
 #define decode_layoutreturn_maxsz (op_decode_hdr_maxsz + \
 				1 + decode_stateid_maxsz)
+#define encode_secinfo_no_name_maxsz (op_encode_hdr_maxsz + 1)
+#define decode_secinfo_no_name_maxsz decode_secinfo_maxsz
+#define encode_test_stateid_maxsz	(op_encode_hdr_maxsz + 2 + \
+					 XDR_QUADLEN(NFS4_STATEID_SIZE))
+#define decode_test_stateid_maxsz	(op_decode_hdr_maxsz + 2 + 1)
+#define encode_free_stateid_maxsz	(op_encode_hdr_maxsz + 1 + \
+					 XDR_QUADLEN(NFS4_STATEID_SIZE))
+#define decode_free_stateid_maxsz	(op_decode_hdr_maxsz + 1)
 #else /* CONFIG_NFS_V4_1 */
 #define encode_sequence_maxsz	0
 #define decode_sequence_maxsz	0
@@ -772,6 +780,26 @@ static int nfs4_stat_to_errno(int);
 				decode_sequence_maxsz + \
 				decode_putfh_maxsz + \
 				decode_layoutreturn_maxsz)
+#define NFS4_enc_secinfo_no_name_sz	(compound_encode_hdr_maxsz + \
+					encode_sequence_maxsz + \
+					encode_putrootfh_maxsz +\
+					encode_secinfo_no_name_maxsz)
+#define NFS4_dec_secinfo_no_name_sz	(compound_decode_hdr_maxsz + \
+					decode_sequence_maxsz + \
+					decode_putrootfh_maxsz + \
+					decode_secinfo_no_name_maxsz)
+#define NFS4_enc_test_stateid_sz	(compound_encode_hdr_maxsz + \
+					 encode_sequence_maxsz + \
+					 encode_test_stateid_maxsz)
+#define NFS4_dec_test_stateid_sz	(compound_decode_hdr_maxsz + \
+					 decode_sequence_maxsz + \
+					 decode_test_stateid_maxsz)
+#define NFS4_enc_free_stateid_sz	(compound_encode_hdr_maxsz + \
+					 encode_sequence_maxsz + \
+					 encode_free_stateid_maxsz)
+#define NFS4_dec_free_stateid_sz	(compound_decode_hdr_maxsz + \
+					 decode_sequence_maxsz + \
+					 decode_free_stateid_maxsz)
 
 const u32 nfs41_maxwrite_overhead = ((RPC_MAX_HEADER_WITH_AUTH +
 				      compound_encode_hdr_maxsz +
@@ -1938,6 +1966,46 @@ encode_layoutreturn(struct xdr_stream *xdr,
 	hdr->nops++;
 	hdr->replen += decode_layoutreturn_maxsz;
 }
+
+static int
+encode_secinfo_no_name(struct xdr_stream *xdr,
+		       const struct nfs41_secinfo_no_name_args *args,
+		       struct compound_hdr *hdr)
+{
+	__be32 *p;
+	p = reserve_space(xdr, 8);
+	*p++ = cpu_to_be32(OP_SECINFO_NO_NAME);
+	*p++ = cpu_to_be32(args->style);
+	hdr->nops++;
+	hdr->replen += decode_secinfo_no_name_maxsz;
+	return 0;
+}
+
+static void encode_test_stateid(struct xdr_stream *xdr,
+				struct nfs41_test_stateid_args *args,
+				struct compound_hdr *hdr)
+{
+	__be32 *p;
+
+	p = reserve_space(xdr, 8 + NFS4_STATEID_SIZE);
+	*p++ = cpu_to_be32(OP_TEST_STATEID);
+	*p++ = cpu_to_be32(1);
+	xdr_encode_opaque_fixed(p, args->stateid->data, NFS4_STATEID_SIZE);
+	hdr->nops++;
+	hdr->replen += decode_test_stateid_maxsz;
+}
+
+static void encode_free_stateid(struct xdr_stream *xdr,
+				struct nfs41_free_stateid_args *args,
+				struct compound_hdr *hdr)
+{
+	__be32 *p;
+	p = reserve_space(xdr, 4 + NFS4_STATEID_SIZE);
+	*p++ = cpu_to_be32(OP_FREE_STATEID);
+	xdr_encode_opaque_fixed(p, args->stateid->data, NFS4_STATEID_SIZE);
+	hdr->nops++;
+	hdr->replen += decode_free_stateid_maxsz;
+}
 #endif /* CONFIG_NFS_V4_1 */
 
 /*
@@ -2790,6 +2858,59 @@ static void nfs4_xdr_enc_layoutreturn(struct rpc_rqst *req,
 	encode_layoutreturn(xdr, args, &hdr);
 	encode_nops(&hdr);
 }
+
+/*
+ * Encode SECINFO_NO_NAME request
+ */
+static int nfs4_xdr_enc_secinfo_no_name(struct rpc_rqst *req,
+					struct xdr_stream *xdr,
+					struct nfs41_secinfo_no_name_args *args)
+{
+	struct compound_hdr hdr = {
+		.minorversion = nfs4_xdr_minorversion(&args->seq_args),
+	};
+
+	encode_compound_hdr(xdr, req, &hdr);
+	encode_sequence(xdr, &args->seq_args, &hdr);
+	encode_putrootfh(xdr, &hdr);
+	encode_secinfo_no_name(xdr, args, &hdr);
+	encode_nops(&hdr);
+	return 0;
+}
+
+/*
+ *  Encode TEST_STATEID request
+ */
+static void nfs4_xdr_enc_test_stateid(struct rpc_rqst *req,
+				      struct xdr_stream *xdr,
+				      struct nfs41_test_stateid_args *args)
+{
+	struct compound_hdr hdr = {
+		.minorversion = nfs4_xdr_minorversion(&args->seq_args),
+	};
+
+	encode_compound_hdr(xdr, req, &hdr);
+	encode_sequence(xdr, &args->seq_args, &hdr);
+	encode_test_stateid(xdr, args, &hdr);
+	encode_nops(&hdr);
+}
+
+/*
+ *  Encode FREE_STATEID request
+ */
+static void nfs4_xdr_enc_free_stateid(struct rpc_rqst *req,
+				     struct xdr_stream *xdr,
+				     struct nfs41_free_stateid_args *args)
+{
+	struct compound_hdr hdr = {
+		.minorversion = nfs4_xdr_minorversion(&args->seq_args),
+	};
+
+	encode_compound_hdr(xdr, req, &hdr);
+	encode_sequence(xdr, &args->seq_args, &hdr);
+	encode_free_stateid(xdr, args, &hdr);
+	encode_nops(&hdr);
+}
 #endif /* CONFIG_NFS_V4_1 */
 
 static void print_overflow_msg(const char *func, const struct xdr_stream *xdr)
@@ -4977,11 +5098,17 @@ static int decode_exchange_id(struct xdr_stream *xdr,
 	if (unlikely(status))
 		return status;
 
-	/* Throw away server_scope */
+	/* Save server_scope */
 	status = decode_opaque_inline(xdr, &dummy, &dummy_str);
 	if (unlikely(status))
 		return status;
 
+	if (unlikely(dummy > NFS4_OPAQUE_LIMIT))
+		return -EIO;
+
+	memcpy(res->server_scope->server_scope, dummy_str, dummy);
+	res->server_scope->server_scope_sz = dummy;
+
 	/* Throw away Implementation id array */
 	status = decode_opaque_inline(xdr, &dummy, &dummy_str);
 	if (unlikely(status))
@@ -5322,6 +5449,55 @@ out_overflow:
 	print_overflow_msg(__func__, xdr);
 	return -EIO;
 }
+
+static int decode_test_stateid(struct xdr_stream *xdr,
+			       struct nfs41_test_stateid_res *res)
+{
+	__be32 *p;
+	int status;
+	int num_res;
+
+	status = decode_op_hdr(xdr, OP_TEST_STATEID);
+	if (status)
+		return status;
+
+	p = xdr_inline_decode(xdr, 4);
+	if (unlikely(!p))
+		goto out_overflow;
+	num_res = be32_to_cpup(p++);
+	if (num_res != 1)
+		goto out;
+
+	p = xdr_inline_decode(xdr, 4);
+	if (unlikely(!p))
+		goto out_overflow;
+	res->status = be32_to_cpup(p++);
+	return res->status;
+out_overflow:
+	print_overflow_msg(__func__, xdr);
+out:
+	return -EIO;
+}
+
+static int decode_free_stateid(struct xdr_stream *xdr,
+			       struct nfs41_free_stateid_res *res)
+{
+	__be32 *p;
+	int status;
+
+	status = decode_op_hdr(xdr, OP_FREE_STATEID);
+	if (status)
+		return status;
+
+	p = xdr_inline_decode(xdr, 4);
+	if (unlikely(!p))
+		goto out_overflow;
+	res->status = be32_to_cpup(p++);
+	return res->status;
+out_overflow:
+	print_overflow_msg(__func__, xdr);
+	return -EIO;
+}
 #endif /* CONFIG_NFS_V4_1 */
 
 /*
@@ -6461,6 +6637,72 @@ static int nfs4_xdr_dec_layoutcommit(struct rpc_rqst *rqstp,
 out:
 	return status;
 }
+
+/*
+ * Decode SECINFO_NO_NAME response
+ */
+static int nfs4_xdr_dec_secinfo_no_name(struct rpc_rqst *rqstp,
+					struct xdr_stream *xdr,
+					struct nfs4_secinfo_res *res)
+{
+	struct compound_hdr hdr;
+	int status;
+
+	status = decode_compound_hdr(xdr, &hdr);
+	if (status)
+		goto out;
+	status = decode_sequence(xdr, &res->seq_res, rqstp);
+	if (status)
+		goto out;
+	status = decode_putrootfh(xdr);
+	if (status)
+		goto out;
+	status = decode_secinfo(xdr, res);
+out:
+	return status;
+}
+
+/*
+ * Decode TEST_STATEID response
+ */
+static int nfs4_xdr_dec_test_stateid(struct rpc_rqst *rqstp,
+				     struct xdr_stream *xdr,
+				     struct nfs41_test_stateid_res *res)
+{
+	struct compound_hdr hdr;
+	int status;
+
+	status = decode_compound_hdr(xdr, &hdr);
+	if (status)
+		goto out;
+	status = decode_sequence(xdr, &res->seq_res, rqstp);
+	if (status)
+		goto out;
+	status = decode_test_stateid(xdr, res);
+out:
+	return status;
+}
+
+/*
+ * Decode FREE_STATEID response
+ */
+static int nfs4_xdr_dec_free_stateid(struct rpc_rqst *rqstp,
+				     struct xdr_stream *xdr,
+				     struct nfs41_free_stateid_res *res)
+{
+	struct compound_hdr hdr;
+	int status;
+
+	status = decode_compound_hdr(xdr, &hdr);
+	if (status)
+		goto out;
+	status = decode_sequence(xdr, &res->seq_res, rqstp);
+	if (status)
+		goto out;
+	status = decode_free_stateid(xdr, res);
+out:
+	return status;
+}
 #endif /* CONFIG_NFS_V4_1 */
 
 /**
@@ -6663,6 +6905,9 @@ struct rpc_procinfo	nfs4_procedures[] = {
 	PROC(LAYOUTGET,		enc_layoutget,		dec_layoutget),
 	PROC(LAYOUTCOMMIT,	enc_layoutcommit,	dec_layoutcommit),
 	PROC(LAYOUTRETURN,	enc_layoutreturn,	dec_layoutreturn),
+	PROC(SECINFO_NO_NAME,	enc_secinfo_no_name,	dec_secinfo_no_name),
+	PROC(TEST_STATEID,	enc_test_stateid,	dec_test_stateid),
+	PROC(FREE_STATEID,	enc_free_stateid,	dec_free_stateid),
 #endif /* CONFIG_NFS_V4_1 */
 };
 

fs/nfs/objlayout/objio_osd.c

@@ -1000,13 +1000,22 @@ static bool objio_pg_test(struct nfs_pageio_descriptor *pgio,
 	if (!pnfs_generic_pg_test(pgio, prev, req))
 		return false;
 
-	if (pgio->pg_lseg == NULL)
-		return true;
-
 	return pgio->pg_count + req->wb_bytes <=
 			OBJIO_LSEG(pgio->pg_lseg)->max_io_size;
 }
 
+static const struct nfs_pageio_ops objio_pg_read_ops = {
+	.pg_init = pnfs_generic_pg_init_read,
+	.pg_test = objio_pg_test,
+	.pg_doio = pnfs_generic_pg_readpages,
+};
+
+static const struct nfs_pageio_ops objio_pg_write_ops = {
+	.pg_init = pnfs_generic_pg_init_write,
+	.pg_test = objio_pg_test,
+	.pg_doio = pnfs_generic_pg_writepages,
+};
+
 static struct pnfs_layoutdriver_type objlayout_type = {
 	.id = LAYOUT_OSD2_OBJECTS,
 	.name = "LAYOUT_OSD2_OBJECTS",
@@ -1020,7 +1029,8 @@ static struct pnfs_layoutdriver_type objlayout_type = {
 
 	.read_pagelist           = objlayout_read_pagelist,
 	.write_pagelist          = objlayout_write_pagelist,
-	.pg_test                 = objio_pg_test,
+	.pg_read_ops             = &objio_pg_read_ops,
+	.pg_write_ops            = &objio_pg_write_ops,
 
 	.free_deviceid_node	 = objio_free_deviceid_node,
 
@@ -1055,5 +1065,7 @@ objlayout_exit(void)
 	       __func__);
 }
 
+MODULE_ALIAS("nfs-layouttype4-2");
+
 module_init(objlayout_init);
 module_exit(objlayout_exit);

fs/nfs/pagelist.c

@@ -230,7 +230,7 @@ EXPORT_SYMBOL_GPL(nfs_generic_pg_test);
  */
 void nfs_pageio_init(struct nfs_pageio_descriptor *desc,
 		     struct inode *inode,
-		     int (*doio)(struct nfs_pageio_descriptor *),
+		     const struct nfs_pageio_ops *pg_ops,
 		     size_t bsize,
 		     int io_flags)
 {
@@ -240,13 +240,12 @@ void nfs_pageio_init(struct nfs_pageio_descriptor *desc,
 	desc->pg_bsize = bsize;
 	desc->pg_base = 0;
 	desc->pg_moreio = 0;
+	desc->pg_recoalesce = 0;
 	desc->pg_inode = inode;
-	desc->pg_doio = doio;
+	desc->pg_ops = pg_ops;
 	desc->pg_ioflags = io_flags;
 	desc->pg_error = 0;
 	desc->pg_lseg = NULL;
-	desc->pg_test = nfs_generic_pg_test;
-	pnfs_pageio_init(desc, inode);
 }
 
 /**
@@ -276,7 +275,7 @@ static bool nfs_can_coalesce_requests(struct nfs_page *prev,
 		return false;
 	if (prev->wb_pgbase + prev->wb_bytes != PAGE_CACHE_SIZE)
 		return false;
-	return pgio->pg_test(pgio, prev, req);
+	return pgio->pg_ops->pg_test(pgio, prev, req);
 }
 
 /**
@@ -297,6 +296,8 @@ static int nfs_pageio_do_add_request(struct nfs_pageio_descriptor *desc,
 		if (!nfs_can_coalesce_requests(prev, req, desc))
 			return 0;
 	} else {
+		if (desc->pg_ops->pg_init)
+			desc->pg_ops->pg_init(desc, req);
 		desc->pg_base = req->wb_pgbase;
 	}
 	nfs_list_remove_request(req);
@@ -311,7 +312,7 @@ static int nfs_pageio_do_add_request(struct nfs_pageio_descriptor *desc,
 static void nfs_pageio_doio(struct nfs_pageio_descriptor *desc)
 {
 	if (!list_empty(&desc->pg_list)) {
-		int error = desc->pg_doio(desc);
+		int error = desc->pg_ops->pg_doio(desc);
 		if (error < 0)
 			desc->pg_error = error;
 		else
@@ -331,7 +332,7 @@ static void nfs_pageio_doio(struct nfs_pageio_descriptor *desc)
  * Returns true if the request 'req' was successfully coalesced into the
  * existing list of pages 'desc'.
  */
-int nfs_pageio_add_request(struct nfs_pageio_descriptor *desc,
+static int __nfs_pageio_add_request(struct nfs_pageio_descriptor *desc,
 			   struct nfs_page *req)
 {
 	while (!nfs_pageio_do_add_request(desc, req)) {
@@ -340,17 +341,67 @@ int nfs_pageio_add_request(struct nfs_pageio_descriptor *desc,
 		if (desc->pg_error < 0)
 			return 0;
 		desc->pg_moreio = 0;
+		if (desc->pg_recoalesce)
+			return 0;
 	}
 	return 1;
 }
 
+static int nfs_do_recoalesce(struct nfs_pageio_descriptor *desc)
+{
+	LIST_HEAD(head);
+
+	do {
+		list_splice_init(&desc->pg_list, &head);
+		desc->pg_bytes_written -= desc->pg_count;
+		desc->pg_count = 0;
+		desc->pg_base = 0;
+		desc->pg_recoalesce = 0;
+
+		while (!list_empty(&head)) {
+			struct nfs_page *req;
+
+			req = list_first_entry(&head, struct nfs_page, wb_list);
+			nfs_list_remove_request(req);
+			if (__nfs_pageio_add_request(desc, req))
+				continue;
+			if (desc->pg_error < 0)
+				return 0;
+			break;
+		}
+	} while (desc->pg_recoalesce);
+	return 1;
+}
+
+int nfs_pageio_add_request(struct nfs_pageio_descriptor *desc,
+		struct nfs_page *req)
+{
+	int ret;
+
+	do {
+		ret = __nfs_pageio_add_request(desc, req);
+		if (ret)
+			break;
+		if (desc->pg_error < 0)
+			break;
+		ret = nfs_do_recoalesce(desc);
+	} while (ret);
+	return ret;
+}
+
 /**
  * nfs_pageio_complete - Complete I/O on an nfs_pageio_descriptor
  * @desc: pointer to io descriptor
  */
 void nfs_pageio_complete(struct nfs_pageio_descriptor *desc)
 {
-	nfs_pageio_doio(desc);
+	for (;;) {
+		nfs_pageio_doio(desc);
+		if (!desc->pg_recoalesce)
+			break;
+		if (!nfs_do_recoalesce(desc))
+			break;
+	}
 }
 
 /**
@@ -369,7 +420,7 @@ void nfs_pageio_cond_complete(struct nfs_pageio_descriptor *desc, pgoff_t index)
 	if (!list_empty(&desc->pg_list)) {
 		struct nfs_page *prev = nfs_list_entry(desc->pg_list.prev);
 		if (index != prev->wb_index + 1)
-			nfs_pageio_doio(desc);
+			nfs_pageio_complete(desc);
 	}
 }
 

fs/nfs/pnfs.c

@@ -28,6 +28,7 @@
  */
 
 #include <linux/nfs_fs.h>
+#include <linux/nfs_page.h>
 #include "internal.h"
 #include "pnfs.h"
 #include "iostat.h"
@@ -448,11 +449,20 @@ pnfs_destroy_layout(struct nfs_inode *nfsi)
 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);
-	list_splice_init(&clp->cl_layouts, &tmp_list);
+	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)) {
@@ -661,6 +671,7 @@ _pnfs_return_layout(struct inode *ino)
 	lrp->args.stateid = stateid;
 	lrp->args.layout_type = NFS_SERVER(ino)->pnfs_curr_ld->id;
 	lrp->args.inode = ino;
+	lrp->args.layout = lo;
 	lrp->clp = NFS_SERVER(ino)->nfs_client;
 
 	status = nfs4_proc_layoutreturn(lrp);
@@ -920,7 +931,8 @@ pnfs_update_layout(struct inode *ino,
 	};
 	unsigned pg_offset;
 	struct nfs_inode *nfsi = NFS_I(ino);
-	struct nfs_client *clp = NFS_SERVER(ino)->nfs_client;
+	struct nfs_server *server = NFS_SERVER(ino);
+	struct nfs_client *clp = server->nfs_client;
 	struct pnfs_layout_hdr *lo;
 	struct pnfs_layout_segment *lseg = NULL;
 	bool first = false;
@@ -964,7 +976,7 @@ pnfs_update_layout(struct inode *ino,
 		 */
 		spin_lock(&clp->cl_lock);
 		BUG_ON(!list_empty(&lo->plh_layouts));
-		list_add_tail(&lo->plh_layouts, &clp->cl_layouts);
+		list_add_tail(&lo->plh_layouts, &server->layouts);
 		spin_unlock(&clp->cl_lock);
 	}
 
@@ -973,7 +985,8 @@ pnfs_update_layout(struct inode *ino,
 		arg.offset -= pg_offset;
 		arg.length += pg_offset;
 	}
-	arg.length = PAGE_CACHE_ALIGN(arg.length);
+	if (arg.length != NFS4_MAX_UINT64)
+		arg.length = PAGE_CACHE_ALIGN(arg.length);
 
 	lseg = send_layoutget(lo, ctx, &arg, gfp_flags);
 	if (!lseg && first) {
@@ -991,6 +1004,7 @@ out_unlock:
 	spin_unlock(&ino->i_lock);
 	goto out;
 }
+EXPORT_SYMBOL_GPL(pnfs_update_layout);
 
 int
 pnfs_layout_process(struct nfs4_layoutget *lgp)
@@ -1048,35 +1062,71 @@ out_forget_reply:
 	goto out;
 }
 
+void
+pnfs_generic_pg_init_read(struct nfs_pageio_descriptor *pgio, struct nfs_page *req)
+{
+	BUG_ON(pgio->pg_lseg != NULL);
+
+	pgio->pg_lseg = pnfs_update_layout(pgio->pg_inode,
+					   req->wb_context,
+					   req_offset(req),
+					   req->wb_bytes,
+					   IOMODE_READ,
+					   GFP_KERNEL);
+	/* If no lseg, fall back to read through mds */
+	if (pgio->pg_lseg == NULL)
+		nfs_pageio_reset_read_mds(pgio);
+
+}
+EXPORT_SYMBOL_GPL(pnfs_generic_pg_init_read);
+
+void
+pnfs_generic_pg_init_write(struct nfs_pageio_descriptor *pgio, struct nfs_page *req)
+{
+	BUG_ON(pgio->pg_lseg != NULL);
+
+	pgio->pg_lseg = pnfs_update_layout(pgio->pg_inode,
+					   req->wb_context,
+					   req_offset(req),
+					   req->wb_bytes,
+					   IOMODE_RW,
+					   GFP_NOFS);
+	/* If no lseg, fall back to write through mds */
+	if (pgio->pg_lseg == NULL)
+		nfs_pageio_reset_write_mds(pgio);
+}
+EXPORT_SYMBOL_GPL(pnfs_generic_pg_init_write);
+
 bool
-pnfs_generic_pg_test(struct nfs_pageio_descriptor *pgio, struct nfs_page *prev,
-		     struct nfs_page *req)
+pnfs_pageio_init_read(struct nfs_pageio_descriptor *pgio, struct inode *inode)
 {
-	enum pnfs_iomode access_type;
-	gfp_t gfp_flags;
+	struct nfs_server *server = NFS_SERVER(inode);
+	struct pnfs_layoutdriver_type *ld = server->pnfs_curr_ld;
 
-	/* We assume that pg_ioflags == 0 iff we're reading a page */
-	if (pgio->pg_ioflags == 0) {
-		access_type = IOMODE_READ;
-		gfp_flags = GFP_KERNEL;
-	} else {
-		access_type = IOMODE_RW;
-		gfp_flags = GFP_NOFS;
-	}
+	if (ld == NULL)
+		return false;
+	nfs_pageio_init(pgio, inode, ld->pg_read_ops, server->rsize, 0);
+	return true;
+}
 
-	if (pgio->pg_lseg == NULL) {
-		if (pgio->pg_count != prev->wb_bytes)
-			return true;
-		/* This is first coelesce call for a series of nfs_pages */
-		pgio->pg_lseg = pnfs_update_layout(pgio->pg_inode,
-						   prev->wb_context,
-						   req_offset(prev),
-						   pgio->pg_count,
-						   access_type,
-						   gfp_flags);
-		if (pgio->pg_lseg == NULL)
-			return true;
-	}
+bool
+pnfs_pageio_init_write(struct nfs_pageio_descriptor *pgio, struct inode *inode, int ioflags)
+{
+	struct nfs_server *server = NFS_SERVER(inode);
+	struct pnfs_layoutdriver_type *ld = server->pnfs_curr_ld;
+
+	if (ld == NULL)
+		return false;
+	nfs_pageio_init(pgio, inode, ld->pg_write_ops, server->wsize, ioflags);
+	return true;
+}
+
+bool
+pnfs_generic_pg_test(struct nfs_pageio_descriptor *pgio, struct nfs_page *prev,
+		     struct nfs_page *req)
+{
+	if (pgio->pg_lseg == NULL)
+		return nfs_generic_pg_test(pgio, prev, req);
 
 	/*
 	 * Test if a nfs_page is fully contained in the pnfs_layout_range.
@@ -1120,15 +1170,30 @@ pnfs_ld_write_done(struct nfs_write_data *data)
 }
 EXPORT_SYMBOL_GPL(pnfs_ld_write_done);
 
-enum pnfs_try_status
+static void
+pnfs_write_through_mds(struct nfs_pageio_descriptor *desc,
+		struct nfs_write_data *data)
+{
+	list_splice_tail_init(&data->pages, &desc->pg_list);
+	if (data->req && list_empty(&data->req->wb_list))
+		nfs_list_add_request(data->req, &desc->pg_list);
+	nfs_pageio_reset_write_mds(desc);
+	desc->pg_recoalesce = 1;
+	nfs_writedata_release(data);
+}
+
+static enum pnfs_try_status
 pnfs_try_to_write_data(struct nfs_write_data *wdata,
-			const struct rpc_call_ops *call_ops, int how)
+			const struct rpc_call_ops *call_ops,
+			struct pnfs_layout_segment *lseg,
+			int how)
 {
 	struct inode *inode = wdata->inode;
 	enum pnfs_try_status trypnfs;
 	struct nfs_server *nfss = NFS_SERVER(inode);
 
 	wdata->mds_ops = call_ops;
+	wdata->lseg = get_lseg(lseg);
 
 	dprintk("%s: Writing ino:%lu %u@%llu (how %d)\n", __func__,
 		inode->i_ino, wdata->args.count, wdata->args.offset, how);
@@ -1144,6 +1209,44 @@ pnfs_try_to_write_data(struct nfs_write_data *wdata,
 	return trypnfs;
 }
 
+static void
+pnfs_do_multiple_writes(struct nfs_pageio_descriptor *desc, struct list_head *head, int how)
+{
+	struct nfs_write_data *data;
+	const struct rpc_call_ops *call_ops = desc->pg_rpc_callops;
+	struct pnfs_layout_segment *lseg = desc->pg_lseg;
+
+	desc->pg_lseg = NULL;
+	while (!list_empty(head)) {
+		enum pnfs_try_status trypnfs;
+
+		data = list_entry(head->next, struct nfs_write_data, list);
+		list_del_init(&data->list);
+
+		trypnfs = pnfs_try_to_write_data(data, call_ops, lseg, how);
+		if (trypnfs == PNFS_NOT_ATTEMPTED)
+			pnfs_write_through_mds(desc, data);
+	}
+	put_lseg(lseg);
+}
+
+int
+pnfs_generic_pg_writepages(struct nfs_pageio_descriptor *desc)
+{
+	LIST_HEAD(head);
+	int ret;
+
+	ret = nfs_generic_flush(desc, &head);
+	if (ret != 0) {
+		put_lseg(desc->pg_lseg);
+		desc->pg_lseg = NULL;
+		return ret;
+	}
+	pnfs_do_multiple_writes(desc, &head, desc->pg_ioflags);
+	return 0;
+}
+EXPORT_SYMBOL_GPL(pnfs_generic_pg_writepages);
+
 /*
  * Called by non rpc-based layout drivers
  */
@@ -1167,18 +1270,32 @@ pnfs_ld_read_done(struct nfs_read_data *data)
 }
 EXPORT_SYMBOL_GPL(pnfs_ld_read_done);
 
+static void
+pnfs_read_through_mds(struct nfs_pageio_descriptor *desc,
+		struct nfs_read_data *data)
+{
+	list_splice_tail_init(&data->pages, &desc->pg_list);
+	if (data->req && list_empty(&data->req->wb_list))
+		nfs_list_add_request(data->req, &desc->pg_list);
+	nfs_pageio_reset_read_mds(desc);
+	desc->pg_recoalesce = 1;
+	nfs_readdata_release(data);
+}
+
 /*
  * Call the appropriate parallel I/O subsystem read function.
  */
-enum pnfs_try_status
+static enum pnfs_try_status
 pnfs_try_to_read_data(struct nfs_read_data *rdata,
-		       const struct rpc_call_ops *call_ops)
+		       const struct rpc_call_ops *call_ops,
+		       struct pnfs_layout_segment *lseg)
 {
 	struct inode *inode = rdata->inode;
 	struct nfs_server *nfss = NFS_SERVER(inode);
 	enum pnfs_try_status trypnfs;
 
 	rdata->mds_ops = call_ops;
+	rdata->lseg = get_lseg(lseg);
 
 	dprintk("%s: Reading ino:%lu %u@%llu\n",
 		__func__, inode->i_ino, rdata->args.count, rdata->args.offset);
@@ -1194,6 +1311,44 @@ pnfs_try_to_read_data(struct nfs_read_data *rdata,
 	return trypnfs;
 }
 
+static void
+pnfs_do_multiple_reads(struct nfs_pageio_descriptor *desc, struct list_head *head)
+{
+	struct nfs_read_data *data;
+	const struct rpc_call_ops *call_ops = desc->pg_rpc_callops;
+	struct pnfs_layout_segment *lseg = desc->pg_lseg;
+
+	desc->pg_lseg = NULL;
+	while (!list_empty(head)) {
+		enum pnfs_try_status trypnfs;
+
+		data = list_entry(head->next, struct nfs_read_data, list);
+		list_del_init(&data->list);
+
+		trypnfs = pnfs_try_to_read_data(data, call_ops, lseg);
+		if (trypnfs == PNFS_NOT_ATTEMPTED)
+			pnfs_read_through_mds(desc, data);
+	}
+	put_lseg(lseg);
+}
+
+int
+pnfs_generic_pg_readpages(struct nfs_pageio_descriptor *desc)
+{
+	LIST_HEAD(head);
+	int ret;
+
+	ret = nfs_generic_pagein(desc, &head);
+	if (ret != 0) {
+		put_lseg(desc->pg_lseg);
+		desc->pg_lseg = NULL;
+		return ret;
+	}
+	pnfs_do_multiple_reads(desc, &head);
+	return 0;
+}
+EXPORT_SYMBOL_GPL(pnfs_generic_pg_readpages);
+
 /*
  * Currently there is only one (whole file) write lseg.
  */

fs/nfs/pnfs.h

@@ -87,7 +87,8 @@ struct pnfs_layoutdriver_type {
 	void (*free_lseg) (struct pnfs_layout_segment *lseg);
 
 	/* test for nfs page cache coalescing */
-	bool (*pg_test)(struct nfs_pageio_descriptor *, struct nfs_page *, struct nfs_page *);
+	const struct nfs_pageio_ops *pg_read_ops;
+	const struct nfs_pageio_ops *pg_write_ops;
 
 	/* Returns true if layoutdriver wants to divert this request to
 	 * driver's commit routine.
@@ -148,16 +149,16 @@ extern int nfs4_proc_layoutreturn(struct nfs4_layoutreturn *lrp);
 /* pnfs.c */
 void get_layout_hdr(struct pnfs_layout_hdr *lo);
 void put_lseg(struct pnfs_layout_segment *lseg);
-struct pnfs_layout_segment *
-pnfs_update_layout(struct inode *ino, struct nfs_open_context *ctx,
-		   loff_t pos, u64 count, enum pnfs_iomode access_type,
-		   gfp_t gfp_flags);
+
+bool pnfs_pageio_init_read(struct nfs_pageio_descriptor *, struct inode *);
+bool pnfs_pageio_init_write(struct nfs_pageio_descriptor *, struct inode *, int);
+
 void set_pnfs_layoutdriver(struct nfs_server *, u32 id);
 void unset_pnfs_layoutdriver(struct nfs_server *);
-enum pnfs_try_status pnfs_try_to_write_data(struct nfs_write_data *,
-					     const struct rpc_call_ops *, int);
-enum pnfs_try_status pnfs_try_to_read_data(struct nfs_read_data *,
-					    const struct rpc_call_ops *);
+void pnfs_generic_pg_init_read(struct nfs_pageio_descriptor *, struct nfs_page *);
+int pnfs_generic_pg_readpages(struct nfs_pageio_descriptor *desc);
+void pnfs_generic_pg_init_write(struct nfs_pageio_descriptor *, struct nfs_page *);
+int pnfs_generic_pg_writepages(struct nfs_pageio_descriptor *desc);
 bool pnfs_generic_pg_test(struct nfs_pageio_descriptor *pgio, struct nfs_page *prev, struct nfs_page *req);
 int pnfs_layout_process(struct nfs4_layoutget *lgp);
 void pnfs_free_lseg_list(struct list_head *tmp_list);
@@ -182,6 +183,19 @@ int pnfs_layoutcommit_inode(struct inode *inode, bool sync);
 int _pnfs_return_layout(struct inode *);
 int pnfs_ld_write_done(struct nfs_write_data *);
 int pnfs_ld_read_done(struct nfs_read_data *);
+struct pnfs_layout_segment *pnfs_update_layout(struct inode *ino,
+					       struct nfs_open_context *ctx,
+					       loff_t pos,
+					       u64 count,
+					       enum pnfs_iomode iomode,
+					       gfp_t gfp_flags);
+
+void nfs4_deviceid_mark_client_invalid(struct nfs_client *clp);
+
+/* nfs4_deviceid_flags */
+enum {
+	NFS_DEVICEID_INVALID = 0,       /* set when MDS clientid recalled */
+};
 
 /* pnfs_dev.c */
 struct nfs4_deviceid_node {
@@ -189,13 +203,13 @@ struct nfs4_deviceid_node {
 	struct hlist_node		tmpnode;
 	const struct pnfs_layoutdriver_type *ld;
 	const struct nfs_client		*nfs_client;
+	unsigned long 			flags;
 	struct nfs4_deviceid		deviceid;
 	atomic_t			ref;
 };
 
 void nfs4_print_deviceid(const struct nfs4_deviceid *dev_id);
 struct nfs4_deviceid_node *nfs4_find_get_deviceid(const struct pnfs_layoutdriver_type *, const struct nfs_client *, const struct nfs4_deviceid *);
-struct nfs4_deviceid_node *nfs4_unhash_put_deviceid(const struct pnfs_layoutdriver_type *, const struct nfs_client *, const struct nfs4_deviceid *);
 void nfs4_delete_deviceid(const struct pnfs_layoutdriver_type *, const struct nfs_client *, const struct nfs4_deviceid *);
 void nfs4_init_deviceid_node(struct nfs4_deviceid_node *,
 			     const struct pnfs_layoutdriver_type *,
@@ -293,15 +307,6 @@ static inline int pnfs_return_layout(struct inode *ino)
 	return 0;
 }
 
-static inline void pnfs_pageio_init(struct nfs_pageio_descriptor *pgio,
-				    struct inode *inode)
-{
-	struct pnfs_layoutdriver_type *ld = NFS_SERVER(inode)->pnfs_curr_ld;
-
-	if (ld)
-		pgio->pg_test = ld->pg_test;
-}
-
 #else  /* CONFIG_NFS_V4_1 */
 
 static inline void pnfs_destroy_all_layouts(struct nfs_client *clp)
@@ -322,28 +327,6 @@ static inline void put_lseg(struct pnfs_layout_segment *lseg)
 {
 }
 
-static inline struct pnfs_layout_segment *
-pnfs_update_layout(struct inode *ino, struct nfs_open_context *ctx,
-		   loff_t pos, u64 count, enum pnfs_iomode access_type,
-		   gfp_t gfp_flags)
-{
-	return NULL;
-}
-
-static inline enum pnfs_try_status
-pnfs_try_to_read_data(struct nfs_read_data *data,
-		      const struct rpc_call_ops *call_ops)
-{
-	return PNFS_NOT_ATTEMPTED;
-}
-
-static inline enum pnfs_try_status
-pnfs_try_to_write_data(struct nfs_write_data *data,
-		       const struct rpc_call_ops *call_ops, int how)
-{
-	return PNFS_NOT_ATTEMPTED;
-}
-
 static inline int pnfs_return_layout(struct inode *ino)
 {
 	return 0;
@@ -385,9 +368,14 @@ static inline void unset_pnfs_layoutdriver(struct nfs_server *s)
 {
 }
 
-static inline void pnfs_pageio_init(struct nfs_pageio_descriptor *pgio,
-				    struct inode *inode)
+static inline bool pnfs_pageio_init_read(struct nfs_pageio_descriptor *pgio, struct inode *inode)
 {
+	return false;
+}
+
+static inline bool pnfs_pageio_init_write(struct nfs_pageio_descriptor *pgio, struct inode *inode, int ioflags)
+{
+	return false;
 }
 
 static inline void

fs/nfs/pnfs_dev.c

@@ -100,8 +100,8 @@ _find_get_deviceid(const struct pnfs_layoutdriver_type *ld,
 
 	rcu_read_lock();
 	d = _lookup_deviceid(ld, clp, id, hash);
-	if (d && !atomic_inc_not_zero(&d->ref))
-		d = NULL;
+	if (d != NULL)
+		atomic_inc(&d->ref);
 	rcu_read_unlock();
 	return d;
 }
@@ -115,15 +115,15 @@ nfs4_find_get_deviceid(const struct pnfs_layoutdriver_type *ld,
 EXPORT_SYMBOL_GPL(nfs4_find_get_deviceid);
 
 /*
- * Unhash and put deviceid
+ * Remove a deviceid from cache
  *
  * @clp nfs_client associated with deviceid
  * @id the deviceid to unhash
  *
  * @ret the unhashed node, if found and dereferenced to zero, NULL otherwise.
  */
-struct nfs4_deviceid_node *
-nfs4_unhash_put_deviceid(const struct pnfs_layoutdriver_type *ld,
+void
+nfs4_delete_deviceid(const struct pnfs_layoutdriver_type *ld,
 			 const struct nfs_client *clp, const struct nfs4_deviceid *id)
 {
 	struct nfs4_deviceid_node *d;
@@ -134,7 +134,7 @@ nfs4_unhash_put_deviceid(const struct pnfs_layoutdriver_type *ld,
 	rcu_read_unlock();
 	if (!d) {
 		spin_unlock(&nfs4_deviceid_lock);
-		return NULL;
+		return;
 	}
 	hlist_del_init_rcu(&d->node);
 	spin_unlock(&nfs4_deviceid_lock);
@@ -142,28 +142,7 @@ nfs4_unhash_put_deviceid(const struct pnfs_layoutdriver_type *ld,
 
 	/* balance the initial ref set in pnfs_insert_deviceid */
 	if (atomic_dec_and_test(&d->ref))
-		return d;
-
-	return NULL;
-}
-EXPORT_SYMBOL_GPL(nfs4_unhash_put_deviceid);
-
-/*
- * Delete a deviceid from cache
- *
- * @clp struct nfs_client qualifying the deviceid
- * @id deviceid to delete
- */
-void
-nfs4_delete_deviceid(const struct pnfs_layoutdriver_type *ld,
-		     const struct nfs_client *clp, const struct nfs4_deviceid *id)
-{
-	struct nfs4_deviceid_node *d;
-
-	d = nfs4_unhash_put_deviceid(ld, clp, id);
-	if (!d)
-		return;
-	d->ld->free_deviceid_node(d);
+		d->ld->free_deviceid_node(d);
 }
 EXPORT_SYMBOL_GPL(nfs4_delete_deviceid);
 
@@ -177,6 +156,7 @@ nfs4_init_deviceid_node(struct nfs4_deviceid_node *d,
 	INIT_HLIST_NODE(&d->tmpnode);
 	d->ld = ld;
 	d->nfs_client = nfs_client;
+	d->flags = 0;
 	d->deviceid = *id;
 	atomic_set(&d->ref, 1);
 }
@@ -221,16 +201,15 @@ EXPORT_SYMBOL_GPL(nfs4_insert_deviceid_node);
  *
  * @d deviceid node to put
  *
- * @ret true iff the node was deleted
+ * return true iff the node was deleted
+ * Note that since the test for d->ref == 0 is sufficient to establish
+ * that the node is no longer hashed in the global device id cache.
  */
 bool
 nfs4_put_deviceid_node(struct nfs4_deviceid_node *d)
 {
-	if (!atomic_dec_and_lock(&d->ref, &nfs4_deviceid_lock))
+	if (!atomic_dec_and_test(&d->ref))
 		return false;
-	hlist_del_init_rcu(&d->node);
-	spin_unlock(&nfs4_deviceid_lock);
-	synchronize_rcu();
 	d->ld->free_deviceid_node(d);
 	return true;
 }
@@ -275,3 +254,22 @@ nfs4_deviceid_purge_client(const struct nfs_client *clp)
 	for (h = 0; h < NFS4_DEVICE_ID_HASH_SIZE; h++)
 		_deviceid_purge_client(clp, h);
 }
+
+/*
+ * Stop use of all deviceids associated with an nfs_client
+ */
+void
+nfs4_deviceid_mark_client_invalid(struct nfs_client *clp)
+{
+	struct nfs4_deviceid_node *d;
+	struct hlist_node *n;
+	int i;
+
+	rcu_read_lock();
+	for (i = 0; i < NFS4_DEVICE_ID_HASH_SIZE; i ++){
+		hlist_for_each_entry_rcu(d, n, &nfs4_deviceid_cache[i], node)
+			if (d->nfs_client == clp)
+				set_bit(NFS_DEVICEID_INVALID, &d->flags);
+	}
+	rcu_read_unlock();
+}

fs/nfs/read.c

@@ -30,8 +30,7 @@
 
 #define NFSDBG_FACILITY		NFSDBG_PAGECACHE
 
-static int nfs_pagein_multi(struct nfs_pageio_descriptor *desc);
-static int nfs_pagein_one(struct nfs_pageio_descriptor *desc);
+static const struct nfs_pageio_ops nfs_pageio_read_ops;
 static const struct rpc_call_ops nfs_read_partial_ops;
 static const struct rpc_call_ops nfs_read_full_ops;
 
@@ -68,7 +67,7 @@ void nfs_readdata_free(struct nfs_read_data *p)
 	mempool_free(p, nfs_rdata_mempool);
 }
 
-static void nfs_readdata_release(struct nfs_read_data *rdata)
+void nfs_readdata_release(struct nfs_read_data *rdata)
 {
 	put_lseg(rdata->lseg);
 	put_nfs_open_context(rdata->args.context);
@@ -113,6 +112,27 @@ static void nfs_readpage_truncate_uninitialised_page(struct nfs_read_data *data)
 	}
 }
 
+static void nfs_pageio_init_read_mds(struct nfs_pageio_descriptor *pgio,
+		struct inode *inode)
+{
+	nfs_pageio_init(pgio, inode, &nfs_pageio_read_ops,
+			NFS_SERVER(inode)->rsize, 0);
+}
+
+void nfs_pageio_reset_read_mds(struct nfs_pageio_descriptor *pgio)
+{
+	pgio->pg_ops = &nfs_pageio_read_ops;
+	pgio->pg_bsize = NFS_SERVER(pgio->pg_inode)->rsize;
+}
+EXPORT_SYMBOL_GPL(nfs_pageio_reset_read_mds);
+
+static void nfs_pageio_init_read(struct nfs_pageio_descriptor *pgio,
+		struct inode *inode)
+{
+	if (!pnfs_pageio_init_read(pgio, inode))
+		nfs_pageio_init_read_mds(pgio, inode);
+}
+
 int nfs_readpage_async(struct nfs_open_context *ctx, struct inode *inode,
 		       struct page *page)
 {
@@ -131,14 +151,9 @@ int nfs_readpage_async(struct nfs_open_context *ctx, struct inode *inode,
 	if (len < PAGE_CACHE_SIZE)
 		zero_user_segment(page, len, PAGE_CACHE_SIZE);
 
-	nfs_pageio_init(&pgio, inode, NULL, 0, 0);
-	nfs_list_add_request(new, &pgio.pg_list);
-	pgio.pg_count = len;
-
-	if (NFS_SERVER(inode)->rsize < PAGE_CACHE_SIZE)
-		nfs_pagein_multi(&pgio);
-	else
-		nfs_pagein_one(&pgio);
+	nfs_pageio_init_read(&pgio, inode);
+	nfs_pageio_add_request(&pgio, new);
+	nfs_pageio_complete(&pgio);
 	return 0;
 }
 
@@ -202,17 +217,14 @@ EXPORT_SYMBOL_GPL(nfs_initiate_read);
 /*
  * Set up the NFS read request struct
  */
-static int nfs_read_rpcsetup(struct nfs_page *req, struct nfs_read_data *data,
-		const struct rpc_call_ops *call_ops,
-		unsigned int count, unsigned int offset,
-		struct pnfs_layout_segment *lseg)
+static void nfs_read_rpcsetup(struct nfs_page *req, struct nfs_read_data *data,
+		unsigned int count, unsigned int offset)
 {
 	struct inode *inode = req->wb_context->dentry->d_inode;
 
 	data->req	  = req;
 	data->inode	  = inode;
 	data->cred	  = req->wb_context->cred;
-	data->lseg	  = get_lseg(lseg);
 
 	data->args.fh     = NFS_FH(inode);
 	data->args.offset = req_offset(req) + offset;
@@ -226,14 +238,36 @@ static int nfs_read_rpcsetup(struct nfs_page *req, struct nfs_read_data *data,
 	data->res.count   = count;
 	data->res.eof     = 0;
 	nfs_fattr_init(&data->fattr);
+}
 
-	if (data->lseg &&
-	    (pnfs_try_to_read_data(data, call_ops) == PNFS_ATTEMPTED))
-		return 0;
+static int nfs_do_read(struct nfs_read_data *data,
+		const struct rpc_call_ops *call_ops)
+{
+	struct inode *inode = data->args.context->dentry->d_inode;
 
 	return nfs_initiate_read(data, NFS_CLIENT(inode), call_ops);
 }
 
+static int
+nfs_do_multiple_reads(struct list_head *head,
+		const struct rpc_call_ops *call_ops)
+{
+	struct nfs_read_data *data;
+	int ret = 0;
+
+	while (!list_empty(head)) {
+		int ret2;
+
+		data = list_entry(head->next, struct nfs_read_data, list);
+		list_del_init(&data->list);
+
+		ret2 = nfs_do_read(data, call_ops);
+		if (ret == 0)
+			ret = ret2;
+	}
+	return ret;
+}
+
 static void
 nfs_async_read_error(struct list_head *head)
 {
@@ -260,20 +294,19 @@ nfs_async_read_error(struct list_head *head)
  * won't see the new data until our attribute cache is updated.  This is more
  * or less conventional NFS client behavior.
  */
-static int nfs_pagein_multi(struct nfs_pageio_descriptor *desc)
+static int nfs_pagein_multi(struct nfs_pageio_descriptor *desc, struct list_head *res)
 {
 	struct nfs_page *req = nfs_list_entry(desc->pg_list.next);
 	struct page *page = req->wb_page;
 	struct nfs_read_data *data;
-	size_t rsize = NFS_SERVER(desc->pg_inode)->rsize, nbytes;
+	size_t rsize = desc->pg_bsize, nbytes;
 	unsigned int offset;
 	int requests = 0;
 	int ret = 0;
-	struct pnfs_layout_segment *lseg;
-	LIST_HEAD(list);
 
 	nfs_list_remove_request(req);
 
+	offset = 0;
 	nbytes = desc->pg_count;
 	do {
 		size_t len = min(nbytes,rsize);
@@ -281,45 +314,21 @@ static int nfs_pagein_multi(struct nfs_pageio_descriptor *desc)
 		data = nfs_readdata_alloc(1);
 		if (!data)
 			goto out_bad;
-		list_add(&data->pages, &list);
+		data->pagevec[0] = page;
+		nfs_read_rpcsetup(req, data, len, offset);
+		list_add(&data->list, res);
 		requests++;
 		nbytes -= len;
+		offset += len;
 	} while(nbytes != 0);
 	atomic_set(&req->wb_complete, requests);
-
-	BUG_ON(desc->pg_lseg != NULL);
-	lseg = pnfs_update_layout(desc->pg_inode, req->wb_context,
-				  req_offset(req), desc->pg_count,
-				  IOMODE_READ, GFP_KERNEL);
 	ClearPageError(page);
-	offset = 0;
-	nbytes = desc->pg_count;
-	do {
-		int ret2;
-
-		data = list_entry(list.next, struct nfs_read_data, pages);
-		list_del_init(&data->pages);
-
-		data->pagevec[0] = page;
-
-		if (nbytes < rsize)
-			rsize = nbytes;
-		ret2 = nfs_read_rpcsetup(req, data, &nfs_read_partial_ops,
-					 rsize, offset, lseg);
-		if (ret == 0)
-			ret = ret2;
-		offset += rsize;
-		nbytes -= rsize;
-	} while (nbytes != 0);
-	put_lseg(lseg);
-	desc->pg_lseg = NULL;
-
+	desc->pg_rpc_callops = &nfs_read_partial_ops;
 	return ret;
-
 out_bad:
-	while (!list_empty(&list)) {
-		data = list_entry(list.next, struct nfs_read_data, pages);
-		list_del(&data->pages);
+	while (!list_empty(res)) {
+		data = list_entry(res->next, struct nfs_read_data, list);
+		list_del(&data->list);
 		nfs_readdata_free(data);
 	}
 	SetPageError(page);
@@ -327,19 +336,19 @@ out_bad:
 	return -ENOMEM;
 }
 
-static int nfs_pagein_one(struct nfs_pageio_descriptor *desc)
+static int nfs_pagein_one(struct nfs_pageio_descriptor *desc, struct list_head *res)
 {
 	struct nfs_page		*req;
 	struct page		**pages;
 	struct nfs_read_data	*data;
 	struct list_head *head = &desc->pg_list;
-	struct pnfs_layout_segment *lseg = desc->pg_lseg;
-	int ret = -ENOMEM;
+	int ret = 0;
 
 	data = nfs_readdata_alloc(nfs_page_array_len(desc->pg_base,
 						     desc->pg_count));
 	if (!data) {
 		nfs_async_read_error(head);
+		ret = -ENOMEM;
 		goto out;
 	}
 
@@ -352,19 +361,37 @@ static int nfs_pagein_one(struct nfs_pageio_descriptor *desc)
 		*pages++ = req->wb_page;
 	}
 	req = nfs_list_entry(data->pages.next);
-	if ((!lseg) && list_is_singular(&data->pages))
-		lseg = pnfs_update_layout(desc->pg_inode, req->wb_context,
-					  req_offset(req), desc->pg_count,
-					  IOMODE_READ, GFP_KERNEL);
 
-	ret = nfs_read_rpcsetup(req, data, &nfs_read_full_ops, desc->pg_count,
-				0, lseg);
+	nfs_read_rpcsetup(req, data, desc->pg_count, 0);
+	list_add(&data->list, res);
+	desc->pg_rpc_callops = &nfs_read_full_ops;
 out:
-	put_lseg(lseg);
-	desc->pg_lseg = NULL;
 	return ret;
 }
 
+int nfs_generic_pagein(struct nfs_pageio_descriptor *desc, struct list_head *head)
+{
+	if (desc->pg_bsize < PAGE_CACHE_SIZE)
+		return nfs_pagein_multi(desc, head);
+	return nfs_pagein_one(desc, head);
+}
+
+static int nfs_generic_pg_readpages(struct nfs_pageio_descriptor *desc)
+{
+	LIST_HEAD(head);
+	int ret;
+
+	ret = nfs_generic_pagein(desc, &head);
+	if (ret == 0)
+		ret = nfs_do_multiple_reads(&head, desc->pg_rpc_callops);
+	return ret;
+}
+
+static const struct nfs_pageio_ops nfs_pageio_read_ops = {
+	.pg_test = nfs_generic_pg_test,
+	.pg_doio = nfs_generic_pg_readpages,
+};
+
 /*
  * This is the callback from RPC telling us whether a reply was
  * received or some error occurred (timeout or socket shutdown).
@@ -635,8 +662,6 @@ int nfs_readpages(struct file *filp, struct address_space *mapping,
 		.pgio = &pgio,
 	};
 	struct inode *inode = mapping->host;
-	struct nfs_server *server = NFS_SERVER(inode);
-	size_t rsize = server->rsize;
 	unsigned long npages;
 	int ret = -ESTALE;
 
@@ -664,10 +689,7 @@ int nfs_readpages(struct file *filp, struct address_space *mapping,
 	if (ret == 0)
 		goto read_complete; /* all pages were read */
 
-	if (rsize < PAGE_CACHE_SIZE)
-		nfs_pageio_init(&pgio, inode, nfs_pagein_multi, rsize, 0);
-	else
-		nfs_pageio_init(&pgio, inode, nfs_pagein_one, rsize, 0);
+	nfs_pageio_init_read(&pgio, inode);
 
 	ret = read_cache_pages(mapping, pages, readpage_async_filler, &desc);
 

fs/nfs/unlink.c

@@ -147,7 +147,7 @@ static int nfs_do_call_unlink(struct dentry *parent, struct inode *dir, struct n
 
 	alias = d_lookup(parent, &data->args.name);
 	if (alias != NULL) {
-		int ret = 0;
+		int ret;
 		void *devname_garbage = NULL;
 
 		/*
@@ -155,14 +155,16 @@ static int nfs_do_call_unlink(struct dentry *parent, struct inode *dir, struct n
 		 * the sillyrename information to the aliased dentry.
 		 */
 		nfs_free_dname(data);
+		ret = nfs_copy_dname(alias, data);
 		spin_lock(&alias->d_lock);
-		if (alias->d_inode != NULL &&
+		if (ret == 0 && alias->d_inode != NULL &&
 		    !(alias->d_flags & DCACHE_NFSFS_RENAMED)) {
 			devname_garbage = alias->d_fsdata;
 			alias->d_fsdata = data;
 			alias->d_flags |= DCACHE_NFSFS_RENAMED;
 			ret = 1;
-		}
+		} else
+			ret = 0;
 		spin_unlock(&alias->d_lock);
 		nfs_dec_sillycount(dir);
 		dput(alias);
@@ -171,8 +173,7 @@ static int nfs_do_call_unlink(struct dentry *parent, struct inode *dir, struct n
 		 * point dentry is definitely not a root, so we won't need
 		 * that anymore.
 		 */
-		if (devname_garbage)
-			kfree(devname_garbage);
+		kfree(devname_garbage);
 		return ret;
 	}
 	data->dir = igrab(dir);
@@ -204,8 +205,6 @@ static int nfs_call_unlink(struct dentry *dentry, struct nfs_unlinkdata *data)
 	if (parent == NULL)
 		goto out_free;
 	dir = parent->d_inode;
-	if (nfs_copy_dname(dentry, data) != 0)
-		goto out_dput;
 	/* Non-exclusive lock protects against concurrent lookup() calls */
 	spin_lock(&dir->i_lock);
 	if (atomic_inc_not_zero(&NFS_I(dir)->silly_count) == 0) {
@@ -366,6 +365,8 @@ static void nfs_async_rename_done(struct rpc_task *task, void *calldata)
 	struct nfs_renamedata *data = calldata;
 	struct inode *old_dir = data->old_dir;
 	struct inode *new_dir = data->new_dir;
+	struct dentry *old_dentry = data->old_dentry;
+	struct dentry *new_dentry = data->new_dentry;
 
 	if (!NFS_PROTO(old_dir)->rename_done(task, old_dir, new_dir)) {
 		nfs_restart_rpc(task, NFS_SERVER(old_dir)->nfs_client);
@@ -373,12 +374,12 @@ static void nfs_async_rename_done(struct rpc_task *task, void *calldata)
 	}
 
 	if (task->tk_status != 0) {
-		nfs_cancel_async_unlink(data->old_dentry);
+		nfs_cancel_async_unlink(old_dentry);
 		return;
 	}
 
-	nfs_set_verifier(data->old_dentry, nfs_save_change_attribute(old_dir));
-	d_move(data->old_dentry, data->new_dentry);
+	d_drop(old_dentry);
+	d_drop(new_dentry);
 }
 
 /**
@@ -501,6 +502,14 @@ nfs_async_rename(struct inode *old_dir, struct inode *new_dir,
  * and only performs the unlink once the last reference to it is put.
  *
  * The final cleanup is done during dentry_iput.
+ *
+ * (Note: NFSv4 is stateful, and has opens, so in theory an NFSv4 server
+ * could take responsibility for keeping open files referenced.  The server
+ * would also need to ensure that opened-but-deleted files were kept over
+ * reboots.  However, we may not assume a server does so.  (RFC 5661
+ * does provide an OPEN4_RESULT_PRESERVE_UNLINKED flag that a server can
+ * use to advertise that it does this; some day we may take advantage of
+ * it.))
  */
 int
 nfs_sillyrename(struct inode *dir, struct dentry *dentry)
@@ -560,6 +569,14 @@ nfs_sillyrename(struct inode *dir, struct dentry *dentry)
 	if (error)
 		goto out_dput;
 
+	/* populate unlinkdata with the right dname */
+	error = nfs_copy_dname(sdentry,
+				(struct nfs_unlinkdata *)dentry->d_fsdata);
+	if (error) {
+		nfs_cancel_async_unlink(dentry);
+		goto out_dput;
+	}
+
 	/* run the rename task, undo unlink if it fails */
 	task = nfs_async_rename(dir, dir, dentry, sdentry);
 	if (IS_ERR(task)) {

fs/nfs/write.c

@@ -97,7 +97,7 @@ void nfs_writedata_free(struct nfs_write_data *p)
 	mempool_free(p, nfs_wdata_mempool);
 }
 
-static void nfs_writedata_release(struct nfs_write_data *wdata)
+void nfs_writedata_release(struct nfs_write_data *wdata)
 {
 	put_lseg(wdata->lseg);
 	put_nfs_open_context(wdata->args.context);
@@ -845,11 +845,9 @@ EXPORT_SYMBOL_GPL(nfs_initiate_write);
 /*
  * Set up the argument/result storage required for the RPC call.
  */
-static int nfs_write_rpcsetup(struct nfs_page *req,
+static void nfs_write_rpcsetup(struct nfs_page *req,
 		struct nfs_write_data *data,
-		const struct rpc_call_ops *call_ops,
 		unsigned int count, unsigned int offset,
-		struct pnfs_layout_segment *lseg,
 		int how)
 {
 	struct inode *inode = req->wb_context->dentry->d_inode;
@@ -860,7 +858,6 @@ static int nfs_write_rpcsetup(struct nfs_page *req,
 	data->req = req;
 	data->inode = inode = req->wb_context->dentry->d_inode;
 	data->cred = req->wb_context->cred;
-	data->lseg = get_lseg(lseg);
 
 	data->args.fh     = NFS_FH(inode);
 	data->args.offset = req_offset(req) + offset;
@@ -872,24 +869,51 @@ static int nfs_write_rpcsetup(struct nfs_page *req,
 	data->args.context = get_nfs_open_context(req->wb_context);
 	data->args.lock_context = req->wb_lock_context;
 	data->args.stable  = NFS_UNSTABLE;
-	if (how & (FLUSH_STABLE | FLUSH_COND_STABLE)) {
-		data->args.stable = NFS_DATA_SYNC;
-		if (!nfs_need_commit(NFS_I(inode)))
-			data->args.stable = NFS_FILE_SYNC;
+	switch (how & (FLUSH_STABLE | FLUSH_COND_STABLE)) {
+	case 0:
+		break;
+	case FLUSH_COND_STABLE:
+		if (nfs_need_commit(NFS_I(inode)))
+			break;
+	default:
+		data->args.stable = NFS_FILE_SYNC;
 	}
 
 	data->res.fattr   = &data->fattr;
 	data->res.count   = count;
 	data->res.verf    = &data->verf;
 	nfs_fattr_init(&data->fattr);
+}
 
-	if (data->lseg &&
-	    (pnfs_try_to_write_data(data, call_ops, how) == PNFS_ATTEMPTED))
-		return 0;
+static int nfs_do_write(struct nfs_write_data *data,
+		const struct rpc_call_ops *call_ops,
+		int how)
+{
+	struct inode *inode = data->args.context->dentry->d_inode;
 
 	return nfs_initiate_write(data, NFS_CLIENT(inode), call_ops, how);
 }
 
+static int nfs_do_multiple_writes(struct list_head *head,
+		const struct rpc_call_ops *call_ops,
+		int how)
+{
+	struct nfs_write_data *data;
+	int ret = 0;
+
+	while (!list_empty(head)) {
+		int ret2;
+
+		data = list_entry(head->next, struct nfs_write_data, list);
+		list_del_init(&data->list);
+		
+		ret2 = nfs_do_write(data, call_ops, how);
+		 if (ret == 0)
+			 ret = ret2;
+	}
+	return ret;
+}
+
 /* If a nfs_flush_* function fails, it should remove reqs from @head and
  * call this on each, which will prepare them to be retried on next
  * writeback using standard nfs.
@@ -907,17 +931,15 @@ static void nfs_redirty_request(struct nfs_page *req)
  * Generate multiple small requests to write out a single
  * contiguous dirty area on one page.
  */
-static int nfs_flush_multi(struct nfs_pageio_descriptor *desc)
+static int nfs_flush_multi(struct nfs_pageio_descriptor *desc, struct list_head *res)
 {
 	struct nfs_page *req = nfs_list_entry(desc->pg_list.next);
 	struct page *page = req->wb_page;
 	struct nfs_write_data *data;
-	size_t wsize = NFS_SERVER(desc->pg_inode)->wsize, nbytes;
+	size_t wsize = desc->pg_bsize, nbytes;
 	unsigned int offset;
 	int requests = 0;
 	int ret = 0;
-	struct pnfs_layout_segment *lseg;
-	LIST_HEAD(list);
 
 	nfs_list_remove_request(req);
 
@@ -927,6 +949,7 @@ static int nfs_flush_multi(struct nfs_pageio_descriptor *desc)
 		desc->pg_ioflags &= ~FLUSH_COND_STABLE;
 
 
+	offset = 0;
 	nbytes = desc->pg_count;
 	do {
 		size_t len = min(nbytes, wsize);
@@ -934,45 +957,21 @@ static int nfs_flush_multi(struct nfs_pageio_descriptor *desc)
 		data = nfs_writedata_alloc(1);
 		if (!data)
 			goto out_bad;
-		list_add(&data->pages, &list);
+		data->pagevec[0] = page;
+		nfs_write_rpcsetup(req, data, wsize, offset, desc->pg_ioflags);
+		list_add(&data->list, res);
 		requests++;
 		nbytes -= len;
+		offset += len;
 	} while (nbytes != 0);
 	atomic_set(&req->wb_complete, requests);
-
-	BUG_ON(desc->pg_lseg);
-	lseg = pnfs_update_layout(desc->pg_inode, req->wb_context,
-				  req_offset(req), desc->pg_count,
-				  IOMODE_RW, GFP_NOFS);
-	ClearPageError(page);
-	offset = 0;
-	nbytes = desc->pg_count;
-	do {
-		int ret2;
-
-		data = list_entry(list.next, struct nfs_write_data, pages);
-		list_del_init(&data->pages);
-
-		data->pagevec[0] = page;
-
-		if (nbytes < wsize)
-			wsize = nbytes;
-		ret2 = nfs_write_rpcsetup(req, data, &nfs_write_partial_ops,
-					  wsize, offset, lseg, desc->pg_ioflags);
-		if (ret == 0)
-			ret = ret2;
-		offset += wsize;
-		nbytes -= wsize;
-	} while (nbytes != 0);
-
-	put_lseg(lseg);
-	desc->pg_lseg = NULL;
+	desc->pg_rpc_callops = &nfs_write_partial_ops;
 	return ret;
 
 out_bad:
-	while (!list_empty(&list)) {
-		data = list_entry(list.next, struct nfs_write_data, pages);
-		list_del(&data->pages);
+	while (!list_empty(res)) {
+		data = list_entry(res->next, struct nfs_write_data, list);
+		list_del(&data->list);
 		nfs_writedata_free(data);
 	}
 	nfs_redirty_request(req);
@@ -987,14 +986,13 @@ out_bad:
  * This is the case if nfs_updatepage detects a conflicting request
  * that has been written but not committed.
  */
-static int nfs_flush_one(struct nfs_pageio_descriptor *desc)
+static int nfs_flush_one(struct nfs_pageio_descriptor *desc, struct list_head *res)
 {
 	struct nfs_page		*req;
 	struct page		**pages;
 	struct nfs_write_data	*data;
 	struct list_head *head = &desc->pg_list;
-	struct pnfs_layout_segment *lseg = desc->pg_lseg;
-	int ret;
+	int ret = 0;
 
 	data = nfs_writedata_alloc(nfs_page_array_len(desc->pg_base,
 						      desc->pg_count));
@@ -1016,32 +1014,62 @@ static int nfs_flush_one(struct nfs_pageio_descriptor *desc)
 		*pages++ = req->wb_page;
 	}
 	req = nfs_list_entry(data->pages.next);
-	if ((!lseg) && list_is_singular(&data->pages))
-		lseg = pnfs_update_layout(desc->pg_inode, req->wb_context,
-					  req_offset(req), desc->pg_count,
-					  IOMODE_RW, GFP_NOFS);
 
 	if ((desc->pg_ioflags & FLUSH_COND_STABLE) &&
 	    (desc->pg_moreio || NFS_I(desc->pg_inode)->ncommit))
 		desc->pg_ioflags &= ~FLUSH_COND_STABLE;
 
 	/* Set up the argument struct */
-	ret = nfs_write_rpcsetup(req, data, &nfs_write_full_ops, desc->pg_count, 0, lseg, desc->pg_ioflags);
+	nfs_write_rpcsetup(req, data, desc->pg_count, 0, desc->pg_ioflags);
+	list_add(&data->list, res);
+	desc->pg_rpc_callops = &nfs_write_full_ops;
 out:
-	put_lseg(lseg); /* Cleans any gotten in ->pg_test */
-	desc->pg_lseg = NULL;
 	return ret;
 }
 
-static void nfs_pageio_init_write(struct nfs_pageio_descriptor *pgio,
+int nfs_generic_flush(struct nfs_pageio_descriptor *desc, struct list_head *head)
+{
+	if (desc->pg_bsize < PAGE_CACHE_SIZE)
+		return nfs_flush_multi(desc, head);
+	return nfs_flush_one(desc, head);
+}
+
+static int nfs_generic_pg_writepages(struct nfs_pageio_descriptor *desc)
+{
+	LIST_HEAD(head);
+	int ret;
+
+	ret = nfs_generic_flush(desc, &head);
+	if (ret == 0)
+		ret = nfs_do_multiple_writes(&head, desc->pg_rpc_callops,
+				desc->pg_ioflags);
+	return ret;
+}
+
+static const struct nfs_pageio_ops nfs_pageio_write_ops = {
+	.pg_test = nfs_generic_pg_test,
+	.pg_doio = nfs_generic_pg_writepages,
+};
+
+static void nfs_pageio_init_write_mds(struct nfs_pageio_descriptor *pgio,
 				  struct inode *inode, int ioflags)
 {
-	size_t wsize = NFS_SERVER(inode)->wsize;
+	nfs_pageio_init(pgio, inode, &nfs_pageio_write_ops,
+				NFS_SERVER(inode)->wsize, ioflags);
+}
+
+void nfs_pageio_reset_write_mds(struct nfs_pageio_descriptor *pgio)
+{
+	pgio->pg_ops = &nfs_pageio_write_ops;
+	pgio->pg_bsize = NFS_SERVER(pgio->pg_inode)->wsize;
+}
+EXPORT_SYMBOL_GPL(nfs_pageio_reset_write_mds);
 
-	if (wsize < PAGE_CACHE_SIZE)
-		nfs_pageio_init(pgio, inode, nfs_flush_multi, wsize, ioflags);
-	else
-		nfs_pageio_init(pgio, inode, nfs_flush_one, wsize, ioflags);
+static void nfs_pageio_init_write(struct nfs_pageio_descriptor *pgio,
+				  struct inode *inode, int ioflags)
+{
+	if (!pnfs_pageio_init_write(pgio, inode, ioflags))
+		nfs_pageio_init_write_mds(pgio, inode, ioflags);
 }
 
 /*

include/linux/nfs4.h

@@ -563,6 +563,9 @@ enum {
 	NFSPROC4_CLNT_GETDEVICEINFO,
 	NFSPROC4_CLNT_LAYOUTCOMMIT,
 	NFSPROC4_CLNT_LAYOUTRETURN,
+	NFSPROC4_CLNT_SECINFO_NO_NAME,
+	NFSPROC4_CLNT_TEST_STATEID,
+	NFSPROC4_CLNT_FREE_STATEID,
 };
 
 /* nfs41 types */

include/linux/nfs_fs_sb.h

@@ -16,6 +16,7 @@ struct nfs4_sequence_args;
 struct nfs4_sequence_res;
 struct nfs_server;
 struct nfs4_minor_version_ops;
+struct server_scope;
 
 /*
  * The nfs_client identifies our client state to the server.
@@ -77,12 +78,13 @@ struct nfs_client {
 	/* The flags used for obtaining the clientid during EXCHANGE_ID */
 	u32			cl_exchange_flags;
 	struct nfs4_session	*cl_session; 	/* sharred session */
-	struct list_head	cl_layouts;
 #endif /* CONFIG_NFS_V4 */
 
 #ifdef CONFIG_NFS_FSCACHE
 	struct fscache_cookie	*fscache;	/* client index cache cookie */
 #endif
+
+	struct server_scope	*server_scope;	/* from exchange_id */
 };
 
 /*
@@ -149,6 +151,7 @@ struct nfs_server {
 	struct rb_root		openowner_id;
 	struct rb_root		lockowner_id;
 #endif
+	struct list_head	layouts;
 	struct list_head	delegations;
 	void (*destroy)(struct nfs_server *);
 

include/linux/nfs_page.h

@@ -55,20 +55,28 @@ struct nfs_page {
 	struct nfs_writeverf	wb_verf;	/* Commit cookie */
 };
 
+struct nfs_pageio_descriptor;
+struct nfs_pageio_ops {
+	void	(*pg_init)(struct nfs_pageio_descriptor *, struct nfs_page *);
+	bool	(*pg_test)(struct nfs_pageio_descriptor *, struct nfs_page *, struct nfs_page *);
+	int	(*pg_doio)(struct nfs_pageio_descriptor *);
+};
+
 struct nfs_pageio_descriptor {
 	struct list_head	pg_list;
 	unsigned long		pg_bytes_written;
 	size_t			pg_count;
 	size_t			pg_bsize;
 	unsigned int		pg_base;
-	char			pg_moreio;
+	unsigned char		pg_moreio : 1,
+				pg_recoalesce : 1;
 
 	struct inode		*pg_inode;
-	int			(*pg_doio)(struct nfs_pageio_descriptor *);
+	const struct nfs_pageio_ops *pg_ops;
 	int 			pg_ioflags;
 	int			pg_error;
+	const struct rpc_call_ops *pg_rpc_callops;
 	struct pnfs_layout_segment *pg_lseg;
-	bool			(*pg_test)(struct nfs_pageio_descriptor *, struct nfs_page *, struct nfs_page *);
 };
 
 #define NFS_WBACK_BUSY(req)	(test_bit(PG_BUSY,&(req)->wb_flags))
@@ -85,7 +93,7 @@ extern	int nfs_scan_list(struct nfs_inode *nfsi, struct list_head *dst,
 			  pgoff_t idx_start, unsigned int npages, int tag);
 extern	void nfs_pageio_init(struct nfs_pageio_descriptor *desc,
 			     struct inode *inode,
-			     int (*doio)(struct nfs_pageio_descriptor *desc),
+			     const struct nfs_pageio_ops *pg_ops,
 			     size_t bsize,
 			     int how);
 extern	int nfs_pageio_add_request(struct nfs_pageio_descriptor *,
@@ -100,7 +108,6 @@ extern	void nfs_unlock_request(struct nfs_page *req);
 extern	int nfs_set_page_tag_locked(struct nfs_page *req);
 extern  void nfs_clear_page_tag_locked(struct nfs_page *req);
 
-
 /*
  * Lock the page of an asynchronous request without getting a new reference
  */

include/linux/nfs_xdr.h

@@ -269,9 +269,10 @@ struct nfs4_layoutcommit_data {
 };
 
 struct nfs4_layoutreturn_args {
-	__u32   layout_type;
+	struct pnfs_layout_hdr *layout;
 	struct inode *inode;
 	nfs4_stateid stateid;
+	__u32   layout_type;
 	struct nfs4_sequence_args seq_args;
 };
 
@@ -1060,6 +1061,7 @@ struct server_scope {
 struct nfs41_exchange_id_res {
 	struct nfs_client		*client;
 	u32				flags;
+	struct server_scope		*server_scope;
 };
 
 struct nfs41_create_session_args {
@@ -1083,6 +1085,34 @@ struct nfs41_reclaim_complete_args {
 struct nfs41_reclaim_complete_res {
 	struct nfs4_sequence_res	seq_res;
 };
+
+#define SECINFO_STYLE_CURRENT_FH 0
+#define SECINFO_STYLE_PARENT 1
+struct nfs41_secinfo_no_name_args {
+	int				style;
+	struct nfs4_sequence_args	seq_args;
+};
+
+struct nfs41_test_stateid_args {
+	nfs4_stateid			*stateid;
+	struct nfs4_sequence_args	seq_args;
+};
+
+struct nfs41_test_stateid_res {
+	unsigned int			status;
+	struct nfs4_sequence_res	seq_res;
+};
+
+struct nfs41_free_stateid_args {
+	nfs4_stateid			*stateid;
+	struct nfs4_sequence_args	seq_args;
+};
+
+struct nfs41_free_stateid_res {
+	unsigned int			status;
+	struct nfs4_sequence_res	seq_res;
+};
+
 #endif /* CONFIG_NFS_V4_1 */
 
 struct nfs_page;
@@ -1096,6 +1126,7 @@ struct nfs_read_data {
 	struct rpc_cred		*cred;
 	struct nfs_fattr	fattr;	/* fattr storage */
 	struct list_head	pages;	/* Coalesced read requests */
+	struct list_head	list;	/* lists of struct nfs_read_data */
 	struct nfs_page		*req;	/* multi ops per nfs_page */
 	struct page		**pagevec;
 	unsigned int		npages;	/* Max length of pagevec */
@@ -1119,6 +1150,7 @@ struct nfs_write_data {
 	struct nfs_fattr	fattr;
 	struct nfs_writeverf	verf;
 	struct list_head	pages;		/* Coalesced requests we wish to flush */
+	struct list_head	list;		/* lists of struct nfs_write_data */
 	struct nfs_page		*req;		/* multi ops per nfs_page */
 	struct page		**pagevec;
 	unsigned int		npages;		/* Max length of pagevec */

include/linux/pnfs_osd_xdr.h

@@ -41,9 +41,6 @@
 
 #include <linux/nfs_fs.h>
 #include <linux/nfs_page.h>
-#include <scsi/osd_protocol.h>
-
-#define PNFS_OSD_OSDNAME_MAXSIZE 256
 
 /*
  * draft-ietf-nfsv4-minorversion-22
@@ -99,12 +96,6 @@ struct pnfs_osd_objid {
 #define _DEVID_HI(oid_device_id) \
 	(unsigned long long)be64_to_cpup(((__be64 *)(oid_device_id)->data) + 1)
 
-static inline int
-pnfs_osd_objid_xdr_sz(void)
-{
-	return (NFS4_DEVICEID4_SIZE / 4) + 2 + 2;
-}
-
 enum pnfs_osd_version {
 	PNFS_OSD_MISSING              = 0,
 	PNFS_OSD_VERSION_1            = 1,
@@ -189,8 +180,6 @@ struct pnfs_osd_targetid {
 	struct nfs4_string		oti_scsi_device_id;
 };
 
-enum { PNFS_OSD_TARGETID_MAX = 1 + PNFS_OSD_OSDNAME_MAXSIZE / 4 };
-
 /*   struct netaddr4 {
  *       // see struct rpcb in RFC1833
  *       string r_netid<>;    // network id
@@ -207,12 +196,6 @@ struct pnfs_osd_targetaddr {
 	struct pnfs_osd_net_addr	ota_netaddr;
 };
 
-enum {
-	NETWORK_ID_MAX = 16 / 4,
-	UNIVERSAL_ADDRESS_MAX = 64 / 4,
-	PNFS_OSD_TARGETADDR_MAX = 3 +  NETWORK_ID_MAX + UNIVERSAL_ADDRESS_MAX,
-};
-
 struct pnfs_osd_deviceaddr {
 	struct pnfs_osd_targetid	oda_targetid;
 	struct pnfs_osd_targetaddr	oda_targetaddr;
@@ -222,15 +205,6 @@ struct pnfs_osd_deviceaddr {
 	struct nfs4_string		oda_osdname;
 };
 
-enum {
-	ODA_OSDNAME_MAX = PNFS_OSD_OSDNAME_MAXSIZE / 4,
-	PNFS_OSD_DEVICEADDR_MAX =
-		PNFS_OSD_TARGETID_MAX + PNFS_OSD_TARGETADDR_MAX +
-		2 /*oda_lun*/ +
-		1 + OSD_SYSTEMID_LEN +
-		1 + ODA_OSDNAME_MAX,
-};
-
 /* LAYOUTCOMMIT: layoutupdate */
 
 /*   union pnfs_osd_deltaspaceused4 switch (bool dsu_valid) {
@@ -279,7 +253,7 @@ struct pnfs_osd_ioerr {
 	u32			oer_errno;
 };
 
-/* OSD XDR API */
+/* OSD XDR Client API */
 /* Layout helpers */
 /* Layout decoding is done in two parts:
  * 1. First Call pnfs_osd_xdr_decode_layout_map to read in only the header part
@@ -337,8 +311,7 @@ extern int
 pnfs_osd_xdr_encode_layoutupdate(struct xdr_stream *xdr,
 				 struct pnfs_osd_layoutupdate *lou);
 
-/* osd_ioerror encoding/decoding (layout_return) */
-/* Client */
+/* osd_ioerror encoding (layout_return) */
 extern __be32 *pnfs_osd_xdr_ioerr_reserve_space(struct xdr_stream *xdr);
 extern void pnfs_osd_xdr_encode_ioerr(__be32 *p, struct pnfs_osd_ioerr *ioerr);
 

include/linux/sunrpc/bc_xprt.h

@@ -31,7 +31,7 @@ SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 #include <linux/sunrpc/xprt.h>
 #include <linux/sunrpc/sched.h>
 
-#ifdef CONFIG_NFS_V4_1
+#ifdef CONFIG_SUNRPC_BACKCHANNEL
 struct rpc_rqst *xprt_alloc_bc_request(struct rpc_xprt *xprt);
 void xprt_free_bc_request(struct rpc_rqst *req);
 int xprt_setup_backchannel(struct rpc_xprt *, unsigned int min_reqs);
@@ -47,7 +47,7 @@ static inline int svc_is_backchannel(const struct svc_rqst *rqstp)
 		return 1;
 	return 0;
 }
-#else /* CONFIG_NFS_V4_1 */
+#else /* CONFIG_SUNRPC_BACKCHANNEL */
 static inline int xprt_setup_backchannel(struct rpc_xprt *xprt,
 					 unsigned int min_reqs)
 {
@@ -62,6 +62,6 @@ static inline int svc_is_backchannel(const struct svc_rqst *rqstp)
 static inline void xprt_free_bc_request(struct rpc_rqst *req)
 {
 }
-#endif /* CONFIG_NFS_V4_1 */
+#endif /* CONFIG_SUNRPC_BACKCHANNEL */
 #endif /* _LINUX_SUNRPC_BC_XPRT_H */
 

include/linux/sunrpc/sched.h

@@ -227,6 +227,10 @@ void		rpc_init_wait_queue(struct rpc_wait_queue *, const char *);
 void		rpc_destroy_wait_queue(struct rpc_wait_queue *);
 void		rpc_sleep_on(struct rpc_wait_queue *, struct rpc_task *,
 					rpc_action action);
+void		rpc_sleep_on_priority(struct rpc_wait_queue *,
+					struct rpc_task *,
+					rpc_action action,
+					int priority);
 void		rpc_wake_up_queued_task(struct rpc_wait_queue *,
 					struct rpc_task *);
 void		rpc_wake_up(struct rpc_wait_queue *);

include/linux/sunrpc/svc.h

@@ -92,7 +92,7 @@ struct svc_serv {
 	struct module *		sv_module;	/* optional module to count when
 						 * adding threads */
 	svc_thread_fn		sv_function;	/* main function for threads */
-#if defined(CONFIG_NFS_V4_1)
+#if defined(CONFIG_SUNRPC_BACKCHANNEL)
 	struct list_head	sv_cb_list;	/* queue for callback requests
 						 * that arrive over the same
 						 * connection */
@@ -100,7 +100,7 @@ struct svc_serv {
 	wait_queue_head_t	sv_cb_waitq;	/* sleep here if there are no
 						 * entries in the svc_cb_list */
 	struct svc_xprt		*sv_bc_xprt;	/* callback on fore channel */
-#endif /* CONFIG_NFS_V4_1 */
+#endif /* CONFIG_SUNRPC_BACKCHANNEL */
 };
 
 /*

include/linux/sunrpc/xprt.h

@@ -22,6 +22,7 @@
 #define RPC_MIN_SLOT_TABLE	(2U)
 #define RPC_DEF_SLOT_TABLE	(16U)
 #define RPC_MAX_SLOT_TABLE	(128U)
+#define RPC_MAX_SLOT_TABLE_LIMIT	(65536U)
 
 /*
  * This describes a timeout strategy
@@ -100,18 +101,18 @@ struct rpc_rqst {
 	ktime_t			rq_xtime;	/* transmit time stamp */
 	int			rq_ntrans;
 
-#if defined(CONFIG_NFS_V4_1)
+#if defined(CONFIG_SUNRPC_BACKCHANNEL)
 	struct list_head	rq_bc_list;	/* Callback service list */
 	unsigned long		rq_bc_pa_state;	/* Backchannel prealloc state */
 	struct list_head	rq_bc_pa_list;	/* Backchannel prealloc list */
-#endif /* CONFIG_NFS_V4_1 */
+#endif /* CONFIG_SUNRPC_BACKCHANEL */
 };
 #define rq_svec			rq_snd_buf.head
 #define rq_slen			rq_snd_buf.len
 
 struct rpc_xprt_ops {
 	void		(*set_buffer_size)(struct rpc_xprt *xprt, size_t sndsize, size_t rcvsize);
-	int		(*reserve_xprt)(struct rpc_task *task);
+	int		(*reserve_xprt)(struct rpc_xprt *xprt, struct rpc_task *task);
 	void		(*release_xprt)(struct rpc_xprt *xprt, struct rpc_task *task);
 	void		(*rpcbind)(struct rpc_task *task);
 	void		(*set_port)(struct rpc_xprt *xprt, unsigned short port);
@@ -164,12 +165,12 @@ struct rpc_xprt {
 
 	struct rpc_wait_queue	binding;	/* requests waiting on rpcbind */
 	struct rpc_wait_queue	sending;	/* requests waiting to send */
-	struct rpc_wait_queue	resend;		/* requests waiting to resend */
 	struct rpc_wait_queue	pending;	/* requests in flight */
 	struct rpc_wait_queue	backlog;	/* waiting for slot */
 	struct list_head	free;		/* free slots */
-	struct rpc_rqst *	slot;		/* slot table storage */
-	unsigned int		max_reqs;	/* total slots */
+	unsigned int		max_reqs;	/* max number of slots */
+	unsigned int		min_reqs;	/* min number of slots */
+	atomic_t		num_reqs;	/* total slots */
 	unsigned long		state;		/* transport state */
 	unsigned char		shutdown   : 1,	/* being shut down */
 				resvport   : 1; /* use a reserved port */
@@ -200,7 +201,7 @@ struct rpc_xprt {
 	u32			xid;		/* Next XID value to use */
 	struct rpc_task *	snd_task;	/* Task blocked in send */
 	struct svc_xprt		*bc_xprt;	/* NFSv4.1 backchannel */
-#if defined(CONFIG_NFS_V4_1)
+#if defined(CONFIG_SUNRPC_BACKCHANNEL)
 	struct svc_serv		*bc_serv;       /* The RPC service which will */
 						/* process the callback */
 	unsigned int		bc_alloc_count;	/* Total number of preallocs */
@@ -208,7 +209,7 @@ struct rpc_xprt {
 						 * items */
 	struct list_head	bc_pa_list;	/* List of preallocated
 						 * backchannel rpc_rqst's */
-#endif /* CONFIG_NFS_V4_1 */
+#endif /* CONFIG_SUNRPC_BACKCHANNEL */
 	struct list_head	recv;
 
 	struct {
@@ -228,15 +229,15 @@ struct rpc_xprt {
 	const char		*address_strings[RPC_DISPLAY_MAX];
 };
 
-#if defined(CONFIG_NFS_V4_1)
+#if defined(CONFIG_SUNRPC_BACKCHANNEL)
 /*
  * Backchannel flags
  */
 #define	RPC_BC_PA_IN_USE	0x0001		/* Preallocated backchannel */
 						/* buffer in use */
-#endif /* CONFIG_NFS_V4_1 */
+#endif /* CONFIG_SUNRPC_BACKCHANNEL */
 
-#if defined(CONFIG_NFS_V4_1)
+#if defined(CONFIG_SUNRPC_BACKCHANNEL)
 static inline int bc_prealloc(struct rpc_rqst *req)
 {
 	return test_bit(RPC_BC_PA_IN_USE, &req->rq_bc_pa_state);
@@ -246,7 +247,7 @@ static inline int bc_prealloc(struct rpc_rqst *req)
 {
 	return 0;
 }
-#endif /* CONFIG_NFS_V4_1 */
+#endif /* CONFIG_SUNRPC_BACKCHANNEL */
 
 struct xprt_create {
 	int			ident;		/* XPRT_TRANSPORT identifier */
@@ -271,8 +272,8 @@ struct xprt_class {
 struct rpc_xprt		*xprt_create_transport(struct xprt_create *args);
 void			xprt_connect(struct rpc_task *task);
 void			xprt_reserve(struct rpc_task *task);
-int			xprt_reserve_xprt(struct rpc_task *task);
-int			xprt_reserve_xprt_cong(struct rpc_task *task);
+int			xprt_reserve_xprt(struct rpc_xprt *xprt, struct rpc_task *task);
+int			xprt_reserve_xprt_cong(struct rpc_xprt *xprt, struct rpc_task *task);
 int			xprt_prepare_transmit(struct rpc_task *task);
 void			xprt_transmit(struct rpc_task *task);
 void			xprt_end_transmit(struct rpc_task *task);
@@ -282,7 +283,9 @@ void			xprt_release_xprt_cong(struct rpc_xprt *xprt, struct rpc_task *task);
 void			xprt_release(struct rpc_task *task);
 struct rpc_xprt *	xprt_get(struct rpc_xprt *xprt);
 void			xprt_put(struct rpc_xprt *xprt);
-struct rpc_xprt *	xprt_alloc(struct net *net, int size, int max_req);
+struct rpc_xprt *	xprt_alloc(struct net *net, size_t size,
+				unsigned int num_prealloc,
+				unsigned int max_req);
 void			xprt_free(struct rpc_xprt *);
 
 static inline __be32 *xprt_skip_transport_header(struct rpc_xprt *xprt, __be32 *p)
@@ -321,7 +324,6 @@ void			xprt_conditional_disconnect(struct rpc_xprt *xprt, unsigned int cookie);
 #define XPRT_CLOSING		(6)
 #define XPRT_CONNECTION_ABORT	(7)
 #define XPRT_CONNECTION_CLOSE	(8)
-#define XPRT_INITIALIZED	(9)
 
 static inline void xprt_set_connected(struct rpc_xprt *xprt)
 {

net/sunrpc/Kconfig

@@ -4,6 +4,10 @@ config SUNRPC
 config SUNRPC_GSS
 	tristate
 
+config SUNRPC_BACKCHANNEL
+	bool
+	depends on SUNRPC
+
 config SUNRPC_XPRT_RDMA
 	tristate
 	depends on SUNRPC && INFINIBAND && INFINIBAND_ADDR_TRANS && EXPERIMENTAL

net/sunrpc/Makefile

@@ -13,6 +13,6 @@ sunrpc-y := clnt.o xprt.o socklib.o xprtsock.o sched.o \
 	    addr.o rpcb_clnt.o timer.o xdr.o \
 	    sunrpc_syms.o cache.o rpc_pipe.o \
 	    svc_xprt.o
-sunrpc-$(CONFIG_NFS_V4_1) += backchannel_rqst.o bc_svc.o
+sunrpc-$(CONFIG_SUNRPC_BACKCHANNEL) += backchannel_rqst.o bc_svc.o
 sunrpc-$(CONFIG_PROC_FS) += stats.o
 sunrpc-$(CONFIG_SYSCTL) += sysctl.o

net/sunrpc/backchannel_rqst.c

@@ -29,8 +29,6 @@ SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 #define RPCDBG_FACILITY	RPCDBG_TRANS
 #endif
 
-#if defined(CONFIG_NFS_V4_1)
-
 /*
  * Helper routines that track the number of preallocation elements
  * on the transport.
@@ -174,7 +172,7 @@ out_free:
 	dprintk("RPC:       setup backchannel transport failed\n");
 	return -1;
 }
-EXPORT_SYMBOL(xprt_setup_backchannel);
+EXPORT_SYMBOL_GPL(xprt_setup_backchannel);
 
 /*
  * Destroys the backchannel preallocated structures.
@@ -204,7 +202,7 @@ void xprt_destroy_backchannel(struct rpc_xprt *xprt, unsigned int max_reqs)
 	dprintk("RPC:        backchannel list empty= %s\n",
 		list_empty(&xprt->bc_pa_list) ? "true" : "false");
 }
-EXPORT_SYMBOL(xprt_destroy_backchannel);
+EXPORT_SYMBOL_GPL(xprt_destroy_backchannel);
 
 /*
  * One or more rpc_rqst structure have been preallocated during the
@@ -279,4 +277,3 @@ void xprt_free_bc_request(struct rpc_rqst *req)
 	spin_unlock_bh(&xprt->bc_pa_lock);
 }
 
-#endif /* CONFIG_NFS_V4_1 */

net/sunrpc/bc_svc.c

@@ -27,8 +27,6 @@ SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
  * reply over an existing open connection previously established by the client.
  */
 
-#if defined(CONFIG_NFS_V4_1)
-
 #include <linux/module.h>
 
 #include <linux/sunrpc/xprt.h>
@@ -63,4 +61,3 @@ int bc_send(struct rpc_rqst *req)
 	return ret;
 }
 
-#endif /* CONFIG_NFS_V4_1 */

net/sunrpc/clnt.c

@@ -64,9 +64,9 @@ static void	call_decode(struct rpc_task *task);
 static void	call_bind(struct rpc_task *task);
 static void	call_bind_status(struct rpc_task *task);
 static void	call_transmit(struct rpc_task *task);
-#if defined(CONFIG_NFS_V4_1)
+#if defined(CONFIG_SUNRPC_BACKCHANNEL)
 static void	call_bc_transmit(struct rpc_task *task);
-#endif /* CONFIG_NFS_V4_1 */
+#endif /* CONFIG_SUNRPC_BACKCHANNEL */
 static void	call_status(struct rpc_task *task);
 static void	call_transmit_status(struct rpc_task *task);
 static void	call_refresh(struct rpc_task *task);
@@ -715,7 +715,7 @@ rpc_call_async(struct rpc_clnt *clnt, const struct rpc_message *msg, int flags,
 }
 EXPORT_SYMBOL_GPL(rpc_call_async);
 
-#if defined(CONFIG_NFS_V4_1)
+#if defined(CONFIG_SUNRPC_BACKCHANNEL)
 /**
  * rpc_run_bc_task - Allocate a new RPC task for backchannel use, then run
  * rpc_execute against it
@@ -758,7 +758,7 @@ out:
 	dprintk("RPC: rpc_run_bc_task: task= %p\n", task);
 	return task;
 }
-#endif /* CONFIG_NFS_V4_1 */
+#endif /* CONFIG_SUNRPC_BACKCHANNEL */
 
 void
 rpc_call_start(struct rpc_task *task)
@@ -1361,7 +1361,7 @@ call_transmit_status(struct rpc_task *task)
 	}
 }
 
-#if defined(CONFIG_NFS_V4_1)
+#if defined(CONFIG_SUNRPC_BACKCHANNEL)
 /*
  * 5b.	Send the backchannel RPC reply.  On error, drop the reply.  In
  * addition, disconnect on connectivity errors.
@@ -1425,7 +1425,7 @@ call_bc_transmit(struct rpc_task *task)
 	}
 	rpc_wake_up_queued_task(&req->rq_xprt->pending, task);
 }
-#endif /* CONFIG_NFS_V4_1 */
+#endif /* CONFIG_SUNRPC_BACKCHANNEL */
 
 /*
  * 6.	Sort out the RPC call status
@@ -1550,8 +1550,7 @@ call_decode(struct rpc_task *task)
 	kxdrdproc_t	decode = task->tk_msg.rpc_proc->p_decode;
 	__be32		*p;
 
-	dprintk("RPC: %5u call_decode (status %d)\n",
-			task->tk_pid, task->tk_status);
+	dprint_status(task);
 
 	if (task->tk_flags & RPC_CALL_MAJORSEEN) {
 		if (clnt->cl_chatty)

net/sunrpc/sched.c

@@ -97,14 +97,16 @@ __rpc_add_timer(struct rpc_wait_queue *queue, struct rpc_task *task)
 /*
  * Add new request to a priority queue.
  */
-static void __rpc_add_wait_queue_priority(struct rpc_wait_queue *queue, struct rpc_task *task)
+static void __rpc_add_wait_queue_priority(struct rpc_wait_queue *queue,
+		struct rpc_task *task,
+		unsigned char queue_priority)
 {
 	struct list_head *q;
 	struct rpc_task *t;
 
 	INIT_LIST_HEAD(&task->u.tk_wait.links);
-	q = &queue->tasks[task->tk_priority];
-	if (unlikely(task->tk_priority > queue->maxpriority))
+	q = &queue->tasks[queue_priority];
+	if (unlikely(queue_priority > queue->maxpriority))
 		q = &queue->tasks[queue->maxpriority];
 	list_for_each_entry(t, q, u.tk_wait.list) {
 		if (t->tk_owner == task->tk_owner) {
@@ -123,12 +125,14 @@ static void __rpc_add_wait_queue_priority(struct rpc_wait_queue *queue, struct r
  * improve overall performance.
  * Everyone else gets appended to the queue to ensure proper FIFO behavior.
  */
-static void __rpc_add_wait_queue(struct rpc_wait_queue *queue, struct rpc_task *task)
+static void __rpc_add_wait_queue(struct rpc_wait_queue *queue,
+		struct rpc_task *task,
+		unsigned char queue_priority)
 {
 	BUG_ON (RPC_IS_QUEUED(task));
 
 	if (RPC_IS_PRIORITY(queue))
-		__rpc_add_wait_queue_priority(queue, task);
+		__rpc_add_wait_queue_priority(queue, task, queue_priority);
 	else if (RPC_IS_SWAPPER(task))
 		list_add(&task->u.tk_wait.list, &queue->tasks[0]);
 	else
@@ -311,13 +315,15 @@ static void rpc_make_runnable(struct rpc_task *task)
  * NB: An RPC task will only receive interrupt-driven events as long
  * as it's on a wait queue.
  */
-static void __rpc_sleep_on(struct rpc_wait_queue *q, struct rpc_task *task,
-			rpc_action action)
+static void __rpc_sleep_on_priority(struct rpc_wait_queue *q,
+		struct rpc_task *task,
+		rpc_action action,
+		unsigned char queue_priority)
 {
 	dprintk("RPC: %5u sleep_on(queue \"%s\" time %lu)\n",
 			task->tk_pid, rpc_qname(q), jiffies);
 
-	__rpc_add_wait_queue(q, task);
+	__rpc_add_wait_queue(q, task, queue_priority);
 
 	BUG_ON(task->tk_callback != NULL);
 	task->tk_callback = action;
@@ -334,11 +340,25 @@ void rpc_sleep_on(struct rpc_wait_queue *q, struct rpc_task *task,
 	 * Protect the queue operations.
 	 */
 	spin_lock_bh(&q->lock);
-	__rpc_sleep_on(q, task, action);
+	__rpc_sleep_on_priority(q, task, action, task->tk_priority);
 	spin_unlock_bh(&q->lock);
 }
 EXPORT_SYMBOL_GPL(rpc_sleep_on);
 
+void rpc_sleep_on_priority(struct rpc_wait_queue *q, struct rpc_task *task,
+		rpc_action action, int priority)
+{
+	/* We shouldn't ever put an inactive task to sleep */
+	BUG_ON(!RPC_IS_ACTIVATED(task));
+
+	/*
+	 * Protect the queue operations.
+	 */
+	spin_lock_bh(&q->lock);
+	__rpc_sleep_on_priority(q, task, action, priority - RPC_PRIORITY_LOW);
+	spin_unlock_bh(&q->lock);
+}
+
 /**
  * __rpc_do_wake_up_task - wake up a single rpc_task
  * @queue: wait queue

net/sunrpc/svc.c

@@ -1252,7 +1252,7 @@ svc_process(struct svc_rqst *rqstp)
 	}
 }
 
-#if defined(CONFIG_NFS_V4_1)
+#if defined(CONFIG_SUNRPC_BACKCHANNEL)
 /*
  * Process a backchannel RPC request that arrived over an existing
  * outbound connection
@@ -1300,8 +1300,8 @@ bc_svc_process(struct svc_serv *serv, struct rpc_rqst *req,
 		return 0;
 	}
 }
-EXPORT_SYMBOL(bc_svc_process);
-#endif /* CONFIG_NFS_V4_1 */
+EXPORT_SYMBOL_GPL(bc_svc_process);
+#endif /* CONFIG_SUNRPC_BACKCHANNEL */
 
 /*
  * Return (transport-specific) limit on the rpc payload.

net/sunrpc/svcsock.c

@@ -68,12 +68,12 @@ static void		svc_sock_free(struct svc_xprt *);
 static struct svc_xprt *svc_create_socket(struct svc_serv *, int,
 					  struct net *, struct sockaddr *,
 					  int, int);
-#if defined(CONFIG_NFS_V4_1)
+#if defined(CONFIG_SUNRPC_BACKCHANNEL)
 static struct svc_xprt *svc_bc_create_socket(struct svc_serv *, int,
 					     struct net *, struct sockaddr *,
 					     int, int);
 static void svc_bc_sock_free(struct svc_xprt *xprt);
-#endif /* CONFIG_NFS_V4_1 */
+#endif /* CONFIG_SUNRPC_BACKCHANNEL */
 
 #ifdef CONFIG_DEBUG_LOCK_ALLOC
 static struct lock_class_key svc_key[2];
@@ -1243,7 +1243,7 @@ static struct svc_xprt *svc_tcp_create(struct svc_serv *serv,
 	return svc_create_socket(serv, IPPROTO_TCP, net, sa, salen, flags);
 }
 
-#if defined(CONFIG_NFS_V4_1)
+#if defined(CONFIG_SUNRPC_BACKCHANNEL)
 static struct svc_xprt *svc_bc_create_socket(struct svc_serv *, int,
 					     struct net *, struct sockaddr *,
 					     int, int);
@@ -1284,7 +1284,7 @@ static void svc_cleanup_bc_xprt_sock(void)
 {
 	svc_unreg_xprt_class(&svc_tcp_bc_class);
 }
-#else /* CONFIG_NFS_V4_1 */
+#else /* CONFIG_SUNRPC_BACKCHANNEL */
 static void svc_init_bc_xprt_sock(void)
 {
 }
@@ -1292,7 +1292,7 @@ static void svc_init_bc_xprt_sock(void)
 static void svc_cleanup_bc_xprt_sock(void)
 {
 }
-#endif /* CONFIG_NFS_V4_1 */
+#endif /* CONFIG_SUNRPC_BACKCHANNEL */
 
 static struct svc_xprt_ops svc_tcp_ops = {
 	.xpo_create = svc_tcp_create,
@@ -1623,7 +1623,7 @@ static void svc_sock_free(struct svc_xprt *xprt)
 	kfree(svsk);
 }
 
-#if defined(CONFIG_NFS_V4_1)
+#if defined(CONFIG_SUNRPC_BACKCHANNEL)
 /*
  * Create a back channel svc_xprt which shares the fore channel socket.
  */
@@ -1662,4 +1662,4 @@ static void svc_bc_sock_free(struct svc_xprt *xprt)
 	if (xprt)
 		kfree(container_of(xprt, struct svc_sock, sk_xprt));
 }
-#endif /* CONFIG_NFS_V4_1 */
+#endif /* CONFIG_SUNRPC_BACKCHANNEL */

net/sunrpc/xdr.c

@@ -126,7 +126,7 @@ xdr_terminate_string(struct xdr_buf *buf, const u32 len)
 	kaddr[buf->page_base + len] = '\0';
 	kunmap_atomic(kaddr, KM_USER0);
 }
-EXPORT_SYMBOL(xdr_terminate_string);
+EXPORT_SYMBOL_GPL(xdr_terminate_string);
 
 void
 xdr_encode_pages(struct xdr_buf *xdr, struct page **pages, unsigned int base,

net/sunrpc/xprt.c

@@ -62,6 +62,7 @@
 /*
  * Local functions
  */
+static void	 xprt_init(struct rpc_xprt *xprt, struct net *net);
 static void	xprt_request_init(struct rpc_task *, struct rpc_xprt *);
 static void	xprt_connect_status(struct rpc_task *task);
 static int      __xprt_get_cong(struct rpc_xprt *, struct rpc_task *);
@@ -191,10 +192,10 @@ EXPORT_SYMBOL_GPL(xprt_load_transport);
  * transport connects from colliding with writes.  No congestion control
  * is provided.
  */
-int xprt_reserve_xprt(struct rpc_task *task)
+int xprt_reserve_xprt(struct rpc_xprt *xprt, struct rpc_task *task)
 {
 	struct rpc_rqst *req = task->tk_rqstp;
-	struct rpc_xprt	*xprt = req->rq_xprt;
+	int priority;
 
 	if (test_and_set_bit(XPRT_LOCKED, &xprt->state)) {
 		if (task == xprt->snd_task)
@@ -202,8 +203,10 @@ int xprt_reserve_xprt(struct rpc_task *task)
 		goto out_sleep;
 	}
 	xprt->snd_task = task;
-	req->rq_bytes_sent = 0;
-	req->rq_ntrans++;
+	if (req != NULL) {
+		req->rq_bytes_sent = 0;
+		req->rq_ntrans++;
+	}
 
 	return 1;
 
@@ -212,10 +215,13 @@ out_sleep:
 			task->tk_pid, xprt);
 	task->tk_timeout = 0;
 	task->tk_status = -EAGAIN;
-	if (req->rq_ntrans)
-		rpc_sleep_on(&xprt->resend, task, NULL);
+	if (req == NULL)
+		priority = RPC_PRIORITY_LOW;
+	else if (!req->rq_ntrans)
+		priority = RPC_PRIORITY_NORMAL;
 	else
-		rpc_sleep_on(&xprt->sending, task, NULL);
+		priority = RPC_PRIORITY_HIGH;
+	rpc_sleep_on_priority(&xprt->sending, task, NULL, priority);
 	return 0;
 }
 EXPORT_SYMBOL_GPL(xprt_reserve_xprt);
@@ -239,22 +245,24 @@ static void xprt_clear_locked(struct rpc_xprt *xprt)
  * integrated into the decision of whether a request is allowed to be
  * woken up and given access to the transport.
  */
-int xprt_reserve_xprt_cong(struct rpc_task *task)
+int xprt_reserve_xprt_cong(struct rpc_xprt *xprt, struct rpc_task *task)
 {
-	struct rpc_xprt	*xprt = task->tk_xprt;
 	struct rpc_rqst *req = task->tk_rqstp;
+	int priority;
 
 	if (test_and_set_bit(XPRT_LOCKED, &xprt->state)) {
 		if (task == xprt->snd_task)
 			return 1;
 		goto out_sleep;
 	}
+	if (req == NULL) {
+		xprt->snd_task = task;
+		return 1;
+	}
 	if (__xprt_get_cong(xprt, task)) {
 		xprt->snd_task = task;
-		if (req) {
-			req->rq_bytes_sent = 0;
-			req->rq_ntrans++;
-		}
+		req->rq_bytes_sent = 0;
+		req->rq_ntrans++;
 		return 1;
 	}
 	xprt_clear_locked(xprt);
@@ -262,10 +270,13 @@ out_sleep:
 	dprintk("RPC: %5u failed to lock transport %p\n", task->tk_pid, xprt);
 	task->tk_timeout = 0;
 	task->tk_status = -EAGAIN;
-	if (req && req->rq_ntrans)
-		rpc_sleep_on(&xprt->resend, task, NULL);
+	if (req == NULL)
+		priority = RPC_PRIORITY_LOW;
+	else if (!req->rq_ntrans)
+		priority = RPC_PRIORITY_NORMAL;
 	else
-		rpc_sleep_on(&xprt->sending, task, NULL);
+		priority = RPC_PRIORITY_HIGH;
+	rpc_sleep_on_priority(&xprt->sending, task, NULL, priority);
 	return 0;
 }
 EXPORT_SYMBOL_GPL(xprt_reserve_xprt_cong);
@@ -275,7 +286,7 @@ static inline int xprt_lock_write(struct rpc_xprt *xprt, struct rpc_task *task)
 	int retval;
 
 	spin_lock_bh(&xprt->transport_lock);
-	retval = xprt->ops->reserve_xprt(task);
+	retval = xprt->ops->reserve_xprt(xprt, task);
 	spin_unlock_bh(&xprt->transport_lock);
 	return retval;
 }
@@ -288,12 +299,9 @@ static void __xprt_lock_write_next(struct rpc_xprt *xprt)
 	if (test_and_set_bit(XPRT_LOCKED, &xprt->state))
 		return;
 
-	task = rpc_wake_up_next(&xprt->resend);
-	if (!task) {
-		task = rpc_wake_up_next(&xprt->sending);
-		if (!task)
-			goto out_unlock;
-	}
+	task = rpc_wake_up_next(&xprt->sending);
+	if (task == NULL)
+		goto out_unlock;
 
 	req = task->tk_rqstp;
 	xprt->snd_task = task;
@@ -310,24 +318,25 @@ out_unlock:
 static void __xprt_lock_write_next_cong(struct rpc_xprt *xprt)
 {
 	struct rpc_task *task;
+	struct rpc_rqst *req;
 
 	if (test_and_set_bit(XPRT_LOCKED, &xprt->state))
 		return;
 	if (RPCXPRT_CONGESTED(xprt))
 		goto out_unlock;
-	task = rpc_wake_up_next(&xprt->resend);
-	if (!task) {
-		task = rpc_wake_up_next(&xprt->sending);
-		if (!task)
-			goto out_unlock;
+	task = rpc_wake_up_next(&xprt->sending);
+	if (task == NULL)
+		goto out_unlock;
+
+	req = task->tk_rqstp;
+	if (req == NULL) {
+		xprt->snd_task = task;
+		return;
 	}
 	if (__xprt_get_cong(xprt, task)) {
-		struct rpc_rqst *req = task->tk_rqstp;
 		xprt->snd_task = task;
-		if (req) {
-			req->rq_bytes_sent = 0;
-			req->rq_ntrans++;
-		}
+		req->rq_bytes_sent = 0;
+		req->rq_ntrans++;
 		return;
 	}
 out_unlock:
@@ -852,7 +861,7 @@ int xprt_prepare_transmit(struct rpc_task *task)
 		err = req->rq_reply_bytes_recvd;
 		goto out_unlock;
 	}
-	if (!xprt->ops->reserve_xprt(task))
+	if (!xprt->ops->reserve_xprt(xprt, task))
 		err = -EAGAIN;
 out_unlock:
 	spin_unlock_bh(&xprt->transport_lock);
@@ -928,28 +937,66 @@ void xprt_transmit(struct rpc_task *task)
 	spin_unlock_bh(&xprt->transport_lock);
 }
 
+static struct rpc_rqst *xprt_dynamic_alloc_slot(struct rpc_xprt *xprt, gfp_t gfp_flags)
+{
+	struct rpc_rqst *req = ERR_PTR(-EAGAIN);
+
+	if (!atomic_add_unless(&xprt->num_reqs, 1, xprt->max_reqs))
+		goto out;
+	req = kzalloc(sizeof(struct rpc_rqst), gfp_flags);
+	if (req != NULL)
+		goto out;
+	atomic_dec(&xprt->num_reqs);
+	req = ERR_PTR(-ENOMEM);
+out:
+	return req;
+}
+
+static bool xprt_dynamic_free_slot(struct rpc_xprt *xprt, struct rpc_rqst *req)
+{
+	if (atomic_add_unless(&xprt->num_reqs, -1, xprt->min_reqs)) {
+		kfree(req);
+		return true;
+	}
+	return false;
+}
+
 static void xprt_alloc_slot(struct rpc_task *task)
 {
 	struct rpc_xprt	*xprt = task->tk_xprt;
+	struct rpc_rqst *req;
 
-	task->tk_status = 0;
-	if (task->tk_rqstp)
-		return;
 	if (!list_empty(&xprt->free)) {
-		struct rpc_rqst	*req = list_entry(xprt->free.next, struct rpc_rqst, rq_list);
-		list_del_init(&req->rq_list);
-		task->tk_rqstp = req;
-		xprt_request_init(task, xprt);
-		return;
+		req = list_entry(xprt->free.next, struct rpc_rqst, rq_list);
+		list_del(&req->rq_list);
+		goto out_init_req;
+	}
+	req = xprt_dynamic_alloc_slot(xprt, GFP_NOWAIT);
+	if (!IS_ERR(req))
+		goto out_init_req;
+	switch (PTR_ERR(req)) {
+	case -ENOMEM:
+		rpc_delay(task, HZ >> 2);
+		dprintk("RPC:       dynamic allocation of request slot "
+				"failed! Retrying\n");
+		break;
+	case -EAGAIN:
+		rpc_sleep_on(&xprt->backlog, task, NULL);
+		dprintk("RPC:       waiting for request slot\n");
 	}
-	dprintk("RPC:       waiting for request slot\n");
 	task->tk_status = -EAGAIN;
-	task->tk_timeout = 0;
-	rpc_sleep_on(&xprt->backlog, task, NULL);
+	return;
+out_init_req:
+	task->tk_status = 0;
+	task->tk_rqstp = req;
+	xprt_request_init(task, xprt);
 }
 
 static void xprt_free_slot(struct rpc_xprt *xprt, struct rpc_rqst *req)
 {
+	if (xprt_dynamic_free_slot(xprt, req))
+		return;
+
 	memset(req, 0, sizeof(*req));	/* mark unused */
 
 	spin_lock(&xprt->reserve_lock);
@@ -958,25 +1005,49 @@ static void xprt_free_slot(struct rpc_xprt *xprt, struct rpc_rqst *req)
 	spin_unlock(&xprt->reserve_lock);
 }
 
-struct rpc_xprt *xprt_alloc(struct net *net, int size, int max_req)
+static void xprt_free_all_slots(struct rpc_xprt *xprt)
+{
+	struct rpc_rqst *req;
+	while (!list_empty(&xprt->free)) {
+		req = list_first_entry(&xprt->free, struct rpc_rqst, rq_list);
+		list_del(&req->rq_list);
+		kfree(req);
+	}
+}
+
+struct rpc_xprt *xprt_alloc(struct net *net, size_t size,
+		unsigned int num_prealloc,
+		unsigned int max_alloc)
 {
 	struct rpc_xprt *xprt;
+	struct rpc_rqst *req;
+	int i;
 
 	xprt = kzalloc(size, GFP_KERNEL);
 	if (xprt == NULL)
 		goto out;
-	atomic_set(&xprt->count, 1);
 
-	xprt->max_reqs = max_req;
-	xprt->slot = kcalloc(max_req, sizeof(struct rpc_rqst), GFP_KERNEL);
-	if (xprt->slot == NULL)
+	xprt_init(xprt, net);
+
+	for (i = 0; i < num_prealloc; i++) {
+		req = kzalloc(sizeof(struct rpc_rqst), GFP_KERNEL);
+		if (!req)
+			break;
+		list_add(&req->rq_list, &xprt->free);
+	}
+	if (i < num_prealloc)
 		goto out_free;
+	if (max_alloc > num_prealloc)
+		xprt->max_reqs = max_alloc;
+	else
+		xprt->max_reqs = num_prealloc;
+	xprt->min_reqs = num_prealloc;
+	atomic_set(&xprt->num_reqs, num_prealloc);
 
-	xprt->xprt_net = get_net(net);
 	return xprt;
 
 out_free:
-	kfree(xprt);
+	xprt_free(xprt);
 out:
 	return NULL;
 }
@@ -985,7 +1056,7 @@ EXPORT_SYMBOL_GPL(xprt_alloc);
 void xprt_free(struct rpc_xprt *xprt)
 {
 	put_net(xprt->xprt_net);
-	kfree(xprt->slot);
+	xprt_free_all_slots(xprt);
 	kfree(xprt);
 }
 EXPORT_SYMBOL_GPL(xprt_free);
@@ -1001,10 +1072,24 @@ void xprt_reserve(struct rpc_task *task)
 {
 	struct rpc_xprt	*xprt = task->tk_xprt;
 
-	task->tk_status = -EIO;
+	task->tk_status = 0;
+	if (task->tk_rqstp != NULL)
+		return;
+
+	/* Note: grabbing the xprt_lock_write() here is not strictly needed,
+	 * but ensures that we throttle new slot allocation if the transport
+	 * is congested (e.g. if reconnecting or if we're out of socket
+	 * write buffer space).
+	 */
+	task->tk_timeout = 0;
+	task->tk_status = -EAGAIN;
+	if (!xprt_lock_write(xprt, task))
+		return;
+
 	spin_lock(&xprt->reserve_lock);
 	xprt_alloc_slot(task);
 	spin_unlock(&xprt->reserve_lock);
+	xprt_release_write(xprt, task);
 }
 
 static inline __be32 xprt_alloc_xid(struct rpc_xprt *xprt)
@@ -1021,6 +1106,7 @@ static void xprt_request_init(struct rpc_task *task, struct rpc_xprt *xprt)
 {
 	struct rpc_rqst	*req = task->tk_rqstp;
 
+	INIT_LIST_HEAD(&req->rq_list);
 	req->rq_timeout = task->tk_client->cl_timeout->to_initval;
 	req->rq_task	= task;
 	req->rq_xprt    = xprt;
@@ -1073,6 +1159,34 @@ void xprt_release(struct rpc_task *task)
 		xprt_free_bc_request(req);
 }
 
+static void xprt_init(struct rpc_xprt *xprt, struct net *net)
+{
+	atomic_set(&xprt->count, 1);
+
+	spin_lock_init(&xprt->transport_lock);
+	spin_lock_init(&xprt->reserve_lock);
+
+	INIT_LIST_HEAD(&xprt->free);
+	INIT_LIST_HEAD(&xprt->recv);
+#if defined(CONFIG_SUNRPC_BACKCHANNEL)
+	spin_lock_init(&xprt->bc_pa_lock);
+	INIT_LIST_HEAD(&xprt->bc_pa_list);
+#endif /* CONFIG_SUNRPC_BACKCHANNEL */
+
+	xprt->last_used = jiffies;
+	xprt->cwnd = RPC_INITCWND;
+	xprt->bind_index = 0;
+
+	rpc_init_wait_queue(&xprt->binding, "xprt_binding");
+	rpc_init_wait_queue(&xprt->pending, "xprt_pending");
+	rpc_init_priority_wait_queue(&xprt->sending, "xprt_sending");
+	rpc_init_priority_wait_queue(&xprt->backlog, "xprt_backlog");
+
+	xprt_init_xid(xprt);
+
+	xprt->xprt_net = get_net(net);
+}
+
 /**
  * xprt_create_transport - create an RPC transport
  * @args: rpc transport creation arguments
@@ -1081,7 +1195,6 @@ void xprt_release(struct rpc_task *task)
 struct rpc_xprt *xprt_create_transport(struct xprt_create *args)
 {
 	struct rpc_xprt	*xprt;
-	struct rpc_rqst	*req;
 	struct xprt_class *t;
 
 	spin_lock(&xprt_list_lock);
@@ -1100,46 +1213,17 @@ found:
 	if (IS_ERR(xprt)) {
 		dprintk("RPC:       xprt_create_transport: failed, %ld\n",
 				-PTR_ERR(xprt));
-		return xprt;
+		goto out;
 	}
-	if (test_and_set_bit(XPRT_INITIALIZED, &xprt->state))
-		/* ->setup returned a pre-initialized xprt: */
-		return xprt;
-
-	spin_lock_init(&xprt->transport_lock);
-	spin_lock_init(&xprt->reserve_lock);
-
-	INIT_LIST_HEAD(&xprt->free);
-	INIT_LIST_HEAD(&xprt->recv);
-#if defined(CONFIG_NFS_V4_1)
-	spin_lock_init(&xprt->bc_pa_lock);
-	INIT_LIST_HEAD(&xprt->bc_pa_list);
-#endif /* CONFIG_NFS_V4_1 */
-
 	INIT_WORK(&xprt->task_cleanup, xprt_autoclose);
 	if (xprt_has_timer(xprt))
 		setup_timer(&xprt->timer, xprt_init_autodisconnect,
 			    (unsigned long)xprt);
 	else
 		init_timer(&xprt->timer);
-	xprt->last_used = jiffies;
-	xprt->cwnd = RPC_INITCWND;
-	xprt->bind_index = 0;
-
-	rpc_init_wait_queue(&xprt->binding, "xprt_binding");
-	rpc_init_wait_queue(&xprt->pending, "xprt_pending");
-	rpc_init_wait_queue(&xprt->sending, "xprt_sending");
-	rpc_init_wait_queue(&xprt->resend, "xprt_resend");
-	rpc_init_priority_wait_queue(&xprt->backlog, "xprt_backlog");
-
-	/* initialize free list */
-	for (req = &xprt->slot[xprt->max_reqs-1]; req >= &xprt->slot[0]; req--)
-		list_add(&req->rq_list, &xprt->free);
-
-	xprt_init_xid(xprt);
-
 	dprintk("RPC:       created transport %p with %u slots\n", xprt,
 			xprt->max_reqs);
+out:
 	return xprt;
 }
 
@@ -1157,7 +1241,6 @@ static void xprt_destroy(struct rpc_xprt *xprt)
 	rpc_destroy_wait_queue(&xprt->binding);
 	rpc_destroy_wait_queue(&xprt->pending);
 	rpc_destroy_wait_queue(&xprt->sending);
-	rpc_destroy_wait_queue(&xprt->resend);
 	rpc_destroy_wait_queue(&xprt->backlog);
 	cancel_work_sync(&xprt->task_cleanup);
 	/*

net/sunrpc/xprtrdma/transport.c

@@ -283,6 +283,7 @@ xprt_setup_rdma(struct xprt_create *args)
 	}
 
 	xprt = xprt_alloc(args->net, sizeof(struct rpcrdma_xprt),
+			xprt_rdma_slot_table_entries,
 			xprt_rdma_slot_table_entries);
 	if (xprt == NULL) {
 		dprintk("RPC:       %s: couldn't allocate rpcrdma_xprt\n",
@@ -452,9 +453,8 @@ xprt_rdma_connect(struct rpc_task *task)
 }
 
 static int
-xprt_rdma_reserve_xprt(struct rpc_task *task)
+xprt_rdma_reserve_xprt(struct rpc_xprt *xprt, struct rpc_task *task)
 {
-	struct rpc_xprt *xprt = task->tk_xprt;
 	struct rpcrdma_xprt *r_xprt = rpcx_to_rdmax(xprt);
 	int credits = atomic_read(&r_xprt->rx_buf.rb_credits);
 
@@ -466,7 +466,7 @@ xprt_rdma_reserve_xprt(struct rpc_task *task)
 		BUG_ON(r_xprt->rx_buf.rb_cwndscale <= 0);
 	}
 	xprt->cwnd = credits * r_xprt->rx_buf.rb_cwndscale;
-	return xprt_reserve_xprt_cong(task);
+	return xprt_reserve_xprt_cong(xprt, task);
 }
 
 /*

net/sunrpc/xprtrdma/xprt_rdma.h

@@ -109,7 +109,7 @@ struct rpcrdma_ep {
  */
 
 /* temporary static scatter/gather max */
-#define RPCRDMA_MAX_DATA_SEGS	(8)	/* max scatter/gather */
+#define RPCRDMA_MAX_DATA_SEGS	(64)	/* max scatter/gather */
 #define RPCRDMA_MAX_SEGS 	(RPCRDMA_MAX_DATA_SEGS + 2) /* head+tail = 2 */
 #define MAX_RPCRDMAHDR	(\
 	/* max supported RPC/RDMA header */ \

net/sunrpc/xprtsock.c

@@ -37,7 +37,7 @@
 #include <linux/sunrpc/svcsock.h>
 #include <linux/sunrpc/xprtsock.h>
 #include <linux/file.h>
-#ifdef CONFIG_NFS_V4_1
+#ifdef CONFIG_SUNRPC_BACKCHANNEL
 #include <linux/sunrpc/bc_xprt.h>
 #endif
 
@@ -54,7 +54,8 @@ static void xs_close(struct rpc_xprt *xprt);
  * xprtsock tunables
  */
 unsigned int xprt_udp_slot_table_entries = RPC_DEF_SLOT_TABLE;
-unsigned int xprt_tcp_slot_table_entries = RPC_DEF_SLOT_TABLE;
+unsigned int xprt_tcp_slot_table_entries = RPC_MIN_SLOT_TABLE;
+unsigned int xprt_max_tcp_slot_table_entries = RPC_MAX_SLOT_TABLE;
 
 unsigned int xprt_min_resvport = RPC_DEF_MIN_RESVPORT;
 unsigned int xprt_max_resvport = RPC_DEF_MAX_RESVPORT;
@@ -75,6 +76,7 @@ static unsigned int xs_tcp_fin_timeout __read_mostly = XS_TCP_LINGER_TO;
 
 static unsigned int min_slot_table_size = RPC_MIN_SLOT_TABLE;
 static unsigned int max_slot_table_size = RPC_MAX_SLOT_TABLE;
+static unsigned int max_tcp_slot_table_limit = RPC_MAX_SLOT_TABLE_LIMIT;
 static unsigned int xprt_min_resvport_limit = RPC_MIN_RESVPORT;
 static unsigned int xprt_max_resvport_limit = RPC_MAX_RESVPORT;
 
@@ -103,6 +105,15 @@ static ctl_table xs_tunables_table[] = {
 		.extra1		= &min_slot_table_size,
 		.extra2		= &max_slot_table_size
 	},
+	{
+		.procname	= "tcp_max_slot_table_entries",
+		.data		= &xprt_max_tcp_slot_table_entries,
+		.maxlen		= sizeof(unsigned int),
+		.mode		= 0644,
+		.proc_handler	= proc_dointvec_minmax,
+		.extra1		= &min_slot_table_size,
+		.extra2		= &max_tcp_slot_table_limit
+	},
 	{
 		.procname	= "min_resvport",
 		.data		= &xprt_min_resvport,
@@ -755,6 +766,8 @@ static void xs_tcp_release_xprt(struct rpc_xprt *xprt, struct rpc_task *task)
 	if (task == NULL)
 		goto out_release;
 	req = task->tk_rqstp;
+	if (req == NULL)
+		goto out_release;
 	if (req->rq_bytes_sent == 0)
 		goto out_release;
 	if (req->rq_bytes_sent == req->rq_snd_buf.len)
@@ -1236,7 +1249,7 @@ static inline int xs_tcp_read_reply(struct rpc_xprt *xprt,
 	return 0;
 }
 
-#if defined(CONFIG_NFS_V4_1)
+#if defined(CONFIG_SUNRPC_BACKCHANNEL)
 /*
  * Obtains an rpc_rqst previously allocated and invokes the common
  * tcp read code to read the data.  The result is placed in the callback
@@ -1299,7 +1312,7 @@ static inline int _xs_tcp_read_data(struct rpc_xprt *xprt,
 {
 	return xs_tcp_read_reply(xprt, desc);
 }
-#endif /* CONFIG_NFS_V4_1 */
+#endif /* CONFIG_SUNRPC_BACKCHANNEL */
 
 /*
  * Read data off the transport.  This can be either an RPC_CALL or an
@@ -2489,7 +2502,8 @@ static int xs_init_anyaddr(const int family, struct sockaddr *sap)
 }
 
 static struct rpc_xprt *xs_setup_xprt(struct xprt_create *args,
-				      unsigned int slot_table_size)
+				      unsigned int slot_table_size,
+				      unsigned int max_slot_table_size)
 {
 	struct rpc_xprt *xprt;
 	struct sock_xprt *new;
@@ -2499,7 +2513,8 @@ static struct rpc_xprt *xs_setup_xprt(struct xprt_create *args,
 		return ERR_PTR(-EBADF);
 	}
 
-	xprt = xprt_alloc(args->net, sizeof(*new), slot_table_size);
+	xprt = xprt_alloc(args->net, sizeof(*new), slot_table_size,
+			max_slot_table_size);
 	if (xprt == NULL) {
 		dprintk("RPC:       xs_setup_xprt: couldn't allocate "
 				"rpc_xprt\n");
@@ -2541,7 +2556,8 @@ static struct rpc_xprt *xs_setup_local(struct xprt_create *args)
 	struct rpc_xprt *xprt;
 	struct rpc_xprt *ret;
 
-	xprt = xs_setup_xprt(args, xprt_tcp_slot_table_entries);
+	xprt = xs_setup_xprt(args, xprt_tcp_slot_table_entries,
+			xprt_max_tcp_slot_table_entries);
 	if (IS_ERR(xprt))
 		return xprt;
 	transport = container_of(xprt, struct sock_xprt, xprt);
@@ -2605,7 +2621,8 @@ static struct rpc_xprt *xs_setup_udp(struct xprt_create *args)
 	struct sock_xprt *transport;
 	struct rpc_xprt *ret;
 
-	xprt = xs_setup_xprt(args, xprt_udp_slot_table_entries);
+	xprt = xs_setup_xprt(args, xprt_udp_slot_table_entries,
+			xprt_udp_slot_table_entries);
 	if (IS_ERR(xprt))
 		return xprt;
 	transport = container_of(xprt, struct sock_xprt, xprt);
@@ -2681,7 +2698,8 @@ static struct rpc_xprt *xs_setup_tcp(struct xprt_create *args)
 	struct sock_xprt *transport;
 	struct rpc_xprt *ret;
 
-	xprt = xs_setup_xprt(args, xprt_tcp_slot_table_entries);
+	xprt = xs_setup_xprt(args, xprt_tcp_slot_table_entries,
+			xprt_max_tcp_slot_table_entries);
 	if (IS_ERR(xprt))
 		return xprt;
 	transport = container_of(xprt, struct sock_xprt, xprt);
@@ -2760,7 +2778,8 @@ static struct rpc_xprt *xs_setup_bc_tcp(struct xprt_create *args)
 		 */
 		 return args->bc_xprt->xpt_bc_xprt;
 	}
-	xprt = xs_setup_xprt(args, xprt_tcp_slot_table_entries);
+	xprt = xs_setup_xprt(args, xprt_tcp_slot_table_entries,
+			xprt_tcp_slot_table_entries);
 	if (IS_ERR(xprt))
 		return xprt;
 	transport = container_of(xprt, struct sock_xprt, xprt);
@@ -2947,8 +2966,26 @@ static struct kernel_param_ops param_ops_slot_table_size = {
 #define param_check_slot_table_size(name, p) \
 	__param_check(name, p, unsigned int);
 
+static int param_set_max_slot_table_size(const char *val,
+				     const struct kernel_param *kp)
+{
+	return param_set_uint_minmax(val, kp,
+			RPC_MIN_SLOT_TABLE,
+			RPC_MAX_SLOT_TABLE_LIMIT);
+}
+
+static struct kernel_param_ops param_ops_max_slot_table_size = {
+	.set = param_set_max_slot_table_size,
+	.get = param_get_uint,
+};
+
+#define param_check_max_slot_table_size(name, p) \
+	__param_check(name, p, unsigned int);
+
 module_param_named(tcp_slot_table_entries, xprt_tcp_slot_table_entries,
 		   slot_table_size, 0644);
+module_param_named(tcp_max_slot_table_entries, xprt_max_tcp_slot_table_entries,
+		   max_slot_table_size, 0644);
 module_param_named(udp_slot_table_entries, xprt_udp_slot_table_entries,
 		   slot_table_size, 0644);