11 files changed, 160 insertions, 223 deletions

diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c
index 8c10b05..621c517 100644
--- a/fs/nfs/direct.c
+++ b/fs/nfs/direct.c
@@ -86,10 +86,10 @@ struct nfs_direct_req {
 	struct nfs_direct_mirror mirrors[NFS_PAGEIO_DESCRIPTOR_MIRROR_MAX];
 	int			mirror_count;
 
+	loff_t			io_start;	/* Start offset for I/O */
 	ssize_t			count,		/* bytes actually processed */
 				max_count,	/* max expected count */
 				bytes_left,	/* bytes left to be sent */
-				io_start,	/* start of IO */
 				error;		/* any reported error */
 	struct completion	completion;	/* wait for i/o completion */
 
diff --git a/fs/nfs/fscache-index.c b/fs/nfs/fscache-index.c
index 0ee4b93..1c5d8d3 100644
--- a/fs/nfs/fscache-index.c
+++ b/fs/nfs/fscache-index.c
@@ -50,59 +50,6 @@ void nfs_fscache_unregister(void)
 }
 
 /*
- * Layout of the key for an NFS server cache object.
- */
-struct nfs_server_key {
-	uint16_t	nfsversion;		/* NFS protocol version */
-	uint16_t	family;			/* address family */
-	uint16_t	port;			/* IP port */
-	union {
-		struct in_addr	ipv4_addr;	/* IPv4 address */
-		struct in6_addr ipv6_addr;	/* IPv6 address */
-	} addr[0];
-};
-
-/*
- * Generate a key to describe a server in the main NFS index
- * - We return the length of the key, or 0 if we can't generate one
- */
-static uint16_t nfs_server_get_key(const void *cookie_netfs_data,
-				   void *buffer, uint16_t bufmax)
-{
-	const struct nfs_client *clp = cookie_netfs_data;
-	const struct sockaddr_in6 *sin6 = (struct sockaddr_in6 *) &clp->cl_addr;
-	const struct sockaddr_in *sin = (struct sockaddr_in *) &clp->cl_addr;
-	struct nfs_server_key *key = buffer;
-	uint16_t len = sizeof(struct nfs_server_key);
-
-	memset(key, 0, len);
-	key->nfsversion = clp->rpc_ops->version;
-	key->family = clp->cl_addr.ss_family;
-
-	switch (clp->cl_addr.ss_family) {
-	case AF_INET:
-		key->port = sin->sin_port;
-		key->addr[0].ipv4_addr = sin->sin_addr;
-		len += sizeof(key->addr[0].ipv4_addr);
-		break;
-
-	case AF_INET6:
-		key->port = sin6->sin6_port;
-		key->addr[0].ipv6_addr = sin6->sin6_addr;
-		len += sizeof(key->addr[0].ipv6_addr);
-		break;
-
-	default:
-		printk(KERN_WARNING "NFS: Unknown network family '%d'\n",
-		       clp->cl_addr.ss_family);
-		len = 0;
-		break;
-	}
-
-	return len;
-}
-
-/*
  * Define the server object for FS-Cache.  This is used to describe a server
  * object to fscache_acquire_cookie().  It is keyed by the NFS protocol and
  * server address parameters.
@@ -110,33 +57,9 @@ static uint16_t nfs_server_get_key(const void *cookie_netfs_data,
 const struct fscache_cookie_def nfs_fscache_server_index_def = {
 	.name		= "NFS.server",
 	.type 		= FSCACHE_COOKIE_TYPE_INDEX,
-	.get_key	= nfs_server_get_key,
 };
 
 /*
- * Generate a key to describe a superblock key in the main NFS index
- */
-static uint16_t nfs_super_get_key(const void *cookie_netfs_data,
-				  void *buffer, uint16_t bufmax)
-{
-	const struct nfs_fscache_key *key;
-	const struct nfs_server *nfss = cookie_netfs_data;
-	uint16_t len;
-
-	key = nfss->fscache_key;
-	len = sizeof(key->key) + key->key.uniq_len;
-	if (len > bufmax) {
-		len = 0;
-	} else {
-		memcpy(buffer, &key->key, sizeof(key->key));
-		memcpy(buffer + sizeof(key->key),
-		       key->key.uniquifier, key->key.uniq_len);
-	}
-
-	return len;
-}
-
-/*
  * Define the superblock object for FS-Cache.  This is used to describe a
  * superblock object to fscache_acquire_cookie().  It is keyed by all the NFS
  * parameters that might cause a separate superblock.
@@ -144,84 +67,9 @@ static uint16_t nfs_super_get_key(const void *cookie_netfs_data,
 const struct fscache_cookie_def nfs_fscache_super_index_def = {
 	.name		= "NFS.super",
 	.type 		= FSCACHE_COOKIE_TYPE_INDEX,
-	.get_key	= nfs_super_get_key,
 };
 
 /*
- * Definition of the auxiliary data attached to NFS inode storage objects
- * within the cache.
- *
- * The contents of this struct are recorded in the on-disk local cache in the
- * auxiliary data attached to the data storage object backing an inode.  This
- * permits coherency to be managed when a new inode binds to an already extant
- * cache object.
- */
-struct nfs_fscache_inode_auxdata {
-	struct timespec	mtime;
-	struct timespec	ctime;
-	loff_t		size;
-	u64		change_attr;
-};
-
-/*
- * Generate a key to describe an NFS inode in an NFS server's index
- */
-static uint16_t nfs_fscache_inode_get_key(const void *cookie_netfs_data,
-					  void *buffer, uint16_t bufmax)
-{
-	const struct nfs_inode *nfsi = cookie_netfs_data;
-	uint16_t nsize;
-
-	/* use the inode's NFS filehandle as the key */
-	nsize = nfsi->fh.size;
-	memcpy(buffer, nfsi->fh.data, nsize);
-	return nsize;
-}
-
-/*
- * Get certain file attributes from the netfs data
- * - This function can be absent for an index
- * - Not permitted to return an error
- * - The netfs data from the cookie being used as the source is presented
- */
-static void nfs_fscache_inode_get_attr(const void *cookie_netfs_data,
-				       uint64_t *size)
-{
-	const struct nfs_inode *nfsi = cookie_netfs_data;
-
-	*size = nfsi->vfs_inode.i_size;
-}
-
-/*
- * Get the auxiliary data from netfs data
- * - This function can be absent if the index carries no state data
- * - Should store the auxiliary data in the buffer
- * - Should return the amount of amount stored
- * - Not permitted to return an error
- * - The netfs data from the cookie being used as the source is presented
- */
-static uint16_t nfs_fscache_inode_get_aux(const void *cookie_netfs_data,
-					  void *buffer, uint16_t bufmax)
-{
-	struct nfs_fscache_inode_auxdata auxdata;
-	const struct nfs_inode *nfsi = cookie_netfs_data;
-
-	memset(&auxdata, 0, sizeof(auxdata));
-	auxdata.size = nfsi->vfs_inode.i_size;
-	auxdata.mtime = nfsi->vfs_inode.i_mtime;
-	auxdata.ctime = nfsi->vfs_inode.i_ctime;
-
-	if (NFS_SERVER(&nfsi->vfs_inode)->nfs_client->rpc_ops->version == 4)
-		auxdata.change_attr = inode_peek_iversion_raw(&nfsi->vfs_inode);
-
-	if (bufmax > sizeof(auxdata))
-		bufmax = sizeof(auxdata);
-
-	memcpy(buffer, &auxdata, bufmax);
-	return bufmax;
-}
-
-/*
  * Consult the netfs about the state of an object
  * - This function can be absent if the index carries no state data
  * - The netfs data from the cookie being used as the target is
@@ -230,7 +78,8 @@ static uint16_t nfs_fscache_inode_get_aux(const void *cookie_netfs_data,
 static
 enum fscache_checkaux nfs_fscache_inode_check_aux(void *cookie_netfs_data,
 						  const void *data,
-						  uint16_t datalen)
+						  uint16_t datalen,
+						  loff_t object_size)
 {
 	struct nfs_fscache_inode_auxdata auxdata;
 	struct nfs_inode *nfsi = cookie_netfs_data;
@@ -239,7 +88,6 @@ enum fscache_checkaux nfs_fscache_inode_check_aux(void *cookie_netfs_data,
 		return FSCACHE_CHECKAUX_OBSOLETE;
 
 	memset(&auxdata, 0, sizeof(auxdata));
-	auxdata.size = nfsi->vfs_inode.i_size;
 	auxdata.mtime = nfsi->vfs_inode.i_mtime;
 	auxdata.ctime = nfsi->vfs_inode.i_ctime;
 
@@ -288,9 +136,6 @@ static void nfs_fh_put_context(void *cookie_netfs_data, void *context)
 const struct fscache_cookie_def nfs_fscache_inode_object_def = {
 	.name		= "NFS.fh",
 	.type		= FSCACHE_COOKIE_TYPE_DATAFILE,
-	.get_key	= nfs_fscache_inode_get_key,
-	.get_attr	= nfs_fscache_inode_get_attr,
-	.get_aux	= nfs_fscache_inode_get_aux,
 	.check_aux	= nfs_fscache_inode_check_aux,
 	.get_context	= nfs_fh_get_context,
 	.put_context	= nfs_fh_put_context,
diff --git a/fs/nfs/fscache.c b/fs/nfs/fscache.c
index d63bea8..b55fc79 100644
--- a/fs/nfs/fscache.c
+++ b/fs/nfs/fscache.c
@@ -18,6 +18,7 @@
 #include <linux/in6.h>
 #include <linux/seq_file.h>
 #include <linux/slab.h>
+#include <linux/iversion.h>
 
 #include "internal.h"
 #include "iostat.h"
@@ -29,6 +30,21 @@ static struct rb_root nfs_fscache_keys = RB_ROOT;
 static DEFINE_SPINLOCK(nfs_fscache_keys_lock);
 
 /*
+ * Layout of the key for an NFS server cache object.
+ */
+struct nfs_server_key {
+	struct {
+		uint16_t	nfsversion;		/* NFS protocol version */
+		uint16_t	family;			/* address family */
+		__be16		port;			/* IP port */
+	} hdr;
+	union {
+		struct in_addr	ipv4_addr;	/* IPv4 address */
+		struct in6_addr ipv6_addr;	/* IPv6 address */
+	};
+} __packed;
+
+/*
  * Get the per-client index cookie for an NFS client if the appropriate mount
  * flag was set
  * - We always try and get an index cookie for the client, but get filehandle
@@ -36,10 +52,41 @@ static DEFINE_SPINLOCK(nfs_fscache_keys_lock);
  */
 void nfs_fscache_get_client_cookie(struct nfs_client *clp)
 {
+	const struct sockaddr_in6 *sin6 = (struct sockaddr_in6 *) &clp->cl_addr;
+	const struct sockaddr_in *sin = (struct sockaddr_in *) &clp->cl_addr;
+	struct nfs_server_key key;
+	uint16_t len = sizeof(key.hdr);
+
+	memset(&key, 0, sizeof(key));
+	key.hdr.nfsversion = clp->rpc_ops->version;
+	key.hdr.family = clp->cl_addr.ss_family;
+
+	switch (clp->cl_addr.ss_family) {
+	case AF_INET:
+		key.hdr.port = sin->sin_port;
+		key.ipv4_addr = sin->sin_addr;
+		len += sizeof(key.ipv4_addr);
+		break;
+
+	case AF_INET6:
+		key.hdr.port = sin6->sin6_port;
+		key.ipv6_addr = sin6->sin6_addr;
+		len += sizeof(key.ipv6_addr);
+		break;
+
+	default:
+		printk(KERN_WARNING "NFS: Unknown network family '%d'\n",
+		       clp->cl_addr.ss_family);
+		clp->fscache = NULL;
+		return;
+	}
+
 	/* create a cache index for looking up filehandles */
 	clp->fscache = fscache_acquire_cookie(nfs_fscache_netfs.primary_index,
 					      &nfs_fscache_server_index_def,
-					      clp, true);
+					      &key, len,
+					      NULL, 0,
+					      clp, 0, true);
 	dfprintk(FSCACHE, "NFS: get client cookie (0x%p/0x%p)\n",
 		 clp, clp->fscache);
 }
@@ -52,7 +99,7 @@ void nfs_fscache_release_client_cookie(struct nfs_client *clp)
 	dfprintk(FSCACHE, "NFS: releasing client cookie (0x%p/0x%p)\n",
 		 clp, clp->fscache);
 
-	fscache_relinquish_cookie(clp->fscache, 0);
+	fscache_relinquish_cookie(clp->fscache, NULL, false);
 	clp->fscache = NULL;
 }
 
@@ -139,7 +186,9 @@ void nfs_fscache_get_super_cookie(struct super_block *sb, const char *uniq, int
 	/* create a cache index for looking up filehandles */
 	nfss->fscache = fscache_acquire_cookie(nfss->nfs_client->fscache,
 					       &nfs_fscache_super_index_def,
-					       nfss, true);
+					       key, sizeof(*key) + ulen,
+					       NULL, 0,
+					       nfss, 0, true);
 	dfprintk(FSCACHE, "NFS: get superblock cookie (0x%p/0x%p)\n",
 		 nfss, nfss->fscache);
 	return;
@@ -163,7 +212,7 @@ void nfs_fscache_release_super_cookie(struct super_block *sb)
 	dfprintk(FSCACHE, "NFS: releasing superblock cookie (0x%p/0x%p)\n",
 		 nfss, nfss->fscache);
 
-	fscache_relinquish_cookie(nfss->fscache, 0);
+	fscache_relinquish_cookie(nfss->fscache, NULL, false);
 	nfss->fscache = NULL;
 
 	if (nfss->fscache_key) {
@@ -180,14 +229,25 @@ void nfs_fscache_release_super_cookie(struct super_block *sb)
  */
 void nfs_fscache_init_inode(struct inode *inode)
 {
+	struct nfs_fscache_inode_auxdata auxdata;
 	struct nfs_inode *nfsi = NFS_I(inode);
 
 	nfsi->fscache = NULL;
 	if (!S_ISREG(inode->i_mode))
 		return;
+
+	memset(&auxdata, 0, sizeof(auxdata));
+	auxdata.mtime = nfsi->vfs_inode.i_mtime;
+	auxdata.ctime = nfsi->vfs_inode.i_ctime;
+
+	if (NFS_SERVER(&nfsi->vfs_inode)->nfs_client->rpc_ops->version == 4)
+		auxdata.change_attr = inode_peek_iversion_raw(&nfsi->vfs_inode);
+
 	nfsi->fscache = fscache_acquire_cookie(NFS_SB(inode->i_sb)->fscache,
 					       &nfs_fscache_inode_object_def,
-					       nfsi, false);
+					       nfsi->fh.data, nfsi->fh.size,
+					       &auxdata, sizeof(auxdata),
+					       nfsi, nfsi->vfs_inode.i_size, false);
 }
 
 /*
@@ -195,12 +255,16 @@ void nfs_fscache_init_inode(struct inode *inode)
  */
 void nfs_fscache_clear_inode(struct inode *inode)
 {
+	struct nfs_fscache_inode_auxdata auxdata;
 	struct nfs_inode *nfsi = NFS_I(inode);
 	struct fscache_cookie *cookie = nfs_i_fscache(inode);
 
 	dfprintk(FSCACHE, "NFS: clear cookie (0x%p/0x%p)\n", nfsi, cookie);
 
-	fscache_relinquish_cookie(cookie, false);
+	memset(&auxdata, 0, sizeof(auxdata));
+	auxdata.mtime = nfsi->vfs_inode.i_mtime;
+	auxdata.ctime = nfsi->vfs_inode.i_ctime;
+	fscache_relinquish_cookie(cookie, &auxdata, false);
 	nfsi->fscache = NULL;
 }
 
@@ -232,20 +296,26 @@ static bool nfs_fscache_can_enable(void *data)
  */
 void nfs_fscache_open_file(struct inode *inode, struct file *filp)
 {
+	struct nfs_fscache_inode_auxdata auxdata;
 	struct nfs_inode *nfsi = NFS_I(inode);
 	struct fscache_cookie *cookie = nfs_i_fscache(inode);
 
 	if (!fscache_cookie_valid(cookie))
 		return;
 
+	memset(&auxdata, 0, sizeof(auxdata));
+	auxdata.mtime = nfsi->vfs_inode.i_mtime;
+	auxdata.ctime = nfsi->vfs_inode.i_ctime;
+
 	if (inode_is_open_for_write(inode)) {
 		dfprintk(FSCACHE, "NFS: nfsi 0x%p disabling cache\n", nfsi);
 		clear_bit(NFS_INO_FSCACHE, &nfsi->flags);
-		fscache_disable_cookie(cookie, true);
+		fscache_disable_cookie(cookie, &auxdata, true);
 		fscache_uncache_all_inode_pages(cookie, inode);
 	} else {
 		dfprintk(FSCACHE, "NFS: nfsi 0x%p enabling cache\n", nfsi);
-		fscache_enable_cookie(cookie, nfs_fscache_can_enable, inode);
+		fscache_enable_cookie(cookie, &auxdata, nfsi->vfs_inode.i_size,
+				      nfs_fscache_can_enable, inode);
 		if (fscache_cookie_enabled(cookie))
 			set_bit(NFS_INO_FSCACHE, &NFS_I(inode)->flags);
 	}
@@ -422,7 +492,8 @@ void __nfs_readpage_to_fscache(struct inode *inode, struct page *page, int sync)
 		 "NFS: readpage_to_fscache(fsc:%p/p:%p(i:%lx f:%lx)/%d)\n",
 		 nfs_i_fscache(inode), page, page->index, page->flags, sync);
 
-	ret = fscache_write_page(nfs_i_fscache(inode), page, GFP_KERNEL);
+	ret = fscache_write_page(nfs_i_fscache(inode), page,
+				 inode->i_size, GFP_KERNEL);
 	dfprintk(FSCACHE,
 		 "NFS:     readpage_to_fscache: p:%p(i:%lu f:%lx) ret %d\n",
 		 page, page->index, page->flags, ret);
diff --git a/fs/nfs/fscache.h b/fs/nfs/fscache.h
index d7fe3e7..161ba2e 100644
--- a/fs/nfs/fscache.h
+++ b/fs/nfs/fscache.h
@@ -57,6 +57,21 @@ struct nfs_fscache_key {
 };
 
 /*
+ * Definition of the auxiliary data attached to NFS inode storage objects
+ * within the cache.
+ *
+ * The contents of this struct are recorded in the on-disk local cache in the
+ * auxiliary data attached to the data storage object backing an inode.  This
+ * permits coherency to be managed when a new inode binds to an already extant
+ * cache object.
+ */
+struct nfs_fscache_inode_auxdata {
+	struct timespec	mtime;
+	struct timespec	ctime;
+	u64		change_attr;
+};
+
+/*
  * fscache-index.c
  */
 extern struct fscache_netfs nfs_fscache_netfs;
diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c
index ba73eda..bd15d0b 100644
--- a/fs/nfs/inode.c
+++ b/fs/nfs/inode.c
@@ -85,11 +85,6 @@ int nfs_wait_bit_killable(struct wait_bit_key *key, int mode)
 }
 EXPORT_SYMBOL_GPL(nfs_wait_bit_killable);
 
-int nfs_wait_atomic_killable(atomic_t *p, unsigned int mode)
-{
-	return nfs_wait_killable(mode);
-}
-
 /**
  * nfs_compat_user_ino64 - returns the user-visible inode number
  * @fileid: 64-bit fileid
diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c
index aa550fb..9b73920 100644
--- a/fs/nfs/nfs4xdr.c
+++ b/fs/nfs/nfs4xdr.c
@@ -52,7 +52,6 @@
 #include <linux/nfs.h>
 #include <linux/nfs4.h>
 #include <linux/nfs_fs.h>
-#include <linux/fs_struct.h>
 
 #include "nfs4_fs.h"
 #include "internal.h"
diff --git a/fs/nfs/pagelist.c b/fs/nfs/pagelist.c
index 18a7626..67d19cd 100644
--- a/fs/nfs/pagelist.c
+++ b/fs/nfs/pagelist.c
@@ -98,8 +98,8 @@ nfs_page_free(struct nfs_page *p)
 int
 nfs_iocounter_wait(struct nfs_lock_context *l_ctx)
 {
-	return wait_on_atomic_t(&l_ctx->io_count, nfs_wait_atomic_killable,
-			TASK_KILLABLE);
+	return wait_var_event_killable(&l_ctx->io_count,
+				       !atomic_read(&l_ctx->io_count));
 }
 
 /**
@@ -395,7 +395,7 @@ static void nfs_clear_request(struct nfs_page *req)
 	}
 	if (l_ctx != NULL) {
 		if (atomic_dec_and_test(&l_ctx->io_count)) {
-			wake_up_atomic_t(&l_ctx->io_count);
+			wake_up_var(&l_ctx->io_count);
 			if (test_bit(NFS_CONTEXT_UNLOCK, &ctx->flags))
 				rpc_wake_up(&NFS_SERVER(d_inode(ctx->dentry))->uoc_rpcwaitq);
 		}
diff --git a/fs/nfs/pnfs.c b/fs/nfs/pnfs.c
index c13e826..ee723aa 100644
--- a/fs/nfs/pnfs.c
+++ b/fs/nfs/pnfs.c
@@ -292,8 +292,11 @@ pnfs_detach_layout_hdr(struct pnfs_layout_hdr *lo)
 void
 pnfs_put_layout_hdr(struct pnfs_layout_hdr *lo)
 {
-	struct inode *inode = lo->plh_inode;
+	struct inode *inode;
 
+	if (!lo)
+		return;
+	inode = lo->plh_inode;
 	pnfs_layoutreturn_before_put_layout_hdr(lo);
 
 	if (refcount_dec_and_lock(&lo->plh_refcount, &inode->i_lock)) {
@@ -1241,10 +1244,12 @@ retry:
 	spin_lock(&ino->i_lock);
 	lo = nfsi->layout;
 	if (!lo || !pnfs_layout_is_valid(lo) ||
-	    test_bit(NFS_LAYOUT_BULK_RECALL, &lo->plh_flags))
+	    test_bit(NFS_LAYOUT_BULK_RECALL, &lo->plh_flags)) {
+		lo = NULL;
 		goto out_noroc;
+	}
+	pnfs_get_layout_hdr(lo);
 	if (test_bit(NFS_LAYOUT_RETURN_LOCK, &lo->plh_flags)) {
-		pnfs_get_layout_hdr(lo);
 		spin_unlock(&ino->i_lock);
 		wait_on_bit(&lo->plh_flags, NFS_LAYOUT_RETURN,
 				TASK_UNINTERRUPTIBLE);
@@ -1312,10 +1317,12 @@ out_noroc:
 		struct pnfs_layoutdriver_type *ld = NFS_SERVER(ino)->pnfs_curr_ld;
 		if (ld->prepare_layoutreturn)
 			ld->prepare_layoutreturn(args);
+		pnfs_put_layout_hdr(lo);
 		return true;
 	}
 	if (layoutreturn)
 		pnfs_send_layoutreturn(lo, &stateid, iomode, true);
+	pnfs_put_layout_hdr(lo);
 	return false;
 }
 
diff --git a/fs/nfs/pnfs_nfs.c b/fs/nfs/pnfs_nfs.c
index 03aaa60..32ba2d4 100644
--- a/fs/nfs/pnfs_nfs.c
+++ b/fs/nfs/pnfs_nfs.c
@@ -245,7 +245,7 @@ pnfs_generic_commit_cancel_empty_pagelist(struct list_head *pages,
 {
 	if (list_empty(pages)) {
 		if (atomic_dec_and_test(&cinfo->mds->rpcs_out))
-			wake_up_atomic_t(&cinfo->mds->rpcs_out);
+			wake_up_var(&cinfo->mds->rpcs_out);
 		/* don't call nfs_commitdata_release - it tries to put
 		 * the open_context which is not acquired until nfs_init_commit
 		 * which has not been called on @data */
diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index 29bacdc..5e470e2 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -2631,6 +2631,8 @@ struct dentry *nfs_fs_mount_common(struct nfs_server *server,
 		/* initial superblock/root creation */
 		mount_info->fill_super(s, mount_info);
 		nfs_get_cache_cookie(s, mount_info->parsed, mount_info->cloned);
+		if (!(server->flags & NFS_MOUNT_UNSHARED))
+			s->s_iflags |= SB_I_MULTIROOT;
 	}
 
 	mntroot = nfs_get_root(s, mount_info->mntfh, dev_name);
diff --git a/fs/nfs/write.c b/fs/nfs/write.c
index 541471a..0193053 100644
--- a/fs/nfs/write.c
+++ b/fs/nfs/write.c
@@ -1624,8 +1624,8 @@ static void nfs_writeback_result(struct rpc_task *task,
 
 static int wait_on_commit(struct nfs_mds_commit_info *cinfo)
 {
-	return wait_on_atomic_t(&cinfo->rpcs_out,
-			nfs_wait_atomic_killable, TASK_KILLABLE);
+	return wait_var_event_killable(&cinfo->rpcs_out,
+				       !atomic_read(&cinfo->rpcs_out));
 }
 
 static void nfs_commit_begin(struct nfs_mds_commit_info *cinfo)
@@ -1636,7 +1636,7 @@ static void nfs_commit_begin(struct nfs_mds_commit_info *cinfo)
 static void nfs_commit_end(struct nfs_mds_commit_info *cinfo)
 {
 	if (atomic_dec_and_test(&cinfo->rpcs_out))
-		wake_up_atomic_t(&cinfo->rpcs_out);
+		wake_up_var(&cinfo->rpcs_out);
 }
 
 void nfs_commitdata_release(struct nfs_commit_data *data)
@@ -1880,40 +1880,43 @@ int nfs_generic_commit_list(struct inode *inode, struct list_head *head,
 	return status;
 }
 
-int nfs_commit_inode(struct inode *inode, int how)
+static int __nfs_commit_inode(struct inode *inode, int how,
+		struct writeback_control *wbc)
 {
 	LIST_HEAD(head);
 	struct nfs_commit_info cinfo;
 	int may_wait = how & FLUSH_SYNC;
-	int error = 0;
-	int res;
+	int ret, nscan;
 
 	nfs_init_cinfo_from_inode(&cinfo, inode);
 	nfs_commit_begin(cinfo.mds);
-	res = nfs_scan_commit(inode, &head, &cinfo);
-	if (res)
-		error = nfs_generic_commit_list(inode, &head, how, &cinfo);
+	for (;;) {
+		ret = nscan = nfs_scan_commit(inode, &head, &cinfo);
+		if (ret <= 0)
+			break;
+		ret = nfs_generic_commit_list(inode, &head, how, &cinfo);
+		if (ret < 0)
+			break;
+		ret = 0;
+		if (wbc && wbc->sync_mode == WB_SYNC_NONE) {
+			if (nscan < wbc->nr_to_write)
+				wbc->nr_to_write -= nscan;
+			else
+				wbc->nr_to_write = 0;
+		}
+		if (nscan < INT_MAX)
+			break;
+		cond_resched();
+	}
 	nfs_commit_end(cinfo.mds);
-	if (res == 0)
-		return res;
-	if (error < 0)
-		goto out_error;
-	if (!may_wait)
-		goto out_mark_dirty;
-	error = wait_on_commit(cinfo.mds);
-	if (error < 0)
-		return error;
-	return res;
-out_error:
-	res = error;
-	/* Note: If we exit without ensuring that the commit is complete,
-	 * we must mark the inode as dirty. Otherwise, future calls to
-	 * sync_inode() with the WB_SYNC_ALL flag set will fail to ensure
-	 * that the data is on the disk.
-	 */
-out_mark_dirty:
-	__mark_inode_dirty(inode, I_DIRTY_DATASYNC);
-	return res;
+	if (ret || !may_wait)
+		return ret;
+	return wait_on_commit(cinfo.mds);
+}
+
+int nfs_commit_inode(struct inode *inode, int how)
+{
+	return __nfs_commit_inode(inode, how, NULL);
 }
 EXPORT_SYMBOL_GPL(nfs_commit_inode);
 
@@ -1923,11 +1926,11 @@ int nfs_write_inode(struct inode *inode, struct writeback_control *wbc)
 	int flags = FLUSH_SYNC;
 	int ret = 0;
 
-	/* no commits means nothing needs to be done */
-	if (!atomic_long_read(&nfsi->commit_info.ncommit))
-		return ret;
-
 	if (wbc->sync_mode == WB_SYNC_NONE) {
+		/* no commits means nothing needs to be done */
+		if (!atomic_long_read(&nfsi->commit_info.ncommit))
+			goto check_requests_outstanding;
+
 		/* Don't commit yet if this is a non-blocking flush and there
 		 * are a lot of outstanding writes for this mapping.
 		 */
@@ -1938,16 +1941,16 @@ int nfs_write_inode(struct inode *inode, struct writeback_control *wbc)
 		flags = 0;
 	}
 
-	ret = nfs_commit_inode(inode, flags);
-	if (ret >= 0) {
-		if (wbc->sync_mode == WB_SYNC_NONE) {
-			if (ret < wbc->nr_to_write)
-				wbc->nr_to_write -= ret;
-			else
-				wbc->nr_to_write = 0;
-		}
-		return 0;
-	}
+	ret = __nfs_commit_inode(inode, flags, wbc);
+	if (!ret) {
+		if (flags & FLUSH_SYNC)
+			return 0;
+	} else if (atomic_long_read(&nfsi->commit_info.ncommit))
+		goto out_mark_dirty;
+
+check_requests_outstanding:
+	if (!atomic_read(&nfsi->commit_info.rpcs_out))
+		return ret;
 out_mark_dirty:
 	__mark_inode_dirty(inode, I_DIRTY_DATASYNC);
 	return ret;