Merge tag 'irqchip-core-3.18' of git://git.infradead.org/users/jcooper/linux into irq/core

irqchip core changes for v3.18

 - renesas: suspend to RAM, runtime PM, cleanups and DT binding docs
 - keystone: add new driver
 - hip04: add Hisilicon HiP04 driver (without touching irq-gic.c)
 - gic: Use defines instead of magic number, preserve v2 bybass bits
 - handle_domain_irq: common low level interrupt entry handler

34 files changed, 1650 insertions, 1068 deletions

diff --git a/fs/nfs/blocklayout/blocklayout.c b/fs/nfs/blocklayout/blocklayout.c
index 9b431f44fad9..cbb1797149d5 100644
--- a/fs/nfs/blocklayout/blocklayout.c
+++ b/fs/nfs/blocklayout/blocklayout.c
@@ -210,8 +210,7 @@ static void bl_end_io_read(struct bio *bio, int err)
 			SetPageUptodate(bvec->bv_page);
 
 	if (err) {
-		struct nfs_pgio_data *rdata = par->data;
-		struct nfs_pgio_header *header = rdata->header;
+		struct nfs_pgio_header *header = par->data;
 
 		if (!header->pnfs_error)
 			header->pnfs_error = -EIO;
@@ -224,43 +223,44 @@ static void bl_end_io_read(struct bio *bio, int err)
 static void bl_read_cleanup(struct work_struct *work)
 {
 	struct rpc_task *task;
-	struct nfs_pgio_data *rdata;
+	struct nfs_pgio_header *hdr;
 	dprintk("%s enter\n", __func__);
 	task = container_of(work, struct rpc_task, u.tk_work);
-	rdata = container_of(task, struct nfs_pgio_data, task);
-	pnfs_ld_read_done(rdata);
+	hdr = container_of(task, struct nfs_pgio_header, task);
+	pnfs_ld_read_done(hdr);
 }
 
 static void
 bl_end_par_io_read(void *data, int unused)
 {
-	struct nfs_pgio_data *rdata = data;
+	struct nfs_pgio_header *hdr = data;
 
-	rdata->task.tk_status = rdata->header->pnfs_error;
-	INIT_WORK(&rdata->task.u.tk_work, bl_read_cleanup);
-	schedule_work(&rdata->task.u.tk_work);
+	hdr->task.tk_status = hdr->pnfs_error;
+	INIT_WORK(&hdr->task.u.tk_work, bl_read_cleanup);
+	schedule_work(&hdr->task.u.tk_work);
 }
 
 static enum pnfs_try_status
-bl_read_pagelist(struct nfs_pgio_data *rdata)
+bl_read_pagelist(struct nfs_pgio_header *hdr)
 {
-	struct nfs_pgio_header *header = rdata->header;
+	struct nfs_pgio_header *header = hdr;
 	int i, hole;
 	struct bio *bio = NULL;
 	struct pnfs_block_extent *be = NULL, *cow_read = NULL;
 	sector_t isect, extent_length = 0;
 	struct parallel_io *par;
-	loff_t f_offset = rdata->args.offset;
-	size_t bytes_left = rdata->args.count;
+	loff_t f_offset = hdr->args.offset;
+	size_t bytes_left = hdr->args.count;
 	unsigned int pg_offset, pg_len;
-	struct page **pages = rdata->args.pages;
-	int pg_index = rdata->args.pgbase >> PAGE_CACHE_SHIFT;
+	struct page **pages = hdr->args.pages;
+	int pg_index = hdr->args.pgbase >> PAGE_CACHE_SHIFT;
 	const bool is_dio = (header->dreq != NULL);
 
 	dprintk("%s enter nr_pages %u offset %lld count %u\n", __func__,
-	       rdata->pages.npages, f_offset, (unsigned int)rdata->args.count);
+		hdr->page_array.npages, f_offset,
+		(unsigned int)hdr->args.count);
 
-	par = alloc_parallel(rdata);
+	par = alloc_parallel(hdr);
 	if (!par)
 		goto use_mds;
 	par->pnfs_callback = bl_end_par_io_read;
@@ -268,7 +268,7 @@ bl_read_pagelist(struct nfs_pgio_data *rdata)
 
 	isect = (sector_t) (f_offset >> SECTOR_SHIFT);
 	/* Code assumes extents are page-aligned */
-	for (i = pg_index; i < rdata->pages.npages; i++) {
+	for (i = pg_index; i < hdr->page_array.npages; i++) {
 		if (!extent_length) {
 			/* We've used up the previous extent */
 			bl_put_extent(be);
@@ -317,7 +317,8 @@ bl_read_pagelist(struct nfs_pgio_data *rdata)
 			struct pnfs_block_extent *be_read;
 
 			be_read = (hole && cow_read) ? cow_read : be;
-			bio = do_add_page_to_bio(bio, rdata->pages.npages - i,
+			bio = do_add_page_to_bio(bio,
+						 hdr->page_array.npages - i,
 						 READ,
 						 isect, pages[i], be_read,
 						 bl_end_io_read, par,
@@ -332,10 +333,10 @@ bl_read_pagelist(struct nfs_pgio_data *rdata)
 		extent_length -= PAGE_CACHE_SECTORS;
 	}
 	if ((isect << SECTOR_SHIFT) >= header->inode->i_size) {
-		rdata->res.eof = 1;
-		rdata->res.count = header->inode->i_size - rdata->args.offset;
+		hdr->res.eof = 1;
+		hdr->res.count = header->inode->i_size - hdr->args.offset;
 	} else {
-		rdata->res.count = (isect << SECTOR_SHIFT) - rdata->args.offset;
+		hdr->res.count = (isect << SECTOR_SHIFT) - hdr->args.offset;
 	}
 out:
 	bl_put_extent(be);
@@ -390,8 +391,7 @@ static void bl_end_io_write_zero(struct bio *bio, int err)
 	}
 
 	if (unlikely(err)) {
-		struct nfs_pgio_data *data = par->data;
-		struct nfs_pgio_header *header = data->header;
+		struct nfs_pgio_header *header = par->data;
 
 		if (!header->pnfs_error)
 			header->pnfs_error = -EIO;
@@ -405,8 +405,7 @@ static void bl_end_io_write(struct bio *bio, int err)
 {
 	struct parallel_io *par = bio->bi_private;
 	const int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags);
-	struct nfs_pgio_data *data = par->data;
-	struct nfs_pgio_header *header = data->header;
+	struct nfs_pgio_header *header = par->data;
 
 	if (!uptodate) {
 		if (!header->pnfs_error)
@@ -423,32 +422,32 @@ static void bl_end_io_write(struct bio *bio, int err)
 static void bl_write_cleanup(struct work_struct *work)
 {
 	struct rpc_task *task;
-	struct nfs_pgio_data *wdata;
+	struct nfs_pgio_header *hdr;
 	dprintk("%s enter\n", __func__);
 	task = container_of(work, struct rpc_task, u.tk_work);
-	wdata = container_of(task, struct nfs_pgio_data, task);
-	if (likely(!wdata->header->pnfs_error)) {
+	hdr = container_of(task, struct nfs_pgio_header, task);
+	if (likely(!hdr->pnfs_error)) {
 		/* Marks for LAYOUTCOMMIT */
-		mark_extents_written(BLK_LSEG2EXT(wdata->header->lseg),
-				     wdata->args.offset, wdata->args.count);
+		mark_extents_written(BLK_LSEG2EXT(hdr->lseg),
+				     hdr->args.offset, hdr->args.count);
 	}
-	pnfs_ld_write_done(wdata);
+	pnfs_ld_write_done(hdr);
 }
 
 /* Called when last of bios associated with a bl_write_pagelist call finishes */
 static void bl_end_par_io_write(void *data, int num_se)
 {
-	struct nfs_pgio_data *wdata = data;
+	struct nfs_pgio_header *hdr = data;
 
-	if (unlikely(wdata->header->pnfs_error)) {
-		bl_free_short_extents(&BLK_LSEG2EXT(wdata->header->lseg)->bl_inval,
+	if (unlikely(hdr->pnfs_error)) {
+		bl_free_short_extents(&BLK_LSEG2EXT(hdr->lseg)->bl_inval,
 					num_se);
 	}
 
-	wdata->task.tk_status = wdata->header->pnfs_error;
-	wdata->verf.committed = NFS_FILE_SYNC;
-	INIT_WORK(&wdata->task.u.tk_work, bl_write_cleanup);
-	schedule_work(&wdata->task.u.tk_work);
+	hdr->task.tk_status = hdr->pnfs_error;
+	hdr->verf.committed = NFS_FILE_SYNC;
+	INIT_WORK(&hdr->task.u.tk_work, bl_write_cleanup);
+	schedule_work(&hdr->task.u.tk_work);
 }
 
 /* FIXME STUB - mark intersection of layout and page as bad, so is not
@@ -673,18 +672,17 @@ check_page:
 }
 
 static enum pnfs_try_status
-bl_write_pagelist(struct nfs_pgio_data *wdata, int sync)
+bl_write_pagelist(struct nfs_pgio_header *header, int sync)
 {
-	struct nfs_pgio_header *header = wdata->header;
 	int i, ret, npg_zero, pg_index, last = 0;
 	struct bio *bio = NULL;
 	struct pnfs_block_extent *be = NULL, *cow_read = NULL;
 	sector_t isect, last_isect = 0, extent_length = 0;
 	struct parallel_io *par = NULL;
-	loff_t offset = wdata->args.offset;
-	size_t count = wdata->args.count;
+	loff_t offset = header->args.offset;
+	size_t count = header->args.count;
 	unsigned int pg_offset, pg_len, saved_len;
-	struct page **pages = wdata->args.pages;
+	struct page **pages = header->args.pages;
 	struct page *page;
 	pgoff_t index;
 	u64 temp;
@@ -699,11 +697,11 @@ bl_write_pagelist(struct nfs_pgio_data *wdata, int sync)
 		dprintk("pnfsblock nonblock aligned DIO writes. Resend MDS\n");
 		goto out_mds;
 	}
-	/* At this point, wdata->pages is a (sequential) list of nfs_pages.
+	/* At this point, header->page_aray is a (sequential) list of nfs_pages.
 	 * We want to write each, and if there is an error set pnfs_error
 	 * to have it redone using nfs.
 	 */
-	par = alloc_parallel(wdata);
+	par = alloc_parallel(header);
 	if (!par)
 		goto out_mds;
 	par->pnfs_callback = bl_end_par_io_write;
@@ -790,8 +788,8 @@ next_page:
 	bio = bl_submit_bio(WRITE, bio);
 
 	/* Middle pages */
-	pg_index = wdata->args.pgbase >> PAGE_CACHE_SHIFT;
-	for (i = pg_index; i < wdata->pages.npages; i++) {
+	pg_index = header->args.pgbase >> PAGE_CACHE_SHIFT;
+	for (i = pg_index; i < header->page_array.npages; i++) {
 		if (!extent_length) {
 			/* We've used up the previous extent */
 			bl_put_extent(be);
@@ -862,7 +860,8 @@ next_page:
 		}
 
 
-		bio = do_add_page_to_bio(bio, wdata->pages.npages - i, WRITE,
+		bio = do_add_page_to_bio(bio, header->page_array.npages - i,
+					 WRITE,
 					 isect, pages[i], be,
 					 bl_end_io_write, par,
 					 pg_offset, pg_len);
@@ -890,7 +889,7 @@ next_page:
 	}
 
 write_done:
-	wdata->res.count = wdata->args.count;
+	header->res.count = header->args.count;
 out:
 	bl_put_extent(be);
 	bl_put_extent(cow_read);
@@ -1063,7 +1062,7 @@ nfs4_blk_get_deviceinfo(struct nfs_server *server, const struct nfs_fh *fh,
 		return ERR_PTR(-ENOMEM);
 	}
 
-	pages = kzalloc(max_pages * sizeof(struct page *), GFP_NOFS);
+	pages = kcalloc(max_pages, sizeof(struct page *), GFP_NOFS);
 	if (pages == NULL) {
 		kfree(dev);
 		return ERR_PTR(-ENOMEM);
diff --git a/fs/nfs/callback.c b/fs/nfs/callback.c
index 073b4cf67ed9..54de482143cc 100644
--- a/fs/nfs/callback.c
+++ b/fs/nfs/callback.c
@@ -428,6 +428,18 @@ check_gss_callback_principal(struct nfs_client *clp, struct svc_rqst *rqstp)
 	if (p == NULL)
 		return 0;
 
+	/*
+	 * Did we get the acceptor from userland during the SETCLIENID
+	 * negotiation?
+	 */
+	if (clp->cl_acceptor)
+		return !strcmp(p, clp->cl_acceptor);
+
+	/*
+	 * Otherwise try to verify it using the cl_hostname. Note that this
+	 * doesn't work if a non-canonical hostname was used in the devname.
+	 */
+
 	/* Expect a GSS_C_NT_HOSTBASED_NAME like "nfs@serverhostname" */
 
 	if (memcmp(p, "nfs@", 4) != 0)
diff --git a/fs/nfs/client.c b/fs/nfs/client.c
index 1d09289c8f0e..1c5ff6d58385 100644
--- a/fs/nfs/client.c
+++ b/fs/nfs/client.c
@@ -110,8 +110,8 @@ struct nfs_subversion *get_nfs_version(unsigned int version)
 		mutex_unlock(&nfs_version_mutex);
 	}
 
-	if (!IS_ERR(nfs))
-		try_module_get(nfs->owner);
+	if (!IS_ERR(nfs) && !try_module_get(nfs->owner))
+		return ERR_PTR(-EAGAIN);
 	return nfs;
 }
 
@@ -158,7 +158,8 @@ struct nfs_client *nfs_alloc_client(const struct nfs_client_initdata *cl_init)
 		goto error_0;
 
 	clp->cl_nfs_mod = cl_init->nfs_mod;
-	try_module_get(clp->cl_nfs_mod->owner);
+	if (!try_module_get(clp->cl_nfs_mod->owner))
+		goto error_dealloc;
 
 	clp->rpc_ops = clp->cl_nfs_mod->rpc_ops;
 
@@ -190,6 +191,7 @@ struct nfs_client *nfs_alloc_client(const struct nfs_client_initdata *cl_init)
 
 error_cleanup:
 	put_nfs_version(clp->cl_nfs_mod);
+error_dealloc:
 	kfree(clp);
 error_0:
 	return ERR_PTR(err);
@@ -252,6 +254,7 @@ void nfs_free_client(struct nfs_client *clp)
 	put_net(clp->cl_net);
 	put_nfs_version(clp->cl_nfs_mod);
 	kfree(clp->cl_hostname);
+	kfree(clp->cl_acceptor);
 	kfree(clp);
 
 	dprintk("<-- nfs_free_client()\n");
@@ -482,8 +485,13 @@ nfs_get_client(const struct nfs_client_initdata *cl_init,
 	struct nfs_net *nn = net_generic(cl_init->net, nfs_net_id);
 	const struct nfs_rpc_ops *rpc_ops = cl_init->nfs_mod->rpc_ops;
 
+	if (cl_init->hostname == NULL) {
+		WARN_ON(1);
+		return NULL;
+	}
+
 	dprintk("--> nfs_get_client(%s,v%u)\n",
-		cl_init->hostname ?: "", rpc_ops->version);
+		cl_init->hostname, rpc_ops->version);
 
 	/* see if the client already exists */
 	do {
@@ -510,7 +518,7 @@ nfs_get_client(const struct nfs_client_initdata *cl_init,
 	} while (!IS_ERR(new));
 
 	dprintk("<-- nfs_get_client() Failed to find %s (%ld)\n",
-		cl_init->hostname ?: "", PTR_ERR(new));
+		cl_init->hostname, PTR_ERR(new));
 	return new;
 }
 EXPORT_SYMBOL_GPL(nfs_get_client);
@@ -1205,7 +1213,7 @@ static const struct file_operations nfs_server_list_fops = {
 	.open		= nfs_server_list_open,
 	.read		= seq_read,
 	.llseek		= seq_lseek,
-	.release	= seq_release,
+	.release	= seq_release_net,
 	.owner		= THIS_MODULE,
 };
 
@@ -1226,7 +1234,7 @@ static const struct file_operations nfs_volume_list_fops = {
 	.open		= nfs_volume_list_open,
 	.read		= seq_read,
 	.llseek		= seq_lseek,
-	.release	= seq_release,
+	.release	= seq_release_net,
 	.owner		= THIS_MODULE,
 };
 
@@ -1236,19 +1244,8 @@ static const struct file_operations nfs_volume_list_fops = {
  */
 static int nfs_server_list_open(struct inode *inode, struct file *file)
 {
-	struct seq_file *m;
-	int ret;
-	struct pid_namespace *pid_ns = file->f_dentry->d_sb->s_fs_info;
-	struct net *net = pid_ns->child_reaper->nsproxy->net_ns;
-
-	ret = seq_open(file, &nfs_server_list_ops);
-	if (ret < 0)
-		return ret;
-
-	m = file->private_data;
-	m->private = net;
-
-	return 0;
+	return seq_open_net(inode, file, &nfs_server_list_ops,
+			   sizeof(struct seq_net_private));
 }
 
 /*
@@ -1256,7 +1253,7 @@ static int nfs_server_list_open(struct inode *inode, struct file *file)
  */
 static void *nfs_server_list_start(struct seq_file *m, loff_t *_pos)
 {
-	struct nfs_net *nn = net_generic(m->private, nfs_net_id);
+	struct nfs_net *nn = net_generic(seq_file_net(m), nfs_net_id);
 
 	/* lock the list against modification */
 	spin_lock(&nn->nfs_client_lock);
@@ -1268,7 +1265,7 @@ static void *nfs_server_list_start(struct seq_file *m, loff_t *_pos)
  */
 static void *nfs_server_list_next(struct seq_file *p, void *v, loff_t *pos)
 {
-	struct nfs_net *nn = net_generic(p->private, nfs_net_id);
+	struct nfs_net *nn = net_generic(seq_file_net(p), nfs_net_id);
 
 	return seq_list_next(v, &nn->nfs_client_list, pos);
 }
@@ -1278,7 +1275,7 @@ static void *nfs_server_list_next(struct seq_file *p, void *v, loff_t *pos)
  */
 static void nfs_server_list_stop(struct seq_file *p, void *v)
 {
-	struct nfs_net *nn = net_generic(p->private, nfs_net_id);
+	struct nfs_net *nn = net_generic(seq_file_net(p), nfs_net_id);
 
 	spin_unlock(&nn->nfs_client_lock);
 }
@@ -1289,7 +1286,7 @@ static void nfs_server_list_stop(struct seq_file *p, void *v)
 static int nfs_server_list_show(struct seq_file *m, void *v)
 {
 	struct nfs_client *clp;
-	struct nfs_net *nn = net_generic(m->private, nfs_net_id);
+	struct nfs_net *nn = net_generic(seq_file_net(m), nfs_net_id);
 
 	/* display header on line 1 */
 	if (v == &nn->nfs_client_list) {
@@ -1321,19 +1318,8 @@ static int nfs_server_list_show(struct seq_file *m, void *v)
  */
 static int nfs_volume_list_open(struct inode *inode, struct file *file)
 {
-	struct seq_file *m;
-	int ret;
-	struct pid_namespace *pid_ns = file->f_dentry->d_sb->s_fs_info;
-	struct net *net = pid_ns->child_reaper->nsproxy->net_ns;
-
-	ret = seq_open(file, &nfs_volume_list_ops);
-	if (ret < 0)
-		return ret;
-
-	m = file->private_data;
-	m->private = net;
-
-	return 0;
+	return seq_open_net(inode, file, &nfs_server_list_ops,
+			   sizeof(struct seq_net_private));
 }
 
 /*
@@ -1341,7 +1327,7 @@ static int nfs_volume_list_open(struct inode *inode, struct file *file)
  */
 static void *nfs_volume_list_start(struct seq_file *m, loff_t *_pos)
 {
-	struct nfs_net *nn = net_generic(m->private, nfs_net_id);
+	struct nfs_net *nn = net_generic(seq_file_net(m), nfs_net_id);
 
 	/* lock the list against modification */
 	spin_lock(&nn->nfs_client_lock);
@@ -1353,7 +1339,7 @@ static void *nfs_volume_list_start(struct seq_file *m, loff_t *_pos)
  */
 static void *nfs_volume_list_next(struct seq_file *p, void *v, loff_t *pos)
 {
-	struct nfs_net *nn = net_generic(p->private, nfs_net_id);
+	struct nfs_net *nn = net_generic(seq_file_net(p), nfs_net_id);
 
 	return seq_list_next(v, &nn->nfs_volume_list, pos);
 }
@@ -1363,7 +1349,7 @@ static void *nfs_volume_list_next(struct seq_file *p, void *v, loff_t *pos)
  */
 static void nfs_volume_list_stop(struct seq_file *p, void *v)
 {
-	struct nfs_net *nn = net_generic(p->private, nfs_net_id);
+	struct nfs_net *nn = net_generic(seq_file_net(p), nfs_net_id);
 
 	spin_unlock(&nn->nfs_client_lock);
 }
@@ -1376,7 +1362,7 @@ static int nfs_volume_list_show(struct seq_file *m, void *v)
 	struct nfs_server *server;
 	struct nfs_client *clp;
 	char dev[8], fsid[17];
-	struct nfs_net *nn = net_generic(m->private, nfs_net_id);
+	struct nfs_net *nn = net_generic(seq_file_net(m), nfs_net_id);
 
 	/* display header on line 1 */
 	if (v == &nn->nfs_volume_list) {
@@ -1407,6 +1393,45 @@ static int nfs_volume_list_show(struct seq_file *m, void *v)
 	return 0;
 }
 
+int nfs_fs_proc_net_init(struct net *net)
+{
+	struct nfs_net *nn = net_generic(net, nfs_net_id);
+	struct proc_dir_entry *p;
+
+	nn->proc_nfsfs = proc_net_mkdir(net, "nfsfs", net->proc_net);
+	if (!nn->proc_nfsfs)
+		goto error_0;
+
+	/* a file of servers with which we're dealing */
+	p = proc_create("servers", S_IFREG|S_IRUGO,
+			nn->proc_nfsfs, &nfs_server_list_fops);
+	if (!p)
+		goto error_1;
+
+	/* a file of volumes that we have mounted */
+	p = proc_create("volumes", S_IFREG|S_IRUGO,
+			nn->proc_nfsfs, &nfs_volume_list_fops);
+	if (!p)
+		goto error_2;
+	return 0;
+
+error_2:
+	remove_proc_entry("servers", nn->proc_nfsfs);
+error_1:
+	remove_proc_entry("fs/nfsfs", NULL);
+error_0:
+	return -ENOMEM;
+}
+
+void nfs_fs_proc_net_exit(struct net *net)
+{
+	struct nfs_net *nn = net_generic(net, nfs_net_id);
+
+	remove_proc_entry("volumes", nn->proc_nfsfs);
+	remove_proc_entry("servers", nn->proc_nfsfs);
+	remove_proc_entry("fs/nfsfs", NULL);
+}
+
 /*
  * initialise the /proc/fs/nfsfs/ directory
  */
@@ -1419,14 +1444,12 @@ int __init nfs_fs_proc_init(void)
 		goto error_0;
 
 	/* a file of servers with which we're dealing */
-	p = proc_create("servers", S_IFREG|S_IRUGO,
-			proc_fs_nfs, &nfs_server_list_fops);
+	p = proc_symlink("servers", proc_fs_nfs, "../../net/nfsfs/servers");
 	if (!p)
 		goto error_1;
 
 	/* a file of volumes that we have mounted */
-	p = proc_create("volumes", S_IFREG|S_IRUGO,
-			proc_fs_nfs, &nfs_volume_list_fops);
+	p = proc_symlink("volumes", proc_fs_nfs, "../../net/nfsfs/volumes");
 	if (!p)
 		goto error_2;
 	return 0;
diff --git a/fs/nfs/delegation.c b/fs/nfs/delegation.c
index 5d8ccecf5f5c..5853f53db732 100644
--- a/fs/nfs/delegation.c
+++ b/fs/nfs/delegation.c
@@ -41,14 +41,8 @@ void nfs_mark_delegation_referenced(struct nfs_delegation *delegation)
 	set_bit(NFS_DELEGATION_REFERENCED, &delegation->flags);
 }
 
-/**
- * nfs_have_delegation - check if inode has a delegation
- * @inode: inode to check
- * @flags: delegation types to check for
- *
- * Returns one if inode has the indicated delegation, otherwise zero.
- */
-int nfs4_have_delegation(struct inode *inode, fmode_t flags)
+static int
+nfs4_do_check_delegation(struct inode *inode, fmode_t flags, bool mark)
 {
 	struct nfs_delegation *delegation;
 	int ret = 0;
@@ -58,12 +52,34 @@ int nfs4_have_delegation(struct inode *inode, fmode_t flags)
 	delegation = rcu_dereference(NFS_I(inode)->delegation);
 	if (delegation != NULL && (delegation->type & flags) == flags &&
 	    !test_bit(NFS_DELEGATION_RETURNING, &delegation->flags)) {
-		nfs_mark_delegation_referenced(delegation);
+		if (mark)
+			nfs_mark_delegation_referenced(delegation);
 		ret = 1;
 	}
 	rcu_read_unlock();
 	return ret;
 }
+/**
+ * nfs_have_delegation - check if inode has a delegation, mark it
+ * NFS_DELEGATION_REFERENCED if there is one.
+ * @inode: inode to check
+ * @flags: delegation types to check for
+ *
+ * Returns one if inode has the indicated delegation, otherwise zero.
+ */
+int nfs4_have_delegation(struct inode *inode, fmode_t flags)
+{
+	return nfs4_do_check_delegation(inode, flags, true);
+}
+
+/*
+ * nfs4_check_delegation - check if inode has a delegation, do not mark
+ * NFS_DELEGATION_REFERENCED if it has one.
+ */
+int nfs4_check_delegation(struct inode *inode, fmode_t flags)
+{
+	return nfs4_do_check_delegation(inode, flags, false);
+}
 
 static int nfs_delegation_claim_locks(struct nfs_open_context *ctx, struct nfs4_state *state, const nfs4_stateid *stateid)
 {
diff --git a/fs/nfs/delegation.h b/fs/nfs/delegation.h
index 9a79c7a99d6d..5c1cce39297f 100644
--- a/fs/nfs/delegation.h
+++ b/fs/nfs/delegation.h
@@ -59,6 +59,7 @@ bool nfs4_copy_delegation_stateid(nfs4_stateid *dst, struct inode *inode, fmode_
 
 void nfs_mark_delegation_referenced(struct nfs_delegation *delegation);
 int nfs4_have_delegation(struct inode *inode, fmode_t flags);
+int nfs4_check_delegation(struct inode *inode, fmode_t flags);
 
 #endif
 
diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index 4a3d4ef76127..36d921f0c602 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -988,9 +988,13 @@ EXPORT_SYMBOL_GPL(nfs_force_lookup_revalidate);
  * A check for whether or not the parent directory has changed.
  * In the case it has, we assume that the dentries are untrustworthy
  * and may need to be looked up again.
+ * If rcu_walk prevents us from performing a full check, return 0.
  */
-static int nfs_check_verifier(struct inode *dir, struct dentry *dentry)
+static int nfs_check_verifier(struct inode *dir, struct dentry *dentry,
+			      int rcu_walk)
 {
+	int ret;
+
 	if (IS_ROOT(dentry))
 		return 1;
 	if (NFS_SERVER(dir)->flags & NFS_MOUNT_LOOKUP_CACHE_NONE)
@@ -998,7 +1002,11 @@ static int nfs_check_verifier(struct inode *dir, struct dentry *dentry)
 	if (!nfs_verify_change_attribute(dir, dentry->d_time))
 		return 0;
 	/* Revalidate nfsi->cache_change_attribute before we declare a match */
-	if (nfs_revalidate_inode(NFS_SERVER(dir), dir) < 0)
+	if (rcu_walk)
+		ret = nfs_revalidate_inode_rcu(NFS_SERVER(dir), dir);
+	else
+		ret = nfs_revalidate_inode(NFS_SERVER(dir), dir);
+	if (ret < 0)
 		return 0;
 	if (!nfs_verify_change_attribute(dir, dentry->d_time))
 		return 0;
@@ -1042,6 +1050,8 @@ int nfs_lookup_verify_inode(struct inode *inode, unsigned int flags)
 out:
 	return (inode->i_nlink == 0) ? -ENOENT : 0;
 out_force:
+	if (flags & LOOKUP_RCU)
+		return -ECHILD;
 	ret = __nfs_revalidate_inode(server, inode);
 	if (ret != 0)
 		return ret;
@@ -1054,6 +1064,9 @@ out_force:
  *
  * If parent mtime has changed, we revalidate, else we wait for a
  * period corresponding to the parent's attribute cache timeout value.
+ *
+ * If LOOKUP_RCU prevents us from performing a full check, return 1
+ * suggesting a reval is needed.
  */
 static inline
 int nfs_neg_need_reval(struct inode *dir, struct dentry *dentry,
@@ -1064,7 +1077,7 @@ int nfs_neg_need_reval(struct inode *dir, struct dentry *dentry,
 		return 0;
 	if (NFS_SERVER(dir)->flags & NFS_MOUNT_LOOKUP_CACHE_NONEG)
 		return 1;
-	return !nfs_check_verifier(dir, dentry);
+	return !nfs_check_verifier(dir, dentry, flags & LOOKUP_RCU);
 }
 
 /*
@@ -1088,21 +1101,30 @@ static int nfs_lookup_revalidate(struct dentry *dentry, unsigned int flags)
 	struct nfs4_label *label = NULL;
 	int error;
 
-	if (flags & LOOKUP_RCU)
-		return -ECHILD;
-
-	parent = dget_parent(dentry);
-	dir = parent->d_inode;
+	if (flags & LOOKUP_RCU) {
+		parent = ACCESS_ONCE(dentry->d_parent);
+		dir = ACCESS_ONCE(parent->d_inode);
+		if (!dir)
+			return -ECHILD;
+	} else {
+		parent = dget_parent(dentry);
+		dir = parent->d_inode;
+	}
 	nfs_inc_stats(dir, NFSIOS_DENTRYREVALIDATE);
 	inode = dentry->d_inode;
 
 	if (!inode) {
-		if (nfs_neg_need_reval(dir, dentry, flags))
+		if (nfs_neg_need_reval(dir, dentry, flags)) {
+			if (flags & LOOKUP_RCU)
+				return -ECHILD;
 			goto out_bad;
+		}
 		goto out_valid_noent;
 	}
 
 	if (is_bad_inode(inode)) {
+		if (flags & LOOKUP_RCU)
+			return -ECHILD;
 		dfprintk(LOOKUPCACHE, "%s: %pd2 has dud inode\n",
 				__func__, dentry);
 		goto out_bad;
@@ -1112,12 +1134,20 @@ static int nfs_lookup_revalidate(struct dentry *dentry, unsigned int flags)
 		goto out_set_verifier;
 
 	/* Force a full look up iff the parent directory has changed */
-	if (!nfs_is_exclusive_create(dir, flags) && nfs_check_verifier(dir, dentry)) {
-		if (nfs_lookup_verify_inode(inode, flags))
+	if (!nfs_is_exclusive_create(dir, flags) &&
+	    nfs_check_verifier(dir, dentry, flags & LOOKUP_RCU)) {
+
+		if (nfs_lookup_verify_inode(inode, flags)) {
+			if (flags & LOOKUP_RCU)
+				return -ECHILD;
 			goto out_zap_parent;
+		}
 		goto out_valid;
 	}
 
+	if (flags & LOOKUP_RCU)
+		return -ECHILD;
+
 	if (NFS_STALE(inode))
 		goto out_bad;
 
@@ -1153,13 +1183,18 @@ out_set_verifier:
 	/* Success: notify readdir to use READDIRPLUS */
 	nfs_advise_use_readdirplus(dir);
  out_valid_noent:
-	dput(parent);
+	if (flags & LOOKUP_RCU) {
+		if (parent != ACCESS_ONCE(dentry->d_parent))
+			return -ECHILD;
+	} else
+		dput(parent);
 	dfprintk(LOOKUPCACHE, "NFS: %s(%pd2) is valid\n",
 			__func__, dentry);
 	return 1;
 out_zap_parent:
 	nfs_zap_caches(dir);
  out_bad:
+	WARN_ON(flags & LOOKUP_RCU);
 	nfs_free_fattr(fattr);
 	nfs_free_fhandle(fhandle);
 	nfs4_label_free(label);
@@ -1185,6 +1220,7 @@ out_zap_parent:
 			__func__, dentry);
 	return 0;
 out_error:
+	WARN_ON(flags & LOOKUP_RCU);
 	nfs_free_fattr(fattr);
 	nfs_free_fhandle(fhandle);
 	nfs4_label_free(label);
@@ -1529,14 +1565,9 @@ EXPORT_SYMBOL_GPL(nfs_atomic_open);
 
 static int nfs4_lookup_revalidate(struct dentry *dentry, unsigned int flags)
 {
-	struct dentry *parent = NULL;
 	struct inode *inode;
-	struct inode *dir;
 	int ret = 0;
 
-	if (flags & LOOKUP_RCU)
-		return -ECHILD;
-
 	if (!(flags & LOOKUP_OPEN) || (flags & LOOKUP_DIRECTORY))
 		goto no_open;
 	if (d_mountpoint(dentry))
@@ -1545,34 +1576,47 @@ static int nfs4_lookup_revalidate(struct dentry *dentry, unsigned int flags)
 		goto no_open;
 
 	inode = dentry->d_inode;
-	parent = dget_parent(dentry);
-	dir = parent->d_inode;
 
 	/* We can't create new files in nfs_open_revalidate(), so we
 	 * optimize away revalidation of negative dentries.
 	 */
 	if (inode == NULL) {
+		struct dentry *parent;
+		struct inode *dir;
+
+		if (flags & LOOKUP_RCU) {
+			parent = ACCESS_ONCE(dentry->d_parent);
+			dir = ACCESS_ONCE(parent->d_inode);
+			if (!dir)
+				return -ECHILD;
+		} else {
+			parent = dget_parent(dentry);
+			dir = parent->d_inode;
+		}
 		if (!nfs_neg_need_reval(dir, dentry, flags))
 			ret = 1;
+		else if (flags & LOOKUP_RCU)
+			ret = -ECHILD;
+		if (!(flags & LOOKUP_RCU))
+			dput(parent);
+		else if (parent != ACCESS_ONCE(dentry->d_parent))
+			return -ECHILD;
 		goto out;
 	}
 
 	/* NFS only supports OPEN on regular files */
 	if (!S_ISREG(inode->i_mode))
-		goto no_open_dput;
+		goto no_open;
 	/* We cannot do exclusive creation on a positive dentry */
 	if (flags & LOOKUP_EXCL)
-		goto no_open_dput;
+		goto no_open;
 
 	/* Let f_op->open() actually open (and revalidate) the file */
 	ret = 1;
 
 out:
-	dput(parent);
 	return ret;
 
-no_open_dput:
-	dput(parent);
 no_open:
 	return nfs_lookup_revalidate(dentry, flags);
 }
@@ -2028,10 +2072,14 @@ static DEFINE_SPINLOCK(nfs_access_lru_lock);
 static LIST_HEAD(nfs_access_lru_list);
 static atomic_long_t nfs_access_nr_entries;
 
+static unsigned long nfs_access_max_cachesize = ULONG_MAX;
+module_param(nfs_access_max_cachesize, ulong, 0644);
+MODULE_PARM_DESC(nfs_access_max_cachesize, "NFS access maximum total cache length");
+
 static void nfs_access_free_entry(struct nfs_access_entry *entry)
 {
 	put_rpccred(entry->cred);
-	kfree(entry);
+	kfree_rcu(entry, rcu_head);
 	smp_mb__before_atomic();
 	atomic_long_dec(&nfs_access_nr_entries);
 	smp_mb__after_atomic();
@@ -2048,19 +2096,14 @@ static void nfs_access_free_list(struct list_head *head)
 	}
 }
 
-unsigned long
-nfs_access_cache_scan(struct shrinker *shrink, struct shrink_control *sc)
+static unsigned long
+nfs_do_access_cache_scan(unsigned int nr_to_scan)
 {
 	LIST_HEAD(head);
 	struct nfs_inode *nfsi, *next;
 	struct nfs_access_entry *cache;
-	int nr_to_scan = sc->nr_to_scan;
-	gfp_t gfp_mask = sc->gfp_mask;
 	long freed = 0;
 
-	if ((gfp_mask & GFP_KERNEL) != GFP_KERNEL)
-		return SHRINK_STOP;
-
 	spin_lock(&nfs_access_lru_lock);
 	list_for_each_entry_safe(nfsi, next, &nfs_access_lru_list, access_cache_inode_lru) {
 		struct inode *inode;
@@ -2094,11 +2137,39 @@ remove_lru_entry:
 }
 
 unsigned long
+nfs_access_cache_scan(struct shrinker *shrink, struct shrink_control *sc)
+{
+	int nr_to_scan = sc->nr_to_scan;
+	gfp_t gfp_mask = sc->gfp_mask;
+
+	if ((gfp_mask & GFP_KERNEL) != GFP_KERNEL)
+		return SHRINK_STOP;
+	return nfs_do_access_cache_scan(nr_to_scan);
+}
+
+
+unsigned long
 nfs_access_cache_count(struct shrinker *shrink, struct shrink_control *sc)
 {
 	return vfs_pressure_ratio(atomic_long_read(&nfs_access_nr_entries));
 }
 
+static void
+nfs_access_cache_enforce_limit(void)
+{
+	long nr_entries = atomic_long_read(&nfs_access_nr_entries);
+	unsigned long diff;
+	unsigned int nr_to_scan;
+
+	if (nr_entries < 0 || nr_entries <= nfs_access_max_cachesize)
+		return;
+	nr_to_scan = 100;
+	diff = nr_entries - nfs_access_max_cachesize;
+	if (diff < nr_to_scan)
+		nr_to_scan = diff;
+	nfs_do_access_cache_scan(nr_to_scan);
+}
+
 static void __nfs_access_zap_cache(struct nfs_inode *nfsi, struct list_head *head)
 {
 	struct rb_root *root_node = &nfsi->access_cache;
@@ -2186,6 +2257,38 @@ out_zap:
 	return -ENOENT;
 }
 
+static int nfs_access_get_cached_rcu(struct inode *inode, struct rpc_cred *cred, struct nfs_access_entry *res)
+{
+	/* Only check the most recently returned cache entry,
+	 * but do it without locking.
+	 */
+	struct nfs_inode *nfsi = NFS_I(inode);
+	struct nfs_access_entry *cache;
+	int err = -ECHILD;
+	struct list_head *lh;
+
+	rcu_read_lock();
+	if (nfsi->cache_validity & NFS_INO_INVALID_ACCESS)
+		goto out;
+	lh = rcu_dereference(nfsi->access_cache_entry_lru.prev);
+	cache = list_entry(lh, struct nfs_access_entry, lru);
+	if (lh == &nfsi->access_cache_entry_lru ||
+	    cred != cache->cred)
+		cache = NULL;
+	if (cache == NULL)
+		goto out;
+	if (!nfs_have_delegated_attributes(inode) &&
+	    !time_in_range_open(jiffies, cache->jiffies, cache->jiffies + nfsi->attrtimeo))
+		goto out;
+	res->jiffies = cache->jiffies;
+	res->cred = cache->cred;
+	res->mask = cache->mask;
+	err = 0;
+out:
+	rcu_read_unlock();
+	return err;
+}
+
 static void nfs_access_add_rbtree(struct inode *inode, struct nfs_access_entry *set)
 {
 	struct nfs_inode *nfsi = NFS_I(inode);
@@ -2229,6 +2332,11 @@ void nfs_access_add_cache(struct inode *inode, struct nfs_access_entry *set)
 	cache->cred = get_rpccred(set->cred);
 	cache->mask = set->mask;
 
+	/* The above field assignments must be visible
+	 * before this item appears on the lru.  We cannot easily
+	 * use rcu_assign_pointer, so just force the memory barrier.
+	 */
+	smp_wmb();
 	nfs_access_add_rbtree(inode, cache);
 
 	/* Update accounting */
@@ -2244,6 +2352,7 @@ void nfs_access_add_cache(struct inode *inode, struct nfs_access_entry *set)
 					&nfs_access_lru_list);
 		spin_unlock(&nfs_access_lru_lock);
 	}
+	nfs_access_cache_enforce_limit();
 }
 EXPORT_SYMBOL_GPL(nfs_access_add_cache);
 
@@ -2267,10 +2376,16 @@ static int nfs_do_access(struct inode *inode, struct rpc_cred *cred, int mask)
 
 	trace_nfs_access_enter(inode);
 
-	status = nfs_access_get_cached(inode, cred, &cache);
+	status = nfs_access_get_cached_rcu(inode, cred, &cache);
+	if (status != 0)
+		status = nfs_access_get_cached(inode, cred, &cache);
 	if (status == 0)
 		goto out_cached;
 
+	status = -ECHILD;
+	if (mask & MAY_NOT_BLOCK)
+		goto out;
+
 	/* Be clever: ask server to check for all possible rights */
 	cache.mask = MAY_EXEC | MAY_WRITE | MAY_READ;
 	cache.cred = cred;
@@ -2321,9 +2436,6 @@ int nfs_permission(struct inode *inode, int mask)
 	struct rpc_cred *cred;
 	int res = 0;
 
-	if (mask & MAY_NOT_BLOCK)
-		return -ECHILD;
-
 	nfs_inc_stats(inode, NFSIOS_VFSACCESS);
 
 	if ((mask & (MAY_READ | MAY_WRITE | MAY_EXEC)) == 0)
@@ -2350,12 +2462,23 @@ force_lookup:
 	if (!NFS_PROTO(inode)->access)
 		goto out_notsup;
 
-	cred = rpc_lookup_cred();
-	if (!IS_ERR(cred)) {
-		res = nfs_do_access(inode, cred, mask);
-		put_rpccred(cred);
-	} else
+	/* Always try fast lookups first */
+	rcu_read_lock();
+	cred = rpc_lookup_cred_nonblock();
+	if (!IS_ERR(cred))
+		res = nfs_do_access(inode, cred, mask|MAY_NOT_BLOCK);
+	else
 		res = PTR_ERR(cred);
+	rcu_read_unlock();
+	if (res == -ECHILD && !(mask & MAY_NOT_BLOCK)) {
+		/* Fast lookup failed, try the slow way */
+		cred = rpc_lookup_cred();
+		if (!IS_ERR(cred)) {
+			res = nfs_do_access(inode, cred, mask);
+			put_rpccred(cred);
+		} else
+			res = PTR_ERR(cred);
+	}
 out:
 	if (!res && (mask & MAY_EXEC) && !execute_ok(inode))
 		res = -EACCES;
@@ -2364,6 +2487,9 @@ out:
 		inode->i_sb->s_id, inode->i_ino, mask, res);
 	return res;
 out_notsup:
+	if (mask & MAY_NOT_BLOCK)
+		return -ECHILD;
+
 	res = nfs_revalidate_inode(NFS_SERVER(inode), inode);
 	if (res == 0)
 		res = generic_permission(inode, mask);
diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c
index 8f98138cbc43..65ef6e00deee 100644
--- a/fs/nfs/direct.c
+++ b/fs/nfs/direct.c
@@ -148,8 +148,8 @@ static void nfs_direct_set_hdr_verf(struct nfs_direct_req *dreq,
 {
 	struct nfs_writeverf *verfp;
 
-	verfp = nfs_direct_select_verf(dreq, hdr->data->ds_clp,
-				      hdr->data->ds_idx);
+	verfp = nfs_direct_select_verf(dreq, hdr->ds_clp,
+				      hdr->ds_idx);
 	WARN_ON_ONCE(verfp->committed >= 0);
 	memcpy(verfp, &hdr->verf, sizeof(struct nfs_writeverf));
 	WARN_ON_ONCE(verfp->committed < 0);
@@ -169,8 +169,8 @@ static int nfs_direct_set_or_cmp_hdr_verf(struct nfs_direct_req *dreq,
 {
 	struct nfs_writeverf *verfp;
 
-	verfp = nfs_direct_select_verf(dreq, hdr->data->ds_clp,
-					 hdr->data->ds_idx);
+	verfp = nfs_direct_select_verf(dreq, hdr->ds_clp,
+					 hdr->ds_idx);
 	if (verfp->committed < 0) {
 		nfs_direct_set_hdr_verf(dreq, hdr);
 		return 0;
@@ -715,7 +715,7 @@ static void nfs_direct_write_completion(struct nfs_pgio_header *hdr)
 {
 	struct nfs_direct_req *dreq = hdr->dreq;
 	struct nfs_commit_info cinfo;
-	int bit = -1;
+	bool request_commit = false;
 	struct nfs_page *req = nfs_list_entry(hdr->pages.next);
 
 	if (test_bit(NFS_IOHDR_REDO, &hdr->flags))
@@ -729,43 +729,32 @@ static void nfs_direct_write_completion(struct nfs_pgio_header *hdr)
 		dreq->flags = 0;
 		dreq->error = hdr->error;
 	}
-	if (dreq->error != 0)
-		bit = NFS_IOHDR_ERROR;
-	else {
+	if (dreq->error == 0) {
 		dreq->count += hdr->good_bytes;
-		if (test_bit(NFS_IOHDR_NEED_RESCHED, &hdr->flags)) {
-			dreq->flags = NFS_ODIRECT_RESCHED_WRITES;
-			bit = NFS_IOHDR_NEED_RESCHED;
-		} else if (test_bit(NFS_IOHDR_NEED_COMMIT, &hdr->flags)) {
+		if (nfs_write_need_commit(hdr)) {
 			if (dreq->flags == NFS_ODIRECT_RESCHED_WRITES)
-				bit = NFS_IOHDR_NEED_RESCHED;
+				request_commit = true;
 			else if (dreq->flags == 0) {
 				nfs_direct_set_hdr_verf(dreq, hdr);
-				bit = NFS_IOHDR_NEED_COMMIT;
+				request_commit = true;
 				dreq->flags = NFS_ODIRECT_DO_COMMIT;
 			} else if (dreq->flags == NFS_ODIRECT_DO_COMMIT) {
-				if (nfs_direct_set_or_cmp_hdr_verf(dreq, hdr)) {
+				request_commit = true;
+				if (nfs_direct_set_or_cmp_hdr_verf(dreq, hdr))
 					dreq->flags =
 						NFS_ODIRECT_RESCHED_WRITES;
-					bit = NFS_IOHDR_NEED_RESCHED;
-				} else
-					bit = NFS_IOHDR_NEED_COMMIT;
 			}
 		}
 	}
 	spin_unlock(&dreq->lock);
 
 	while (!list_empty(&hdr->pages)) {
-		bool do_destroy = true;
 
 		req = nfs_list_entry(hdr->pages.next);
 		nfs_list_remove_request(req);
-		switch (bit) {
-		case NFS_IOHDR_NEED_RESCHED:
-		case NFS_IOHDR_NEED_COMMIT:
+		if (request_commit) {
 			kref_get(&req->wb_kref);
 			nfs_mark_request_commit(req, hdr->lseg, &cinfo);
-			do_destroy = false;
 		}
 		nfs_unlock_and_release_request(req);
 	}
diff --git a/fs/nfs/file.c b/fs/nfs/file.c
index 4042ff58fe3f..524dd80d1898 100644
--- a/fs/nfs/file.c
+++ b/fs/nfs/file.c
@@ -361,8 +361,8 @@ start:
 	 * Prevent starvation issues if someone is doing a consistency
 	 * sync-to-disk
 	 */
-	ret = wait_on_bit(&NFS_I(mapping->host)->flags, NFS_INO_FLUSHING,
-			nfs_wait_bit_killable, TASK_KILLABLE);
+	ret = wait_on_bit_action(&NFS_I(mapping->host)->flags, NFS_INO_FLUSHING,
+				 nfs_wait_bit_killable, TASK_KILLABLE);
 	if (ret)
 		return ret;
 
diff --git a/fs/nfs/filelayout/filelayout.c b/fs/nfs/filelayout/filelayout.c
index d2eba1c13b7e..1359c4a27393 100644
--- a/fs/nfs/filelayout/filelayout.c
+++ b/fs/nfs/filelayout/filelayout.c
@@ -84,45 +84,37 @@ filelayout_get_dserver_offset(struct pnfs_layout_segment *lseg, loff_t offset)
 	BUG();
 }
 
-static void filelayout_reset_write(struct nfs_pgio_data *data)
+static void filelayout_reset_write(struct nfs_pgio_header *hdr)
 {
-	struct nfs_pgio_header *hdr = data->header;
-	struct rpc_task *task = &data->task;
+	struct rpc_task *task = &hdr->task;
 
 	if (!test_and_set_bit(NFS_IOHDR_REDO, &hdr->flags)) {
 		dprintk("%s Reset task %5u for i/o through MDS "
 			"(req %s/%llu, %u bytes @ offset %llu)\n", __func__,
-			data->task.tk_pid,
+			hdr->task.tk_pid,
 			hdr->inode->i_sb->s_id,
 			(unsigned long long)NFS_FILEID(hdr->inode),
-			data->args.count,
-			(unsigned long long)data->args.offset);
+			hdr->args.count,
+			(unsigned long long)hdr->args.offset);
 
-		task->tk_status = pnfs_write_done_resend_to_mds(hdr->inode,
-							&hdr->pages,
-							hdr->completion_ops,
-							hdr->dreq);
+		task->tk_status = pnfs_write_done_resend_to_mds(hdr);
 	}
 }
 
-static void filelayout_reset_read(struct nfs_pgio_data *data)
+static void filelayout_reset_read(struct nfs_pgio_header *hdr)
 {
-	struct nfs_pgio_header *hdr = data->header;
-	struct rpc_task *task = &data->task;
+	struct rpc_task *task = &hdr->task;
 
 	if (!test_and_set_bit(NFS_IOHDR_REDO, &hdr->flags)) {
 		dprintk("%s Reset task %5u for i/o through MDS "
 			"(req %s/%llu, %u bytes @ offset %llu)\n", __func__,
-			data->task.tk_pid,
+			hdr->task.tk_pid,
 			hdr->inode->i_sb->s_id,
 			(unsigned long long)NFS_FILEID(hdr->inode),
-			data->args.count,
-			(unsigned long long)data->args.offset);
+			hdr->args.count,
+			(unsigned long long)hdr->args.offset);
 
-		task->tk_status = pnfs_read_done_resend_to_mds(hdr->inode,
-							&hdr->pages,
-							hdr->completion_ops,
-							hdr->dreq);
+		task->tk_status = pnfs_read_done_resend_to_mds(hdr);
 	}
 }
 
@@ -243,18 +235,17 @@ wait_on_recovery:
 /* NFS_PROTO call done callback routines */
 
 static int filelayout_read_done_cb(struct rpc_task *task,
-				struct nfs_pgio_data *data)
+				struct nfs_pgio_header *hdr)
 {
-	struct nfs_pgio_header *hdr = data->header;
 	int err;
 
-	trace_nfs4_pnfs_read(data, task->tk_status);
-	err = filelayout_async_handle_error(task, data->args.context->state,
-					    data->ds_clp, hdr->lseg);
+	trace_nfs4_pnfs_read(hdr, task->tk_status);
+	err = filelayout_async_handle_error(task, hdr->args.context->state,
+					    hdr->ds_clp, hdr->lseg);
 
 	switch (err) {
 	case -NFS4ERR_RESET_TO_MDS:
-		filelayout_reset_read(data);
+		filelayout_reset_read(hdr);
 		return task->tk_status;
 	case -EAGAIN:
 		rpc_restart_call_prepare(task);
@@ -270,15 +261,14 @@ static int filelayout_read_done_cb(struct rpc_task *task,
  * rfc5661 is not clear about which credential should be used.
  */
 static void
-filelayout_set_layoutcommit(struct nfs_pgio_data *wdata)
+filelayout_set_layoutcommit(struct nfs_pgio_header *hdr)
 {
-	struct nfs_pgio_header *hdr = wdata->header;
 
 	if (FILELAYOUT_LSEG(hdr->lseg)->commit_through_mds ||
-	    wdata->res.verf->committed == NFS_FILE_SYNC)
+	    hdr->res.verf->committed == NFS_FILE_SYNC)
 		return;
 
-	pnfs_set_layoutcommit(wdata);
+	pnfs_set_layoutcommit(hdr);
 	dprintk("%s inode %lu pls_end_pos %lu\n", __func__, hdr->inode->i_ino,
 		(unsigned long) NFS_I(hdr->inode)->layout->plh_lwb);
 }
@@ -305,83 +295,82 @@ filelayout_reset_to_mds(struct pnfs_layout_segment *lseg)
  */
 static void filelayout_read_prepare(struct rpc_task *task, void *data)
 {
-	struct nfs_pgio_data *rdata = data;
+	struct nfs_pgio_header *hdr = data;
 
-	if (unlikely(test_bit(NFS_CONTEXT_BAD, &rdata->args.context->flags))) {
+	if (unlikely(test_bit(NFS_CONTEXT_BAD, &hdr->args.context->flags))) {
 		rpc_exit(task, -EIO);
 		return;
 	}
-	if (filelayout_reset_to_mds(rdata->header->lseg)) {
+	if (filelayout_reset_to_mds(hdr->lseg)) {
 		dprintk("%s task %u reset io to MDS\n", __func__, task->tk_pid);
-		filelayout_reset_read(rdata);
+		filelayout_reset_read(hdr);
 		rpc_exit(task, 0);
 		return;
 	}
-	rdata->pgio_done_cb = filelayout_read_done_cb;
+	hdr->pgio_done_cb = filelayout_read_done_cb;
 
-	if (nfs41_setup_sequence(rdata->ds_clp->cl_session,
-			&rdata->args.seq_args,
-			&rdata->res.seq_res,
+	if (nfs41_setup_sequence(hdr->ds_clp->cl_session,
+			&hdr->args.seq_args,
+			&hdr->res.seq_res,
 			task))
 		return;
-	if (nfs4_set_rw_stateid(&rdata->args.stateid, rdata->args.context,
-			rdata->args.lock_context, FMODE_READ) == -EIO)
+	if (nfs4_set_rw_stateid(&hdr->args.stateid, hdr->args.context,
+			hdr->args.lock_context, FMODE_READ) == -EIO)
 		rpc_exit(task, -EIO); /* lost lock, terminate I/O */
 }
 
 static void filelayout_read_call_done(struct rpc_task *task, void *data)
 {
-	struct nfs_pgio_data *rdata = data;
+	struct nfs_pgio_header *hdr = data;
 
 	dprintk("--> %s task->tk_status %d\n", __func__, task->tk_status);
 
-	if (test_bit(NFS_IOHDR_REDO, &rdata->header->flags) &&
+	if (test_bit(NFS_IOHDR_REDO, &hdr->flags) &&
 	    task->tk_status == 0) {
-		nfs41_sequence_done(task, &rdata->res.seq_res);
+		nfs41_sequence_done(task, &hdr->res.seq_res);
 		return;
 	}
 
 	/* Note this may cause RPC to be resent */
-	rdata->header->mds_ops->rpc_call_done(task, data);
+	hdr->mds_ops->rpc_call_done(task, data);
 }
 
 static void filelayout_read_count_stats(struct rpc_task *task, void *data)
 {
-	struct nfs_pgio_data *rdata = data;
+	struct nfs_pgio_header *hdr = data;
 
-	rpc_count_iostats(task, NFS_SERVER(rdata->header->inode)->client->cl_metrics);
+	rpc_count_iostats(task, NFS_SERVER(hdr->inode)->client->cl_metrics);
 }
 
 static void filelayout_read_release(void *data)
 {
-	struct nfs_pgio_data *rdata = data;
-	struct pnfs_layout_hdr *lo = rdata->header->lseg->pls_layout;
+	struct nfs_pgio_header *hdr = data;
+	struct pnfs_layout_hdr *lo = hdr->lseg->pls_layout;
 
 	filelayout_fenceme(lo->plh_inode, lo);
-	nfs_put_client(rdata->ds_clp);
-	rdata->header->mds_ops->rpc_release(data);
+	nfs_put_client(hdr->ds_clp);
+	hdr->mds_ops->rpc_release(data);
 }
 
 static int filelayout_write_done_cb(struct rpc_task *task,
-				struct nfs_pgio_data *data)
+				struct nfs_pgio_header *hdr)
 {
-	struct nfs_pgio_header *hdr = data->header;
 	int err;
 
-	trace_nfs4_pnfs_write(data, task->tk_status);
-	err = filelayout_async_handle_error(task, data->args.context->state,
-					    data->ds_clp, hdr->lseg);
+	trace_nfs4_pnfs_write(hdr, task->tk_status);
+	err = filelayout_async_handle_error(task, hdr->args.context->state,
+					    hdr->ds_clp, hdr->lseg);
 
 	switch (err) {
 	case -NFS4ERR_RESET_TO_MDS:
-		filelayout_reset_write(data);
+		filelayout_reset_write(hdr);
 		return task->tk_status;
 	case -EAGAIN:
 		rpc_restart_call_prepare(task);
 		return -EAGAIN;
 	}
 
-	filelayout_set_layoutcommit(data);
+	filelayout_set_layoutcommit(hdr);
 	return 0;
 }
 
@@ -419,57 +408,57 @@ static int filelayout_commit_done_cb(struct rpc_task *task,
 
 static void filelayout_write_prepare(struct rpc_task *task, void *data)
 {
-	struct nfs_pgio_data *wdata = data;
+	struct nfs_pgio_header *hdr = data;
 
-	if (unlikely(test_bit(NFS_CONTEXT_BAD, &wdata->args.context->flags))) {
+	if (unlikely(test_bit(NFS_CONTEXT_BAD, &hdr->args.context->flags))) {
 		rpc_exit(task, -EIO);
 		return;
 	}
-	if (filelayout_reset_to_mds(wdata->header->lseg)) {
+	if (filelayout_reset_to_mds(hdr->lseg)) {
 		dprintk("%s task %u reset io to MDS\n", __func__, task->tk_pid);
-		filelayout_reset_write(wdata);
+		filelayout_reset_write(hdr);
 		rpc_exit(task, 0);
 		return;
 	}
-	if (nfs41_setup_sequence(wdata->ds_clp->cl_session,
-			&wdata->args.seq_args,
-			&wdata->res.seq_res,
+	if (nfs41_setup_sequence(hdr->ds_clp->cl_session,
+			&hdr->args.seq_args,
+			&hdr->res.seq_res,
 			task))
 		return;
-	if (nfs4_set_rw_stateid(&wdata->args.stateid, wdata->args.context,
-			wdata->args.lock_context, FMODE_WRITE) == -EIO)
+	if (nfs4_set_rw_stateid(&hdr->args.stateid, hdr->args.context,
+			hdr->args.lock_context, FMODE_WRITE) == -EIO)
 		rpc_exit(task, -EIO); /* lost lock, terminate I/O */
 }
 
 static void filelayout_write_call_done(struct rpc_task *task, void *data)
 {
-	struct nfs_pgio_data *wdata = data;
+	struct nfs_pgio_header *hdr = data;
 
-	if (test_bit(NFS_IOHDR_REDO, &wdata->header->flags) &&
+	if (test_bit(NFS_IOHDR_REDO, &hdr->flags) &&
 	    task->tk_status == 0) {
-		nfs41_sequence_done(task, &wdata->res.seq_res);
+		nfs41_sequence_done(task, &hdr->res.seq_res);
 		return;
 	}
 
 	/* Note this may cause RPC to be resent */
-	wdata->header->mds_ops->rpc_call_done(task, data);
+	hdr->mds_ops->rpc_call_done(task, data);
 }
 
 static void filelayout_write_count_stats(struct rpc_task *task, void *data)
 {
-	struct nfs_pgio_data *wdata = data;
+	struct nfs_pgio_header *hdr = data;
 
-	rpc_count_iostats(task, NFS_SERVER(wdata->header->inode)->client->cl_metrics);
+	rpc_count_iostats(task, NFS_SERVER(hdr->inode)->client->cl_metrics);
 }
 
 static void filelayout_write_release(void *data)
 {
-	struct nfs_pgio_data *wdata = data;
-	struct pnfs_layout_hdr *lo = wdata->header->lseg->pls_layout;
+	struct nfs_pgio_header *hdr = data;
+	struct pnfs_layout_hdr *lo = hdr->lseg->pls_layout;
 
 	filelayout_fenceme(lo->plh_inode, lo);
-	nfs_put_client(wdata->ds_clp);
-	wdata->header->mds_ops->rpc_release(data);
+	nfs_put_client(hdr->ds_clp);
+	hdr->mds_ops->rpc_release(data);
 }
 
 static void filelayout_commit_prepare(struct rpc_task *task, void *data)
@@ -529,19 +518,18 @@ static const struct rpc_call_ops filelayout_commit_call_ops = {
 };
 
 static enum pnfs_try_status
-filelayout_read_pagelist(struct nfs_pgio_data *data)
+filelayout_read_pagelist(struct nfs_pgio_header *hdr)
 {
-	struct nfs_pgio_header *hdr = data->header;
 	struct pnfs_layout_segment *lseg = hdr->lseg;
 	struct nfs4_pnfs_ds *ds;
 	struct rpc_clnt *ds_clnt;
-	loff_t offset = data->args.offset;
+	loff_t offset = hdr->args.offset;
 	u32 j, idx;
 	struct nfs_fh *fh;
 
 	dprintk("--> %s ino %lu pgbase %u req %Zu@%llu\n",
 		__func__, hdr->inode->i_ino,
-		data->args.pgbase, (size_t)data->args.count, offset);
+		hdr->args.pgbase, (size_t)hdr->args.count, offset);
 
 	/* Retrieve the correct rpc_client for the byte range */
 	j = nfs4_fl_calc_j_index(lseg, offset);
@@ -559,30 +547,29 @@ filelayout_read_pagelist(struct nfs_pgio_data *data)
 
 	/* No multipath support. Use first DS */
 	atomic_inc(&ds->ds_clp->cl_count);
-	data->ds_clp = ds->ds_clp;
-	data->ds_idx = idx;
+	hdr->ds_clp = ds->ds_clp;
+	hdr->ds_idx = idx;
 	fh = nfs4_fl_select_ds_fh(lseg, j);
 	if (fh)
-		data->args.fh = fh;
+		hdr->args.fh = fh;
 
-	data->args.offset = filelayout_get_dserver_offset(lseg, offset);
-	data->mds_offset = offset;
+	hdr->args.offset = filelayout_get_dserver_offset(lseg, offset);
+	hdr->mds_offset = offset;
 
 	/* Perform an asynchronous read to ds */
-	nfs_initiate_pgio(ds_clnt, data,
+	nfs_initiate_pgio(ds_clnt, hdr,
 			    &filelayout_read_call_ops, 0, RPC_TASK_SOFTCONN);
 	return PNFS_ATTEMPTED;
 }
 
 /* Perform async writes. */
 static enum pnfs_try_status
-filelayout_write_pagelist(struct nfs_pgio_data *data, int sync)
+filelayout_write_pagelist(struct nfs_pgio_header *hdr, int sync)
 {
-	struct nfs_pgio_header *hdr = data->header;
 	struct pnfs_layout_segment *lseg = hdr->lseg;
 	struct nfs4_pnfs_ds *ds;
 	struct rpc_clnt *ds_clnt;
-	loff_t offset = data->args.offset;
+	loff_t offset = hdr->args.offset;
 	u32 j, idx;
 	struct nfs_fh *fh;
 
@@ -598,21 +585,20 @@ filelayout_write_pagelist(struct nfs_pgio_data *data, int sync)
 		return PNFS_NOT_ATTEMPTED;
 
 	dprintk("%s ino %lu sync %d req %Zu@%llu DS: %s cl_count %d\n",
-		__func__, hdr->inode->i_ino, sync, (size_t) data->args.count,
+		__func__, hdr->inode->i_ino, sync, (size_t) hdr->args.count,
 		offset, ds->ds_remotestr, atomic_read(&ds->ds_clp->cl_count));
 
-	data->pgio_done_cb = filelayout_write_done_cb;
+	hdr->pgio_done_cb = filelayout_write_done_cb;
 	atomic_inc(&ds->ds_clp->cl_count);
-	data->ds_clp = ds->ds_clp;
-	data->ds_idx = idx;
+	hdr->ds_clp = ds->ds_clp;
+	hdr->ds_idx = idx;
 	fh = nfs4_fl_select_ds_fh(lseg, j);
 	if (fh)
-		data->args.fh = fh;
-
-	data->args.offset = filelayout_get_dserver_offset(lseg, offset);
+		hdr->args.fh = fh;
+	hdr->args.offset = filelayout_get_dserver_offset(lseg, offset);
 
 	/* Perform an asynchronous write */
-	nfs_initiate_pgio(ds_clnt, data,
+	nfs_initiate_pgio(ds_clnt, hdr,
 				    &filelayout_write_call_ops, sync,
 				    RPC_TASK_SOFTCONN);
 	return PNFS_ATTEMPTED;
@@ -1023,6 +1009,7 @@ static u32 select_bucket_index(struct nfs4_filelayout_segment *fl, u32 j)
 
 /* The generic layer is about to remove the req from the commit list.
  * If this will make the bucket empty, it will need to put the lseg reference.
+ * Note this is must be called holding the inode (/cinfo) lock
  */
 static void
 filelayout_clear_request_commit(struct nfs_page *req,
@@ -1030,7 +1017,6 @@ filelayout_clear_request_commit(struct nfs_page *req,
 {
 	struct pnfs_layout_segment *freeme = NULL;
 
-	spin_lock(cinfo->lock);
 	if (!test_and_clear_bit(PG_COMMIT_TO_DS, &req->wb_flags))
 		goto out;
 	cinfo->ds->nwritten--;
@@ -1045,22 +1031,25 @@ filelayout_clear_request_commit(struct nfs_page *req,
 	}
 out:
 	nfs_request_remove_commit_list(req, cinfo);
-	spin_unlock(cinfo->lock);
-	pnfs_put_lseg(freeme);
+	pnfs_put_lseg_async(freeme);
 }
 
-static struct list_head *
-filelayout_choose_commit_list(struct nfs_page *req,
-			      struct pnfs_layout_segment *lseg,
-			      struct nfs_commit_info *cinfo)
+static void
+filelayout_mark_request_commit(struct nfs_page *req,
+			       struct pnfs_layout_segment *lseg,
+			       struct nfs_commit_info *cinfo)
+
 {
 	struct nfs4_filelayout_segment *fl = FILELAYOUT_LSEG(lseg);
 	u32 i, j;
 	struct list_head *list;
 	struct pnfs_commit_bucket *buckets;
 
-	if (fl->commit_through_mds)
-		return &cinfo->mds->list;
+	if (fl->commit_through_mds) {
+		list = &cinfo->mds->list;
+		spin_lock(cinfo->lock);
+		goto mds_commit;
+	}
 
 	/* Note that we are calling nfs4_fl_calc_j_index on each page
 	 * that ends up being committed to a data server.  An attractive
@@ -1084,19 +1073,22 @@ filelayout_choose_commit_list(struct nfs_page *req,
 	}
 	set_bit(PG_COMMIT_TO_DS, &req->wb_flags);
 	cinfo->ds->nwritten++;
-	spin_unlock(cinfo->lock);
-	return list;
-}
 
-static void
-filelayout_mark_request_commit(struct nfs_page *req,
-			       struct pnfs_layout_segment *lseg,
-			       struct nfs_commit_info *cinfo)
-{
-	struct list_head *list;
-
-	list = filelayout_choose_commit_list(req, lseg, cinfo);
-	nfs_request_add_commit_list(req, list, cinfo);
+mds_commit:
+	/* nfs_request_add_commit_list(). We need to add req to list without
+	 * dropping cinfo lock.
+	 */
+	set_bit(PG_CLEAN, &(req)->wb_flags);
+	nfs_list_add_request(req, list);
+	cinfo->mds->ncommit++;
+	spin_unlock(cinfo->lock);
+	if (!cinfo->dreq) {
+		inc_zone_page_state(req->wb_page, NR_UNSTABLE_NFS);
+		inc_bdi_stat(page_file_mapping(req->wb_page)->backing_dev_info,
+			     BDI_RECLAIMABLE);
+		__mark_inode_dirty(req->wb_context->dentry->d_inode,
+				   I_DIRTY_DATASYNC);
+	}
 }
 
 static u32 calc_ds_index_from_commit(struct pnfs_layout_segment *lseg, u32 i)
@@ -1244,15 +1236,63 @@ restart:
 	spin_unlock(cinfo->lock);
 }
 
+/* filelayout_search_commit_reqs - Search lists in @cinfo for the head reqest
+ *				   for @page
+ * @cinfo - commit info for current inode
+ * @page - page to search for matching head request
+ *
+ * Returns a the head request if one is found, otherwise returns NULL.
+ */
+static struct nfs_page *
+filelayout_search_commit_reqs(struct nfs_commit_info *cinfo, struct page *page)
+{
+	struct nfs_page *freq, *t;
+	struct pnfs_commit_bucket *b;
+	int i;
+
+	/* Linearly search the commit lists for each bucket until a matching
+	 * request is found */
+	for (i = 0, b = cinfo->ds->buckets; i < cinfo->ds->nbuckets; i++, b++) {
+		list_for_each_entry_safe(freq, t, &b->written, wb_list) {
+			if (freq->wb_page == page)
+				return freq->wb_head;
+		}
+		list_for_each_entry_safe(freq, t, &b->committing, wb_list) {
+			if (freq->wb_page == page)
+				return freq->wb_head;
+		}
+	}
+
+	return NULL;
+}
+
+static void filelayout_retry_commit(struct nfs_commit_info *cinfo, int idx)
+{
+	struct pnfs_ds_commit_info *fl_cinfo = cinfo->ds;
+	struct pnfs_commit_bucket *bucket = fl_cinfo->buckets;
+	struct pnfs_layout_segment *freeme;
+	int i;
+
+	for (i = idx; i < fl_cinfo->nbuckets; i++, bucket++) {
+		if (list_empty(&bucket->committing))
+			continue;
+		nfs_retry_commit(&bucket->committing, bucket->clseg, cinfo);
+		spin_lock(cinfo->lock);
+		freeme = bucket->clseg;
+		bucket->clseg = NULL;
+		spin_unlock(cinfo->lock);
+		pnfs_put_lseg(freeme);
+	}
+}
+
 static unsigned int
 alloc_ds_commits(struct nfs_commit_info *cinfo, struct list_head *list)
 {
 	struct pnfs_ds_commit_info *fl_cinfo;
 	struct pnfs_commit_bucket *bucket;
 	struct nfs_commit_data *data;
-	int i, j;
+	int i;
 	unsigned int nreq = 0;
-	struct pnfs_layout_segment *freeme;
 
 	fl_cinfo = cinfo->ds;
 	bucket = fl_cinfo->buckets;
@@ -1272,16 +1312,7 @@ alloc_ds_commits(struct nfs_commit_info *cinfo, struct list_head *list)
 	}
 
 	/* Clean up on error */
-	for (j = i; j < fl_cinfo->nbuckets; j++, bucket++) {
-		if (list_empty(&bucket->committing))
-			continue;
-		nfs_retry_commit(&bucket->committing, bucket->clseg, cinfo);
-		spin_lock(cinfo->lock);
-		freeme = bucket->clseg;
-		bucket->clseg = NULL;
-		spin_unlock(cinfo->lock);
-		pnfs_put_lseg(freeme);
-	}
+	filelayout_retry_commit(cinfo, i);
 	/* Caller will clean up entries put on list */
 	return nreq;
 }
@@ -1301,8 +1332,12 @@ filelayout_commit_pagelist(struct inode *inode, struct list_head *mds_pages,
 			data->lseg = NULL;
 			list_add(&data->pages, &list);
 			nreq++;
-		} else
+		} else {
 			nfs_retry_commit(mds_pages, NULL, cinfo);
+			filelayout_retry_commit(cinfo, 0);
+			cinfo->completion_ops->error_cleanup(NFS_I(inode));
+			return -ENOMEM;
+		}
 	}
 
 	nreq += alloc_ds_commits(cinfo, &list);
@@ -1380,6 +1415,7 @@ static struct pnfs_layoutdriver_type filelayout_type = {
 	.clear_request_commit	= filelayout_clear_request_commit,
 	.scan_commit_lists	= filelayout_scan_commit_lists,
 	.recover_commit_reqs	= filelayout_recover_commit_reqs,
+	.search_commit_reqs	= filelayout_search_commit_reqs,
 	.commit_pagelist	= filelayout_commit_pagelist,
 	.read_pagelist		= filelayout_read_pagelist,
 	.write_pagelist		= filelayout_write_pagelist,
diff --git a/fs/nfs/filelayout/filelayoutdev.c b/fs/nfs/filelayout/filelayoutdev.c
index 44bf0140a4c7..8540516f4d71 100644
--- a/fs/nfs/filelayout/filelayoutdev.c
+++ b/fs/nfs/filelayout/filelayoutdev.c
@@ -695,7 +695,7 @@ filelayout_get_device_info(struct inode *inode,
 	if (pdev == NULL)
 		return NULL;
 
-	pages = kzalloc(max_pages * sizeof(struct page *), gfp_flags);
+	pages = kcalloc(max_pages, sizeof(struct page *), gfp_flags);
 	if (pages == NULL) {
 		kfree(pdev);
 		return NULL;
@@ -783,8 +783,8 @@ nfs4_fl_select_ds_fh(struct pnfs_layout_segment *lseg, u32 j)
 static void nfs4_wait_ds_connect(struct nfs4_pnfs_ds *ds)
 {
 	might_sleep();
-	wait_on_bit(&ds->ds_state, NFS4DS_CONNECTING,
-			nfs_wait_bit_killable, TASK_KILLABLE);
+	wait_on_bit_action(&ds->ds_state, NFS4DS_CONNECTING,
+			   nfs_wait_bit_killable, TASK_KILLABLE);
 }
 
 static void nfs4_clear_ds_conn_bit(struct nfs4_pnfs_ds *ds)
diff --git a/fs/nfs/getroot.c b/fs/nfs/getroot.c
index b94f80420a58..880618a8b048 100644
--- a/fs/nfs/getroot.c
+++ b/fs/nfs/getroot.c
@@ -112,7 +112,7 @@ struct dentry *nfs_get_root(struct super_block *sb, struct nfs_fh *mntfh,
 	 * if the dentry tree reaches them; however if the dentry already
 	 * exists, we'll pick it up at this point and use it as the root
 	 */
-	ret = d_obtain_alias(inode);
+	ret = d_obtain_root(inode);
 	if (IS_ERR(ret)) {
 		dprintk("nfs_get_root: get root dentry failed\n");
 		goto out;
diff --git a/fs/nfs/idmap.c b/fs/nfs/idmap.c
index 567983d2c0eb..7dd55b745c4d 100644
--- a/fs/nfs/idmap.c
+++ b/fs/nfs/idmap.c
@@ -174,7 +174,9 @@ static int nfs_map_numeric_to_string(__u32 id, char *buf, size_t buflen)
 
 static struct key_type key_type_id_resolver = {
 	.name		= "id_resolver",
-	.instantiate	= user_instantiate,
+	.preparse	= user_preparse,
+	.free_preparse	= user_free_preparse,
+	.instantiate	= generic_key_instantiate,
 	.match		= user_match,
 	.revoke		= user_revoke,
 	.destroy	= user_destroy,
@@ -282,6 +284,8 @@ static struct key *nfs_idmap_request_key(const char *name, size_t namelen,
 						desc, "", 0, idmap);
 		mutex_unlock(&idmap->idmap_mutex);
 	}
+	if (!IS_ERR(rkey))
+		set_bit(KEY_FLAG_ROOT_CAN_INVAL, &rkey->flags);
 
 	kfree(desc);
 	return rkey;
@@ -394,7 +398,9 @@ static const struct rpc_pipe_ops idmap_upcall_ops = {
 
 static struct key_type key_type_id_resolver_legacy = {
 	.name		= "id_legacy",
-	.instantiate	= user_instantiate,
+	.preparse	= user_preparse,
+	.free_preparse	= user_free_preparse,
+	.instantiate	= generic_key_instantiate,
 	.match		= user_match,
 	.revoke		= user_revoke,
 	.destroy	= user_destroy,
diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c
index c496f8a74639..577a36f0a510 100644
--- a/fs/nfs/inode.c
+++ b/fs/nfs/inode.c
@@ -75,7 +75,7 @@ nfs_fattr_to_ino_t(struct nfs_fattr *fattr)
  * nfs_wait_bit_killable - helper for functions that are sleeping on bit locks
  * @word: long word containing the bit lock
  */
-int nfs_wait_bit_killable(void *word)
+int nfs_wait_bit_killable(struct wait_bit_key *key)
 {
 	if (fatal_signal_pending(current))
 		return -ERESTARTSYS;
@@ -147,6 +147,17 @@ int nfs_sync_mapping(struct address_space *mapping)
 	return ret;
 }
 
+static void nfs_set_cache_invalid(struct inode *inode, unsigned long flags)
+{
+	struct nfs_inode *nfsi = NFS_I(inode);
+
+	if (inode->i_mapping->nrpages == 0)
+		flags &= ~NFS_INO_INVALID_DATA;
+	nfsi->cache_validity |= flags;
+	if (flags & NFS_INO_INVALID_DATA)
+		nfs_fscache_invalidate(inode);
+}
+
 /*
  * Invalidate the local caches
  */
@@ -162,17 +173,16 @@ static void nfs_zap_caches_locked(struct inode *inode)
 
 	memset(NFS_I(inode)->cookieverf, 0, sizeof(NFS_I(inode)->cookieverf));
 	if (S_ISREG(mode) || S_ISDIR(mode) || S_ISLNK(mode)) {
-		nfs_fscache_invalidate(inode);
-		nfsi->cache_validity |= NFS_INO_INVALID_ATTR
+		nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATTR
 					| NFS_INO_INVALID_DATA
 					| NFS_INO_INVALID_ACCESS
 					| NFS_INO_INVALID_ACL
-					| NFS_INO_REVAL_PAGECACHE;
+					| NFS_INO_REVAL_PAGECACHE);
 	} else
-		nfsi->cache_validity |= NFS_INO_INVALID_ATTR
+		nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATTR
 					| NFS_INO_INVALID_ACCESS
 					| NFS_INO_INVALID_ACL
-					| NFS_INO_REVAL_PAGECACHE;
+					| NFS_INO_REVAL_PAGECACHE);
 	nfs_zap_label_cache_locked(nfsi);
 }
 
@@ -187,8 +197,7 @@ void nfs_zap_mapping(struct inode *inode, struct address_space *mapping)
 {
 	if (mapping->nrpages != 0) {
 		spin_lock(&inode->i_lock);
-		NFS_I(inode)->cache_validity |= NFS_INO_INVALID_DATA;
-		nfs_fscache_invalidate(inode);
+		nfs_set_cache_invalid(inode, NFS_INO_INVALID_DATA);
 		spin_unlock(&inode->i_lock);
 	}
 }
@@ -209,7 +218,7 @@ EXPORT_SYMBOL_GPL(nfs_zap_acl_cache);
 void nfs_invalidate_atime(struct inode *inode)
 {
 	spin_lock(&inode->i_lock);
-	NFS_I(inode)->cache_validity |= NFS_INO_INVALID_ATIME;
+	nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATIME);
 	spin_unlock(&inode->i_lock);
 }
 EXPORT_SYMBOL_GPL(nfs_invalidate_atime);
@@ -369,7 +378,7 @@ nfs_fhget(struct super_block *sb, struct nfs_fh *fh, struct nfs_fattr *fattr, st
 		inode->i_mode = fattr->mode;
 		if ((fattr->valid & NFS_ATTR_FATTR_MODE) == 0
 				&& nfs_server_capable(inode, NFS_CAP_MODE))
-			nfsi->cache_validity |= NFS_INO_INVALID_ATTR;
+			nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATTR);
 		/* Why so? Because we want revalidate for devices/FIFOs, and
 		 * that's precisely what we have in nfs_file_inode_operations.
 		 */
@@ -415,36 +424,36 @@ nfs_fhget(struct super_block *sb, struct nfs_fh *fh, struct nfs_fattr *fattr, st
 		if (fattr->valid & NFS_ATTR_FATTR_ATIME)
 			inode->i_atime = fattr->atime;
 		else if (nfs_server_capable(inode, NFS_CAP_ATIME))
-			nfsi->cache_validity |= NFS_INO_INVALID_ATTR;
+			nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATTR);
 		if (fattr->valid & NFS_ATTR_FATTR_MTIME)
 			inode->i_mtime = fattr->mtime;
 		else if (nfs_server_capable(inode, NFS_CAP_MTIME))
-			nfsi->cache_validity |= NFS_INO_INVALID_ATTR;
+			nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATTR);
 		if (fattr->valid & NFS_ATTR_FATTR_CTIME)
 			inode->i_ctime = fattr->ctime;
 		else if (nfs_server_capable(inode, NFS_CAP_CTIME))
-			nfsi->cache_validity |= NFS_INO_INVALID_ATTR;
+			nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATTR);
 		if (fattr->valid & NFS_ATTR_FATTR_CHANGE)
 			inode->i_version = fattr->change_attr;
 		else if (nfs_server_capable(inode, NFS_CAP_CHANGE_ATTR))
-			nfsi->cache_validity |= NFS_INO_INVALID_ATTR;
+			nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATTR);
 		if (fattr->valid & NFS_ATTR_FATTR_SIZE)
 			inode->i_size = nfs_size_to_loff_t(fattr->size);
 		else
-			nfsi->cache_validity |= NFS_INO_INVALID_ATTR
-				| NFS_INO_REVAL_PAGECACHE;
+			nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATTR
+				| NFS_INO_REVAL_PAGECACHE);
 		if (fattr->valid & NFS_ATTR_FATTR_NLINK)
 			set_nlink(inode, fattr->nlink);
 		else if (nfs_server_capable(inode, NFS_CAP_NLINK))
-			nfsi->cache_validity |= NFS_INO_INVALID_ATTR;
+			nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATTR);
 		if (fattr->valid & NFS_ATTR_FATTR_OWNER)
 			inode->i_uid = fattr->uid;
 		else if (nfs_server_capable(inode, NFS_CAP_OWNER))
-			nfsi->cache_validity |= NFS_INO_INVALID_ATTR;
+			nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATTR);
 		if (fattr->valid & NFS_ATTR_FATTR_GROUP)
 			inode->i_gid = fattr->gid;
 		else if (nfs_server_capable(inode, NFS_CAP_OWNER_GROUP))
-			nfsi->cache_validity |= NFS_INO_INVALID_ATTR;
+			nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATTR);
 		if (fattr->valid & NFS_ATTR_FATTR_BLOCKS_USED)
 			inode->i_blocks = fattr->du.nfs2.blocks;
 		if (fattr->valid & NFS_ATTR_FATTR_SPACE_USED) {
@@ -550,6 +559,9 @@ static int nfs_vmtruncate(struct inode * inode, loff_t offset)
 
 	spin_lock(&inode->i_lock);
 	i_size_write(inode, offset);
+	/* Optimisation */
+	if (offset == 0)
+		NFS_I(inode)->cache_validity &= ~NFS_INO_INVALID_DATA;
 	spin_unlock(&inode->i_lock);
 
 	truncate_pagecache(inode, offset);
@@ -578,7 +590,8 @@ void nfs_setattr_update_inode(struct inode *inode, struct iattr *attr)
 			inode->i_uid = attr->ia_uid;
 		if ((attr->ia_valid & ATTR_GID) != 0)
 			inode->i_gid = attr->ia_gid;
-		NFS_I(inode)->cache_validity |= NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL;
+		nfs_set_cache_invalid(inode, NFS_INO_INVALID_ACCESS
+				| NFS_INO_INVALID_ACL);
 		spin_unlock(&inode->i_lock);
 	}
 	if ((attr->ia_valid & ATTR_SIZE) != 0) {
@@ -989,6 +1002,15 @@ int nfs_revalidate_inode(struct nfs_server *server, struct inode *inode)
 }
 EXPORT_SYMBOL_GPL(nfs_revalidate_inode);
 
+int nfs_revalidate_inode_rcu(struct nfs_server *server, struct inode *inode)
+{
+	if (!(NFS_I(inode)->cache_validity &
+			(NFS_INO_INVALID_ATTR|NFS_INO_INVALID_LABEL))
+			&& !nfs_attribute_cache_expired(inode))
+		return NFS_STALE(inode) ? -ESTALE : 0;
+	return -ECHILD;
+}
+
 static int nfs_invalidate_mapping(struct inode *inode, struct address_space *mapping)
 {
 	struct nfs_inode *nfsi = NFS_I(inode);
@@ -1061,8 +1083,8 @@ int nfs_revalidate_mapping(struct inode *inode, struct address_space *mapping)
 	 * the bit lock here if it looks like we're going to be doing that.
 	 */
 	for (;;) {
-		ret = wait_on_bit(bitlock, NFS_INO_INVALIDATING,
-				  nfs_wait_bit_killable, TASK_KILLABLE);
+		ret = wait_on_bit_action(bitlock, NFS_INO_INVALIDATING,
+					 nfs_wait_bit_killable, TASK_KILLABLE);
 		if (ret)
 			goto out;
 		spin_lock(&inode->i_lock);
@@ -1101,7 +1123,7 @@ static unsigned long nfs_wcc_update_inode(struct inode *inode, struct nfs_fattr
 			&& inode->i_version == fattr->pre_change_attr) {
 		inode->i_version = fattr->change_attr;
 		if (S_ISDIR(inode->i_mode))
-			nfsi->cache_validity |= NFS_INO_INVALID_DATA;
+			nfs_set_cache_invalid(inode, NFS_INO_INVALID_DATA);
 		ret |= NFS_INO_INVALID_ATTR;
 	}
 	/* If we have atomic WCC data, we may update some attributes */
@@ -1117,7 +1139,7 @@ static unsigned long nfs_wcc_update_inode(struct inode *inode, struct nfs_fattr
 			&& timespec_equal(&inode->i_mtime, &fattr->pre_mtime)) {
 		memcpy(&inode->i_mtime, &fattr->mtime, sizeof(inode->i_mtime));
 		if (S_ISDIR(inode->i_mode))
-			nfsi->cache_validity |= NFS_INO_INVALID_DATA;
+			nfs_set_cache_invalid(inode, NFS_INO_INVALID_DATA);
 		ret |= NFS_INO_INVALID_ATTR;
 	}
 	if ((fattr->valid & NFS_ATTR_FATTR_PRESIZE)
@@ -1128,9 +1150,6 @@ static unsigned long nfs_wcc_update_inode(struct inode *inode, struct nfs_fattr
 		ret |= NFS_INO_INVALID_ATTR;
 	}
 
-	if (nfsi->cache_validity & NFS_INO_INVALID_DATA)
-		nfs_fscache_invalidate(inode);
-
 	return ret;
 }
 
@@ -1189,7 +1208,7 @@ static int nfs_check_inode_attributes(struct inode *inode, struct nfs_fattr *fat
 		invalid |= NFS_INO_INVALID_ATIME;
 
 	if (invalid != 0)
-		nfsi->cache_validity |= invalid;
+		nfs_set_cache_invalid(inode, invalid);
 
 	nfsi->read_cache_jiffies = fattr->time_start;
 	return 0;
@@ -1402,13 +1421,11 @@ EXPORT_SYMBOL_GPL(nfs_refresh_inode);
 
 static int nfs_post_op_update_inode_locked(struct inode *inode, struct nfs_fattr *fattr)
 {
-	struct nfs_inode *nfsi = NFS_I(inode);
+	unsigned long invalid = NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE;
 
-	nfsi->cache_validity |= NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE;
-	if (S_ISDIR(inode->i_mode)) {
-		nfsi->cache_validity |= NFS_INO_INVALID_DATA;
-		nfs_fscache_invalidate(inode);
-	}
+	if (S_ISDIR(inode->i_mode))
+		invalid |= NFS_INO_INVALID_DATA;
+	nfs_set_cache_invalid(inode, invalid);
 	if ((fattr->valid & NFS_ATTR_FATTR) == 0)
 		return 0;
 	return nfs_refresh_inode_locked(inode, fattr);
@@ -1601,6 +1618,7 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr)
 			if ((nfsi->npages == 0) || new_isize > cur_isize) {
 				i_size_write(inode, new_isize);
 				invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA;
+				invalid &= ~NFS_INO_REVAL_PAGECACHE;
 			}
 			dprintk("NFS: isize change on server for file %s/%ld "
 					"(%Ld to %Ld)\n",
@@ -1702,10 +1720,7 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr)
 		invalid &= ~NFS_INO_INVALID_DATA;
 	if (!NFS_PROTO(inode)->have_delegation(inode, FMODE_READ) ||
 			(save_cache_validity & NFS_INO_REVAL_FORCED))
-		nfsi->cache_validity |= invalid;
-
-	if (invalid & NFS_INO_INVALID_DATA)
-		nfs_fscache_invalidate(inode);
+		nfs_set_cache_invalid(inode, invalid);
 
 	return 0;
  out_err:
@@ -1834,11 +1849,12 @@ EXPORT_SYMBOL_GPL(nfs_net_id);
 static int nfs_net_init(struct net *net)
 {
 	nfs_clients_init(net);
-	return 0;
+	return nfs_fs_proc_net_init(net);
 }
 
 static void nfs_net_exit(struct net *net)
 {
+	nfs_fs_proc_net_exit(net);
 	nfs_cleanup_cb_ident_idr(net);
 }
 
diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h
index 82ddbf46660e..9056622d2230 100644
--- a/fs/nfs/internal.h
+++ b/fs/nfs/internal.h
@@ -195,7 +195,16 @@ extern struct rpc_clnt *nfs4_find_or_create_ds_client(struct nfs_client *,
 #ifdef CONFIG_PROC_FS
 extern int __init nfs_fs_proc_init(void);
 extern void nfs_fs_proc_exit(void);
+extern int nfs_fs_proc_net_init(struct net *net);
+extern void nfs_fs_proc_net_exit(struct net *net);
 #else
+static inline int nfs_fs_proc_net_init(struct net *net)
+{
+	return 0;
+}
+static inline void nfs_fs_proc_net_exit(struct net *net)
+{
+}
 static inline int nfs_fs_proc_init(void)
 {
 	return 0;
@@ -238,12 +247,13 @@ void nfs_set_pgio_error(struct nfs_pgio_header *hdr, int error, loff_t pos);
 int nfs_iocounter_wait(struct nfs_io_counter *c);
 
 extern const struct nfs_pageio_ops nfs_pgio_rw_ops;
-struct nfs_rw_header *nfs_rw_header_alloc(const struct nfs_rw_ops *);
-void nfs_rw_header_free(struct nfs_pgio_header *);
-void nfs_pgio_data_release(struct nfs_pgio_data *);
+struct nfs_pgio_header *nfs_pgio_header_alloc(const struct nfs_rw_ops *);
+void nfs_pgio_header_free(struct nfs_pgio_header *);
+void nfs_pgio_data_destroy(struct nfs_pgio_header *);
 int nfs_generic_pgio(struct nfs_pageio_descriptor *, struct nfs_pgio_header *);
-int nfs_initiate_pgio(struct rpc_clnt *, struct nfs_pgio_data *,
+int nfs_initiate_pgio(struct rpc_clnt *, struct nfs_pgio_header *,
 		      const struct rpc_call_ops *, int, int);
+void nfs_free_request(struct nfs_page *req);
 
 static inline void nfs_iocounter_init(struct nfs_io_counter *c)
 {
@@ -347,7 +357,7 @@ extern int nfs_drop_inode(struct inode *);
 extern void nfs_clear_inode(struct inode *);
 extern void nfs_evict_inode(struct inode *);
 void nfs_zap_acl_cache(struct inode *inode);
-extern int nfs_wait_bit_killable(void *word);
+extern int nfs_wait_bit_killable(struct wait_bit_key *key);
 
 /* super.c */
 extern const struct super_operations nfs_sops;
@@ -441,6 +451,7 @@ int nfs_scan_commit(struct inode *inode, struct list_head *dst,
 void nfs_mark_request_commit(struct nfs_page *req,
 			     struct pnfs_layout_segment *lseg,
 			     struct nfs_commit_info *cinfo);
+int nfs_write_need_commit(struct nfs_pgio_header *);
 int nfs_generic_commit_list(struct inode *inode, struct list_head *head,
 			    int how, struct nfs_commit_info *cinfo);
 void nfs_retry_commit(struct list_head *page_list,
@@ -481,7 +492,7 @@ static inline void nfs_inode_dio_wait(struct inode *inode)
 extern ssize_t nfs_dreq_bytes_left(struct nfs_direct_req *dreq);
 
 /* nfs4proc.c */
-extern void __nfs4_read_done_cb(struct nfs_pgio_data *);
+extern void __nfs4_read_done_cb(struct nfs_pgio_header *);
 extern struct nfs_client *nfs4_init_client(struct nfs_client *clp,
 			    const struct rpc_timeout *timeparms,
 			    const char *ip_addr);
diff --git a/fs/nfs/netns.h b/fs/nfs/netns.h
index 8ee1fab83268..ef221fb8a183 100644
--- a/fs/nfs/netns.h
+++ b/fs/nfs/netns.h
@@ -29,6 +29,9 @@ struct nfs_net {
 #endif
 	spinlock_t nfs_client_lock;
 	struct timespec boot_time;
+#ifdef CONFIG_PROC_FS
+	struct proc_dir_entry *proc_nfsfs;
+#endif
 };
 
 extern int nfs_net_id;
diff --git a/fs/nfs/nfs3acl.c b/fs/nfs/nfs3acl.c
index 871d6eda8dba..d0fec260132a 100644
--- a/fs/nfs/nfs3acl.c
+++ b/fs/nfs/nfs3acl.c
@@ -247,3 +247,46 @@ const struct xattr_handler *nfs3_xattr_handlers[] = {
 	&posix_acl_default_xattr_handler,
 	NULL,
 };
+
+static int
+nfs3_list_one_acl(struct inode *inode, int type, const char *name, void *data,
+		size_t size, ssize_t *result)
+{
+	struct posix_acl *acl;
+	char *p = data + *result;
+
+	acl = get_acl(inode, type);
+	if (IS_ERR_OR_NULL(acl))
+		return 0;
+
+	posix_acl_release(acl);
+
+	*result += strlen(name);
+	*result += 1;
+	if (!size)
+		return 0;
+	if (*result > size)
+		return -ERANGE;
+
+	strcpy(p, name);
+	return 0;
+}
+
+ssize_t
+nfs3_listxattr(struct dentry *dentry, char *data, size_t size)
+{
+	struct inode *inode = dentry->d_inode;
+	ssize_t result = 0;
+	int error;
+
+	error = nfs3_list_one_acl(inode, ACL_TYPE_ACCESS,
+			POSIX_ACL_XATTR_ACCESS, data, size, &result);
+	if (error)
+		return error;
+
+	error = nfs3_list_one_acl(inode, ACL_TYPE_DEFAULT,
+			POSIX_ACL_XATTR_DEFAULT, data, size, &result);
+	if (error)
+		return error;
+	return result;
+}
diff --git a/fs/nfs/nfs3proc.c b/fs/nfs/nfs3proc.c
index e7daa42bbc86..809670eba52a 100644
--- a/fs/nfs/nfs3proc.c
+++ b/fs/nfs/nfs3proc.c
@@ -795,41 +795,44 @@ nfs3_proc_pathconf(struct nfs_server *server, struct nfs_fh *fhandle,
 	return status;
 }
 
-static int nfs3_read_done(struct rpc_task *task, struct nfs_pgio_data *data)
+static int nfs3_read_done(struct rpc_task *task, struct nfs_pgio_header *hdr)
 {
-	struct inode *inode = data->header->inode;
+	struct inode *inode = hdr->inode;
 
 	if (nfs3_async_handle_jukebox(task, inode))
 		return -EAGAIN;
 
 	nfs_invalidate_atime(inode);
-	nfs_refresh_inode(inode, &data->fattr);
+	nfs_refresh_inode(inode, &hdr->fattr);
 	return 0;
 }
 
-static void nfs3_proc_read_setup(struct nfs_pgio_data *data, struct rpc_message *msg)
+static void nfs3_proc_read_setup(struct nfs_pgio_header *hdr,
+				 struct rpc_message *msg)
 {
 	msg->rpc_proc = &nfs3_procedures[NFS3PROC_READ];
 }
 
-static int nfs3_proc_pgio_rpc_prepare(struct rpc_task *task, struct nfs_pgio_data *data)
+static int nfs3_proc_pgio_rpc_prepare(struct rpc_task *task,
+				      struct nfs_pgio_header *hdr)
 {
 	rpc_call_start(task);
 	return 0;
 }
 
-static int nfs3_write_done(struct rpc_task *task, struct nfs_pgio_data *data)
+static int nfs3_write_done(struct rpc_task *task, struct nfs_pgio_header *hdr)
 {
-	struct inode *inode = data->header->inode;
+	struct inode *inode = hdr->inode;
 
 	if (nfs3_async_handle_jukebox(task, inode))
 		return -EAGAIN;
 	if (task->tk_status >= 0)
-		nfs_post_op_update_inode_force_wcc(inode, data->res.fattr);
+		nfs_post_op_update_inode_force_wcc(inode, hdr->res.fattr);
 	return 0;
 }
 
-static void nfs3_proc_write_setup(struct nfs_pgio_data *data, struct rpc_message *msg)
+static void nfs3_proc_write_setup(struct nfs_pgio_header *hdr,
+				  struct rpc_message *msg)
 {
 	msg->rpc_proc = &nfs3_procedures[NFS3PROC_WRITE];
 }
@@ -885,7 +888,7 @@ static const struct inode_operations nfs3_dir_inode_operations = {
 	.getattr	= nfs_getattr,
 	.setattr	= nfs_setattr,
 #ifdef CONFIG_NFS_V3_ACL
-	.listxattr	= generic_listxattr,
+	.listxattr	= nfs3_listxattr,
 	.getxattr	= generic_getxattr,
 	.setxattr	= generic_setxattr,
 	.removexattr	= generic_removexattr,
@@ -899,7 +902,7 @@ static const struct inode_operations nfs3_file_inode_operations = {
 	.getattr	= nfs_getattr,
 	.setattr	= nfs_setattr,
 #ifdef CONFIG_NFS_V3_ACL
-	.listxattr	= generic_listxattr,
+	.listxattr	= nfs3_listxattr,
 	.getxattr	= generic_getxattr,
 	.setxattr	= generic_setxattr,
 	.removexattr	= generic_removexattr,
diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h
index f63cb87cd730..92193eddb41d 100644
--- a/fs/nfs/nfs4_fs.h
+++ b/fs/nfs/nfs4_fs.h
@@ -54,7 +54,7 @@ struct nfs4_minor_version_ops {
 			const nfs4_stateid *);
 	int	(*find_root_sec)(struct nfs_server *, struct nfs_fh *,
 			struct nfs_fsinfo *);
-	int	(*free_lock_state)(struct nfs_server *,
+	void	(*free_lock_state)(struct nfs_server *,
 			struct nfs4_lock_state *);
 	const struct rpc_call_ops *call_sync_ops;
 	const struct nfs4_state_recovery_ops *reboot_recovery_ops;
@@ -129,27 +129,17 @@ enum {
  * LOCK: one nfs4_state (LOCK) to hold the lock stateid nfs4_state(OPEN)
  */
 
-struct nfs4_lock_owner {
-	unsigned int lo_type;
-#define NFS4_ANY_LOCK_TYPE	(0U)
-#define NFS4_FLOCK_LOCK_TYPE	(1U << 0)
-#define NFS4_POSIX_LOCK_TYPE	(1U << 1)
-	union {
-		fl_owner_t posix_owner;
-		pid_t flock_owner;
-	} lo_u;
-};
-
 struct nfs4_lock_state {
-	struct list_head	ls_locks;	/* Other lock stateids */
-	struct nfs4_state *	ls_state;	/* Pointer to open state */
+	struct list_head		ls_locks;   /* Other lock stateids */
+	struct nfs4_state *		ls_state;   /* Pointer to open state */
 #define NFS_LOCK_INITIALIZED 0
 #define NFS_LOCK_LOST        1
-	unsigned long		ls_flags;
+	unsigned long			ls_flags;
 	struct nfs_seqid_counter	ls_seqid;
-	nfs4_stateid		ls_stateid;
-	atomic_t		ls_count;
-	struct nfs4_lock_owner	ls_owner;
+	nfs4_stateid			ls_stateid;
+	atomic_t			ls_count;
+	fl_owner_t			ls_owner;
+	struct work_struct		ls_release;
 };
 
 /* bits for nfs4_state->flags */
@@ -230,7 +220,7 @@ int nfs_atomic_open(struct inode *, struct dentry *, struct file *,
 extern struct file_system_type nfs4_fs_type;
 
 /* nfs4namespace.c */
-struct rpc_clnt *nfs4_create_sec_client(struct rpc_clnt *, struct inode *, struct qstr *);
+struct rpc_clnt *nfs4_negotiate_security(struct rpc_clnt *, struct inode *, struct qstr *);
 struct vfsmount *nfs4_submount(struct nfs_server *, struct dentry *,
 			       struct nfs_fh *, struct nfs_fattr *);
 int nfs4_replace_transport(struct nfs_server *server,
@@ -337,11 +327,11 @@ nfs4_state_protect(struct nfs_client *clp, unsigned long sp4_mode,
  */
 static inline void
 nfs4_state_protect_write(struct nfs_client *clp, struct rpc_clnt **clntp,
-			 struct rpc_message *msg, struct nfs_pgio_data *wdata)
+			 struct rpc_message *msg, struct nfs_pgio_header *hdr)
 {
 	if (_nfs4_state_protect(clp, NFS_SP4_MACH_CRED_WRITE, clntp, msg) &&
 	    !test_bit(NFS_SP4_MACH_CRED_COMMIT, &clp->cl_sp4_flags))
-		wdata->args.stable = NFS_FILE_SYNC;
+		hdr->args.stable = NFS_FILE_SYNC;
 }
 #else /* CONFIG_NFS_v4_1 */
 static inline struct nfs4_session *nfs4_get_session(const struct nfs_server *server)
@@ -369,7 +359,7 @@ nfs4_state_protect(struct nfs_client *clp, unsigned long sp4_flags,
 
 static inline void
 nfs4_state_protect_write(struct nfs_client *clp, struct rpc_clnt **clntp,
-			 struct rpc_message *msg, struct nfs_pgio_data *wdata)
+			 struct rpc_message *msg, struct nfs_pgio_header *hdr)
 {
 }
 #endif /* CONFIG_NFS_V4_1 */
diff --git a/fs/nfs/nfs4client.c b/fs/nfs/nfs4client.c
index aa9ef4876046..53e435a95260 100644
--- a/fs/nfs/nfs4client.c
+++ b/fs/nfs/nfs4client.c
@@ -855,6 +855,11 @@ struct nfs_client *nfs4_set_ds_client(struct nfs_client* mds_clp,
 	};
 	struct rpc_timeout ds_timeout;
 	struct nfs_client *clp;
+	char buf[INET6_ADDRSTRLEN + 1];
+
+	if (rpc_ntop(ds_addr, buf, sizeof(buf)) <= 0)
+		return ERR_PTR(-EINVAL);
+	cl_init.hostname = buf;
 
 	/*
 	 * Set an authflavor equual to the MDS value. Use the MDS nfs_client
diff --git a/fs/nfs/nfs4namespace.c b/fs/nfs/nfs4namespace.c
index 3d5dbf80d46a..3d83cb1fdc70 100644
--- a/fs/nfs/nfs4namespace.c
+++ b/fs/nfs/nfs4namespace.c
@@ -139,16 +139,22 @@ static size_t nfs_parse_server_name(char *string, size_t len,
  * @server: NFS server struct
  * @flavors: List of security tuples returned by SECINFO procedure
  *
- * Return the pseudoflavor of the first security mechanism in
- * "flavors" that is locally supported.  Return RPC_AUTH_UNIX if
- * no matching flavor is found in the array.  The "flavors" array
+ * Return an rpc client that uses the first security mechanism in
+ * "flavors" that is locally supported.  The "flavors" array
  * is searched in the order returned from the server, per RFC 3530
- * recommendation.
+ * recommendation and each flavor is checked for membership in the
+ * sec= mount option list if it exists.
+ *
+ * Return -EPERM if no matching flavor is found in the array.
+ *
+ * Please call rpc_shutdown_client() when you are done with this rpc client.
+ *
  */
-static rpc_authflavor_t nfs_find_best_sec(struct nfs_server *server,
+static struct rpc_clnt *nfs_find_best_sec(struct rpc_clnt *clnt,
+					  struct nfs_server *server,
 					  struct nfs4_secinfo_flavors *flavors)
 {
-	rpc_authflavor_t pseudoflavor;
+	rpc_authflavor_t pflavor;
 	struct nfs4_secinfo4 *secinfo;
 	unsigned int i;
 
@@ -159,62 +165,73 @@ static rpc_authflavor_t nfs_find_best_sec(struct nfs_server *server,
 		case RPC_AUTH_NULL:
 		case RPC_AUTH_UNIX:
 		case RPC_AUTH_GSS:
-			pseudoflavor = rpcauth_get_pseudoflavor(secinfo->flavor,
+			pflavor = rpcauth_get_pseudoflavor(secinfo->flavor,
 							&secinfo->flavor_info);
-			/* make sure pseudoflavor matches sec= mount opt */
-			if (pseudoflavor != RPC_AUTH_MAXFLAVOR &&
-			    nfs_auth_info_match(&server->auth_info,
-						pseudoflavor))
-				return pseudoflavor;
-			break;
+			/* does the pseudoflavor match a sec= mount opt? */
+			if (pflavor != RPC_AUTH_MAXFLAVOR &&
+			    nfs_auth_info_match(&server->auth_info, pflavor)) {
+				struct rpc_clnt *new;
+				struct rpc_cred *cred;
+
+				/* Cloning creates an rpc_auth for the flavor */
+				new = rpc_clone_client_set_auth(clnt, pflavor);
+				if (IS_ERR(new))
+					continue;
+				/**
+				* Check that the user actually can use the
+				* flavor. This is mostly for RPC_AUTH_GSS
+				* where cr_init obtains a gss context
+				*/
+				cred = rpcauth_lookupcred(new->cl_auth, 0);
+				if (IS_ERR(cred)) {
+					rpc_shutdown_client(new);
+					continue;
+				}
+				put_rpccred(cred);
+				return new;
+			}
 		}
 	}
-
-	/* if there were any sec= options then nothing matched */
-	if (server->auth_info.flavor_len > 0)
-		return -EPERM;
-
-	return RPC_AUTH_UNIX;
+	return ERR_PTR(-EPERM);
 }
 
-static rpc_authflavor_t nfs4_negotiate_security(struct inode *inode, struct qstr *name)
+/**
+ * nfs4_negotiate_security - in response to an NFS4ERR_WRONGSEC on lookup,
+ * return an rpc_clnt that uses the best available security flavor with
+ * respect to the secinfo flavor list and the sec= mount options.
+ *
+ * @clnt: RPC client to clone
+ * @inode: directory inode
+ * @name: lookup name
+ *
+ * Please call rpc_shutdown_client() when you are done with this rpc client.
+ */
+struct rpc_clnt *
+nfs4_negotiate_security(struct rpc_clnt *clnt, struct inode *inode,
+					struct qstr *name)
 {
 	struct page *page;
 	struct nfs4_secinfo_flavors *flavors;
-	rpc_authflavor_t flavor;
+	struct rpc_clnt *new;
 	int err;
 
 	page = alloc_page(GFP_KERNEL);
 	if (!page)
-		return -ENOMEM;
+		return ERR_PTR(-ENOMEM);
+
 	flavors = page_address(page);
 
 	err = nfs4_proc_secinfo(inode, name, flavors);
 	if (err < 0) {
-		flavor = err;
+		new = ERR_PTR(err);
 		goto out;
 	}
 
-	flavor = nfs_find_best_sec(NFS_SERVER(inode), flavors);
+	new = nfs_find_best_sec(clnt, NFS_SERVER(inode), flavors);
 
 out:
 	put_page(page);
-	return flavor;
-}
-
-/*
- * Please call rpc_shutdown_client() when you are done with this client.
- */
-struct rpc_clnt *nfs4_create_sec_client(struct rpc_clnt *clnt, struct inode *inode,
-					struct qstr *name)
-{
-	rpc_authflavor_t flavor;
-
-	flavor = nfs4_negotiate_security(inode, name);
-	if ((int)flavor < 0)
-		return ERR_PTR((int)flavor);
-
-	return rpc_clone_client_set_auth(clnt, flavor);
+	return new;
 }
 
 static struct vfsmount *try_location(struct nfs_clone_mount *mountdata,
@@ -397,11 +414,6 @@ struct vfsmount *nfs4_submount(struct nfs_server *server, struct dentry *dentry,
 
 	if (client->cl_auth->au_flavor != flavor)
 		flavor = client->cl_auth->au_flavor;
-	else {
-		rpc_authflavor_t new = nfs4_negotiate_security(dir, name);
-		if ((int)new >= 0)
-			flavor = new;
-	}
 	mnt = nfs_do_submount(dentry, fh, fattr, flavor);
 out:
 	rpc_shutdown_client(client);
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 285ad5334018..75ae8d22f067 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -1952,6 +1952,14 @@ static int _nfs4_recover_proc_open(struct nfs4_opendata *data)
 	return status;
 }
 
+/*
+ * Additional permission checks in order to distinguish between an
+ * open for read, and an open for execute. This works around the
+ * fact that NFSv4 OPEN treats read and execute permissions as being
+ * the same.
+ * Note that in the non-execute case, we want to turn off permission
+ * checking if we just created a new file (POSIX open() semantics).
+ */
 static int nfs4_opendata_access(struct rpc_cred *cred,
 				struct nfs4_opendata *opendata,
 				struct nfs4_state *state, fmode_t fmode,
@@ -1966,14 +1974,14 @@ static int nfs4_opendata_access(struct rpc_cred *cred,
 		return 0;
 
 	mask = 0;
-	/* don't check MAY_WRITE - a newly created file may not have
-	 * write mode bits, but POSIX allows the creating process to write.
-	 * use openflags to check for exec, because fmode won't
-	 * always have FMODE_EXEC set when file open for exec. */
+	/*
+	 * Use openflags to check for exec, because fmode won't
+	 * always have FMODE_EXEC set when file open for exec.
+	 */
 	if (openflags & __FMODE_EXEC) {
 		/* ONLY check for exec rights */
 		mask = MAY_EXEC;
-	} else if (fmode & FMODE_READ)
+	} else if ((fmode & FMODE_READ) && !opendata->file_created)
 		mask = MAY_READ;
 
 	cache.cred = cred;
@@ -2216,8 +2224,15 @@ static int _nfs4_open_and_get_state(struct nfs4_opendata *opendata,
 	seq = raw_seqcount_begin(&sp->so_reclaim_seqcount);
 
 	ret = _nfs4_proc_open(opendata);
-	if (ret != 0)
+	if (ret != 0) {
+		if (ret == -ENOENT) {
+			d_drop(opendata->dentry);
+			d_add(opendata->dentry, NULL);
+			nfs_set_verifier(opendata->dentry,
+					 nfs_save_change_attribute(opendata->dir->d_inode));
+		}
 		goto out;
+	}
 
 	state = nfs4_opendata_to_nfs4_state(opendata);
 	ret = PTR_ERR(state);
@@ -2647,6 +2662,48 @@ static const struct rpc_call_ops nfs4_close_ops = {
 	.rpc_release = nfs4_free_closedata,
 };
 
+static bool nfs4_state_has_opener(struct nfs4_state *state)
+{
+	/* first check existing openers */
+	if (test_bit(NFS_O_RDONLY_STATE, &state->flags) != 0 &&
+	    state->n_rdonly != 0)
+		return true;
+
+	if (test_bit(NFS_O_WRONLY_STATE, &state->flags) != 0 &&
+	    state->n_wronly != 0)
+		return true;
+
+	if (test_bit(NFS_O_RDWR_STATE, &state->flags) != 0 &&
+	    state->n_rdwr != 0)
+		return true;
+
+	return false;
+}
+
+static bool nfs4_roc(struct inode *inode)
+{
+	struct nfs_inode *nfsi = NFS_I(inode);
+	struct nfs_open_context *ctx;
+	struct nfs4_state *state;
+
+	spin_lock(&inode->i_lock);
+	list_for_each_entry(ctx, &nfsi->open_files, list) {
+		state = ctx->state;
+		if (state == NULL)
+			continue;
+		if (nfs4_state_has_opener(state)) {
+			spin_unlock(&inode->i_lock);
+			return false;
+		}
+	}
+	spin_unlock(&inode->i_lock);
+
+	if (nfs4_check_delegation(inode, FMODE_READ))
+		return false;
+
+	return pnfs_roc(inode);
+}
+
 /* 
  * It is possible for data to be read/written from a mem-mapped file 
  * after the sys_close call (which hits the vfs layer as a flush).
@@ -2697,7 +2754,7 @@ int nfs4_do_close(struct nfs4_state *state, gfp_t gfp_mask, int wait)
 	calldata->res.fattr = &calldata->fattr;
 	calldata->res.seqid = calldata->arg.seqid;
 	calldata->res.server = server;
-	calldata->roc = pnfs_roc(state->inode);
+	calldata->roc = nfs4_roc(state->inode);
 	nfs_sb_active(calldata->inode->i_sb);
 
 	msg.rpc_argp = &calldata->arg;
@@ -3247,7 +3304,7 @@ static int nfs4_proc_lookup_common(struct rpc_clnt **clnt, struct inode *dir,
 			err = -EPERM;
 			if (client != *clnt)
 				goto out;
-			client = nfs4_create_sec_client(client, dir, name);
+			client = nfs4_negotiate_security(client, dir, name);
 			if (IS_ERR(client))
 				return PTR_ERR(client);
 
@@ -4033,24 +4090,25 @@ static bool nfs4_error_stateid_expired(int err)
 	return false;
 }
 
-void __nfs4_read_done_cb(struct nfs_pgio_data *data)
+void __nfs4_read_done_cb(struct nfs_pgio_header *hdr)
 {
-	nfs_invalidate_atime(data->header->inode);
+	nfs_invalidate_atime(hdr->inode);
 }
 
-static int nfs4_read_done_cb(struct rpc_task *task, struct nfs_pgio_data *data)
+static int nfs4_read_done_cb(struct rpc_task *task, struct nfs_pgio_header *hdr)
 {
-	struct nfs_server *server = NFS_SERVER(data->header->inode);
+	struct nfs_server *server = NFS_SERVER(hdr->inode);
 
-	trace_nfs4_read(data, task->tk_status);
-	if (nfs4_async_handle_error(task, server, data->args.context->state) == -EAGAIN) {
+	trace_nfs4_read(hdr, task->tk_status);
+	if (nfs4_async_handle_error(task, server,
+				    hdr->args.context->state) == -EAGAIN) {
 		rpc_restart_call_prepare(task);
 		return -EAGAIN;
 	}
 
-	__nfs4_read_done_cb(data);
+	__nfs4_read_done_cb(hdr);
 	if (task->tk_status > 0)
-		renew_lease(server, data->timestamp);
+		renew_lease(server, hdr->timestamp);
 	return 0;
 }
 
@@ -4068,54 +4126,59 @@ static bool nfs4_read_stateid_changed(struct rpc_task *task,
 	return true;
 }
 
-static int nfs4_read_done(struct rpc_task *task, struct nfs_pgio_data *data)
+static int nfs4_read_done(struct rpc_task *task, struct nfs_pgio_header *hdr)
 {
 
 	dprintk("--> %s\n", __func__);
 
-	if (!nfs4_sequence_done(task, &data->res.seq_res))
+	if (!nfs4_sequence_done(task, &hdr->res.seq_res))
 		return -EAGAIN;
-	if (nfs4_read_stateid_changed(task, &data->args))
+	if (nfs4_read_stateid_changed(task, &hdr->args))
 		return -EAGAIN;
-	return data->pgio_done_cb ? data->pgio_done_cb(task, data) :
-				    nfs4_read_done_cb(task, data);
+	return hdr->pgio_done_cb ? hdr->pgio_done_cb(task, hdr) :
+				    nfs4_read_done_cb(task, hdr);
 }
 
-static void nfs4_proc_read_setup(struct nfs_pgio_data *data, struct rpc_message *msg)
+static void nfs4_proc_read_setup(struct nfs_pgio_header *hdr,
+				 struct rpc_message *msg)
 {
-	data->timestamp   = jiffies;
-	data->pgio_done_cb = nfs4_read_done_cb;
+	hdr->timestamp   = jiffies;
+	hdr->pgio_done_cb = nfs4_read_done_cb;
 	msg->rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_READ];
-	nfs4_init_sequence(&data->args.seq_args, &data->res.seq_res, 0);
+	nfs4_init_sequence(&hdr->args.seq_args, &hdr->res.seq_res, 0);
 }
 
-static int nfs4_proc_pgio_rpc_prepare(struct rpc_task *task, struct nfs_pgio_data *data)
+static int nfs4_proc_pgio_rpc_prepare(struct rpc_task *task,
+				      struct nfs_pgio_header *hdr)
 {
-	if (nfs4_setup_sequence(NFS_SERVER(data->header->inode),
-			&data->args.seq_args,
-			&data->res.seq_res,
+	if (nfs4_setup_sequence(NFS_SERVER(hdr->inode),
+			&hdr->args.seq_args,
+			&hdr->res.seq_res,
 			task))
 		return 0;
-	if (nfs4_set_rw_stateid(&data->args.stateid, data->args.context,
-				data->args.lock_context, data->header->rw_ops->rw_mode) == -EIO)
+	if (nfs4_set_rw_stateid(&hdr->args.stateid, hdr->args.context,
+				hdr->args.lock_context,
+				hdr->rw_ops->rw_mode) == -EIO)
 		return -EIO;
-	if (unlikely(test_bit(NFS_CONTEXT_BAD, &data->args.context->flags)))
+	if (unlikely(test_bit(NFS_CONTEXT_BAD, &hdr->args.context->flags)))
 		return -EIO;
 	return 0;
 }
 
-static int nfs4_write_done_cb(struct rpc_task *task, struct nfs_pgio_data *data)
+static int nfs4_write_done_cb(struct rpc_task *task,
+			      struct nfs_pgio_header *hdr)
 {
-	struct inode *inode = data->header->inode;
+	struct inode *inode = hdr->inode;
 	
-	trace_nfs4_write(data, task->tk_status);
-	if (nfs4_async_handle_error(task, NFS_SERVER(inode), data->args.context->state) == -EAGAIN) {
+	trace_nfs4_write(hdr, task->tk_status);
+	if (nfs4_async_handle_error(task, NFS_SERVER(inode),
+				    hdr->args.context->state) == -EAGAIN) {
 		rpc_restart_call_prepare(task);
 		return -EAGAIN;
 	}
 	if (task->tk_status >= 0) {
-		renew_lease(NFS_SERVER(inode), data->timestamp);
-		nfs_post_op_update_inode_force_wcc(inode, &data->fattr);
+		renew_lease(NFS_SERVER(inode), hdr->timestamp);
+		nfs_post_op_update_inode_force_wcc(inode, &hdr->fattr);
 	}
 	return 0;
 }
@@ -4134,23 +4197,21 @@ static bool nfs4_write_stateid_changed(struct rpc_task *task,
 	return true;
 }
 
-static int nfs4_write_done(struct rpc_task *task, struct nfs_pgio_data *data)
+static int nfs4_write_done(struct rpc_task *task, struct nfs_pgio_header *hdr)
 {
-	if (!nfs4_sequence_done(task, &data->res.seq_res))
+	if (!nfs4_sequence_done(task, &hdr->res.seq_res))
 		return -EAGAIN;
-	if (nfs4_write_stateid_changed(task, &data->args))
+	if (nfs4_write_stateid_changed(task, &hdr->args))
 		return -EAGAIN;
-	return data->pgio_done_cb ? data->pgio_done_cb(task, data) :
-		nfs4_write_done_cb(task, data);
+	return hdr->pgio_done_cb ? hdr->pgio_done_cb(task, hdr) :
+		nfs4_write_done_cb(task, hdr);
 }
 
 static
-bool nfs4_write_need_cache_consistency_data(const struct nfs_pgio_data *data)
+bool nfs4_write_need_cache_consistency_data(struct nfs_pgio_header *hdr)
 {
-	const struct nfs_pgio_header *hdr = data->header;
-
 	/* Don't request attributes for pNFS or O_DIRECT writes */
-	if (data->ds_clp != NULL || hdr->dreq != NULL)
+	if (hdr->ds_clp != NULL || hdr->dreq != NULL)
 		return false;
 	/* Otherwise, request attributes if and only if we don't hold
 	 * a delegation
@@ -4158,23 +4219,24 @@ bool nfs4_write_need_cache_consistency_data(const struct nfs_pgio_data *data)
 	return nfs4_have_delegation(hdr->inode, FMODE_READ) == 0;
 }
 
-static void nfs4_proc_write_setup(struct nfs_pgio_data *data, struct rpc_message *msg)
+static void nfs4_proc_write_setup(struct nfs_pgio_header *hdr,
+				  struct rpc_message *msg)
 {
-	struct nfs_server *server = NFS_SERVER(data->header->inode);
+	struct nfs_server *server = NFS_SERVER(hdr->inode);
 
-	if (!nfs4_write_need_cache_consistency_data(data)) {
-		data->args.bitmask = NULL;
-		data->res.fattr = NULL;
+	if (!nfs4_write_need_cache_consistency_data(hdr)) {
+		hdr->args.bitmask = NULL;
+		hdr->res.fattr = NULL;
 	} else
-		data->args.bitmask = server->cache_consistency_bitmask;
+		hdr->args.bitmask = server->cache_consistency_bitmask;
 
-	if (!data->pgio_done_cb)
-		data->pgio_done_cb = nfs4_write_done_cb;
-	data->res.server = server;
-	data->timestamp   = jiffies;
+	if (!hdr->pgio_done_cb)
+		hdr->pgio_done_cb = nfs4_write_done_cb;
+	hdr->res.server = server;
+	hdr->timestamp   = jiffies;
 
 	msg->rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_WRITE];
-	nfs4_init_sequence(&data->args.seq_args, &data->res.seq_res, 1);
+	nfs4_init_sequence(&hdr->args.seq_args, &hdr->res.seq_res, 1);
 }
 
 static void nfs4_proc_commit_rpc_prepare(struct rpc_task *task, struct nfs_commit_data *data)
@@ -4881,6 +4943,18 @@ nfs4_init_callback_netid(const struct nfs_client *clp, char *buf, size_t len)
 		return scnprintf(buf, len, "tcp");
 }
 
+static void nfs4_setclientid_done(struct rpc_task *task, void *calldata)
+{
+	struct nfs4_setclientid *sc = calldata;
+
+	if (task->tk_status == 0)
+		sc->sc_cred = get_rpccred(task->tk_rqstp->rq_cred);
+}
+
+static const struct rpc_call_ops nfs4_setclientid_ops = {
+	.rpc_call_done = nfs4_setclientid_done,
+};
+
 /**
  * nfs4_proc_setclientid - Negotiate client ID
  * @clp: state data structure
@@ -4907,6 +4981,14 @@ int nfs4_proc_setclientid(struct nfs_client *clp, u32 program,
 		.rpc_resp = res,
 		.rpc_cred = cred,
 	};
+	struct rpc_task *task;
+	struct rpc_task_setup task_setup_data = {
+		.rpc_client = clp->cl_rpcclient,
+		.rpc_message = &msg,
+		.callback_ops = &nfs4_setclientid_ops,
+		.callback_data = &setclientid,
+		.flags = RPC_TASK_TIMEOUT,
+	};
 	int status;
 
 	/* nfs_client_id4 */
@@ -4933,7 +5015,18 @@ int nfs4_proc_setclientid(struct nfs_client *clp, u32 program,
 	dprintk("NFS call  setclientid auth=%s, '%.*s'\n",
 		clp->cl_rpcclient->cl_auth->au_ops->au_name,
 		setclientid.sc_name_len, setclientid.sc_name);
-	status = rpc_call_sync(clp->cl_rpcclient, &msg, RPC_TASK_TIMEOUT);
+	task = rpc_run_task(&task_setup_data);
+	if (IS_ERR(task)) {
+		status = PTR_ERR(task);
+		goto out;
+	}
+	status = task->tk_status;
+	if (setclientid.sc_cred) {
+		clp->cl_acceptor = rpcauth_stringify_acceptor(setclientid.sc_cred);
+		put_rpccred(setclientid.sc_cred);
+	}
+	rpc_put_task(task);
+out:
 	trace_nfs4_setclientid(clp, status);
 	dprintk("NFS reply setclientid: %d\n", status);
 	return status;
@@ -4975,6 +5068,9 @@ struct nfs4_delegreturndata {
 	unsigned long timestamp;
 	struct nfs_fattr fattr;
 	int rpc_status;
+	struct inode *inode;
+	bool roc;
+	u32 roc_barrier;
 };
 
 static void nfs4_delegreturn_done(struct rpc_task *task, void *calldata)
@@ -4988,7 +5084,6 @@ static void nfs4_delegreturn_done(struct rpc_task *task, void *calldata)
 	switch (task->tk_status) {
 	case 0:
 		renew_lease(data->res.server, data->timestamp);
-		break;
 	case -NFS4ERR_ADMIN_REVOKED:
 	case -NFS4ERR_DELEG_REVOKED:
 	case -NFS4ERR_BAD_STATEID:
@@ -4996,6 +5091,8 @@ static void nfs4_delegreturn_done(struct rpc_task *task, void *calldata)
 	case -NFS4ERR_STALE_STATEID:
 	case -NFS4ERR_EXPIRED:
 		task->tk_status = 0;
+		if (data->roc)
+			pnfs_roc_set_barrier(data->inode, data->roc_barrier);
 		break;
 	default:
 		if (nfs4_async_handle_error(task, data->res.server, NULL) ==
@@ -5009,6 +5106,10 @@ static void nfs4_delegreturn_done(struct rpc_task *task, void *calldata)
 
 static void nfs4_delegreturn_release(void *calldata)
 {
+	struct nfs4_delegreturndata *data = calldata;
+
+	if (data->roc)
+		pnfs_roc_release(data->inode);
 	kfree(calldata);
 }
 
@@ -5018,6 +5119,10 @@ static void nfs4_delegreturn_prepare(struct rpc_task *task, void *data)
 
 	d_data = (struct nfs4_delegreturndata *)data;
 
+	if (d_data->roc &&
+	    pnfs_roc_drain(d_data->inode, &d_data->roc_barrier, task))
+		return;
+
 	nfs4_setup_sequence(d_data->res.server,
 			&d_data->args.seq_args,
 			&d_data->res.seq_res,
@@ -5061,6 +5166,9 @@ static int _nfs4_proc_delegreturn(struct inode *inode, struct rpc_cred *cred, co
 	nfs_fattr_init(data->res.fattr);
 	data->timestamp = jiffies;
 	data->rpc_status = 0;
+	data->inode = inode;
+	data->roc = list_empty(&NFS_I(inode)->open_files) ?
+		    pnfs_roc(inode) : false;
 
 	task_setup_data.callback_data = data;
 	msg.rpc_argp = &data->args;
@@ -5834,8 +5942,10 @@ struct nfs_release_lockowner_data {
 static void nfs4_release_lockowner_prepare(struct rpc_task *task, void *calldata)
 {
 	struct nfs_release_lockowner_data *data = calldata;
-	nfs40_setup_sequence(data->server,
-				&data->args.seq_args, &data->res.seq_res, task);
+	struct nfs_server *server = data->server;
+	nfs40_setup_sequence(server, &data->args.seq_args,
+				&data->res.seq_res, task);
+	data->args.lock_owner.clientid = server->nfs_client->cl_clientid;
 	data->timestamp = jiffies;
 }
 
@@ -5852,6 +5962,8 @@ static void nfs4_release_lockowner_done(struct rpc_task *task, void *calldata)
 		break;
 	case -NFS4ERR_STALE_CLIENTID:
 	case -NFS4ERR_EXPIRED:
+		nfs4_schedule_lease_recovery(server->nfs_client);
+		break;
 	case -NFS4ERR_LEASE_MOVED:
 	case -NFS4ERR_DELAY:
 		if (nfs4_async_handle_error(task, server, NULL) == -EAGAIN)
@@ -5872,7 +5984,8 @@ static const struct rpc_call_ops nfs4_release_lockowner_ops = {
 	.rpc_release = nfs4_release_lockowner_release,
 };
 
-static int nfs4_release_lockowner(struct nfs_server *server, struct nfs4_lock_state *lsp)
+static void
+nfs4_release_lockowner(struct nfs_server *server, struct nfs4_lock_state *lsp)
 {
 	struct nfs_release_lockowner_data *data;
 	struct rpc_message msg = {
@@ -5880,11 +5993,11 @@ static int nfs4_release_lockowner(struct nfs_server *server, struct nfs4_lock_st
 	};
 
 	if (server->nfs_client->cl_mvops->minor_version != 0)
-		return -EINVAL;
+		return;
 
 	data = kmalloc(sizeof(*data), GFP_NOFS);
 	if (!data)
-		return -ENOMEM;
+		return;
 	data->lsp = lsp;
 	data->server = server;
 	data->args.lock_owner.clientid = server->nfs_client->cl_clientid;
@@ -5895,7 +6008,6 @@ static int nfs4_release_lockowner(struct nfs_server *server, struct nfs4_lock_st
 	msg.rpc_resp = &data->res;
 	nfs4_init_sequence(&data->args.seq_args, &data->res.seq_res, 0);
 	rpc_call_async(server->client, &msg, 0, &nfs4_release_lockowner_ops, data);
-	return 0;
 }
 
 #define XATTR_NAME_NFSV4_ACL "system.nfs4_acl"
@@ -8182,7 +8294,8 @@ static int nfs41_free_stateid(struct nfs_server *server,
 	return ret;
 }
 
-static int nfs41_free_lock_state(struct nfs_server *server, struct nfs4_lock_state *lsp)
+static void
+nfs41_free_lock_state(struct nfs_server *server, struct nfs4_lock_state *lsp)
 {
 	struct rpc_task *task;
 	struct rpc_cred *cred = lsp->ls_state->owner->so_cred;
@@ -8190,9 +8303,8 @@ static int nfs41_free_lock_state(struct nfs_server *server, struct nfs4_lock_sta
 	task = _nfs41_free_stateid(server, &lsp->ls_stateid, cred, false);
 	nfs4_free_lock_state(server, lsp);
 	if (IS_ERR(task))
-		return PTR_ERR(task);
+		return;
 	rpc_put_task(task);
-	return 0;
 }
 
 static bool nfs41_match_stateid(const nfs4_stateid *s1,
diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c
index 848f6853c59e..a043f618cd5a 100644
--- a/fs/nfs/nfs4state.c
+++ b/fs/nfs/nfs4state.c
@@ -787,33 +787,36 @@ void nfs4_close_sync(struct nfs4_state *state, fmode_t fmode)
  * that is compatible with current->files
  */
 static struct nfs4_lock_state *
-__nfs4_find_lock_state(struct nfs4_state *state, fl_owner_t fl_owner, pid_t fl_pid, unsigned int type)
+__nfs4_find_lock_state(struct nfs4_state *state, fl_owner_t fl_owner)
 {
 	struct nfs4_lock_state *pos;
 	list_for_each_entry(pos, &state->lock_states, ls_locks) {
-		if (type != NFS4_ANY_LOCK_TYPE && pos->ls_owner.lo_type != type)
+		if (pos->ls_owner != fl_owner)
 			continue;
-		switch (pos->ls_owner.lo_type) {
-		case NFS4_POSIX_LOCK_TYPE:
-			if (pos->ls_owner.lo_u.posix_owner != fl_owner)
-				continue;
-			break;
-		case NFS4_FLOCK_LOCK_TYPE:
-			if (pos->ls_owner.lo_u.flock_owner != fl_pid)
-				continue;
-		}
 		atomic_inc(&pos->ls_count);
 		return pos;
 	}
 	return NULL;
 }
 
+static void
+free_lock_state_work(struct work_struct *work)
+{
+	struct nfs4_lock_state *lsp = container_of(work,
+					struct nfs4_lock_state, ls_release);
+	struct nfs4_state *state = lsp->ls_state;
+	struct nfs_server *server = state->owner->so_server;
+	struct nfs_client *clp = server->nfs_client;
+
+	clp->cl_mvops->free_lock_state(server, lsp);
+}
+
 /*
  * Return a compatible lock_state. If no initialized lock_state structure
  * exists, return an uninitialized one.
  *
  */
-static struct nfs4_lock_state *nfs4_alloc_lock_state(struct nfs4_state *state, fl_owner_t fl_owner, pid_t fl_pid, unsigned int type)
+static struct nfs4_lock_state *nfs4_alloc_lock_state(struct nfs4_state *state, fl_owner_t fl_owner)
 {
 	struct nfs4_lock_state *lsp;
 	struct nfs_server *server = state->owner->so_server;
@@ -824,21 +827,12 @@ static struct nfs4_lock_state *nfs4_alloc_lock_state(struct nfs4_state *state, f
 	nfs4_init_seqid_counter(&lsp->ls_seqid);
 	atomic_set(&lsp->ls_count, 1);
 	lsp->ls_state = state;
-	lsp->ls_owner.lo_type = type;
-	switch (lsp->ls_owner.lo_type) {
-	case NFS4_FLOCK_LOCK_TYPE:
-		lsp->ls_owner.lo_u.flock_owner = fl_pid;
-		break;
-	case NFS4_POSIX_LOCK_TYPE:
-		lsp->ls_owner.lo_u.posix_owner = fl_owner;
-		break;
-	default:
-		goto out_free;
-	}
+	lsp->ls_owner = fl_owner;
 	lsp->ls_seqid.owner_id = ida_simple_get(&server->lockowner_id, 0, 0, GFP_NOFS);
 	if (lsp->ls_seqid.owner_id < 0)
 		goto out_free;
 	INIT_LIST_HEAD(&lsp->ls_locks);
+	INIT_WORK(&lsp->ls_release, free_lock_state_work);
 	return lsp;
 out_free:
 	kfree(lsp);
@@ -857,13 +851,13 @@ void nfs4_free_lock_state(struct nfs_server *server, struct nfs4_lock_state *lsp
  * exists, return an uninitialized one.
  *
  */
-static struct nfs4_lock_state *nfs4_get_lock_state(struct nfs4_state *state, fl_owner_t owner, pid_t pid, unsigned int type)
+static struct nfs4_lock_state *nfs4_get_lock_state(struct nfs4_state *state, fl_owner_t owner)
 {
 	struct nfs4_lock_state *lsp, *new = NULL;
 	
 	for(;;) {
 		spin_lock(&state->state_lock);
-		lsp = __nfs4_find_lock_state(state, owner, pid, type);
+		lsp = __nfs4_find_lock_state(state, owner);
 		if (lsp != NULL)
 			break;
 		if (new != NULL) {
@@ -874,7 +868,7 @@ static struct nfs4_lock_state *nfs4_get_lock_state(struct nfs4_state *state, fl_
 			break;
 		}
 		spin_unlock(&state->state_lock);
-		new = nfs4_alloc_lock_state(state, owner, pid, type);
+		new = nfs4_alloc_lock_state(state, owner);
 		if (new == NULL)
 			return NULL;
 	}
@@ -902,13 +896,12 @@ void nfs4_put_lock_state(struct nfs4_lock_state *lsp)
 	if (list_empty(&state->lock_states))
 		clear_bit(LK_STATE_IN_USE, &state->flags);
 	spin_unlock(&state->state_lock);
-	server = state->owner->so_server;
-	if (test_bit(NFS_LOCK_INITIALIZED, &lsp->ls_flags)) {
-		struct nfs_client *clp = server->nfs_client;
-
-		clp->cl_mvops->free_lock_state(server, lsp);
-	} else
+	if (test_bit(NFS_LOCK_INITIALIZED, &lsp->ls_flags))
+		queue_work(nfsiod_workqueue, &lsp->ls_release);
+	else {
+		server = state->owner->so_server;
 		nfs4_free_lock_state(server, lsp);
+	}
 }
 
 static void nfs4_fl_copy_lock(struct file_lock *dst, struct file_lock *src)
@@ -935,13 +928,7 @@ int nfs4_set_lock_state(struct nfs4_state *state, struct file_lock *fl)
 
 	if (fl->fl_ops != NULL)
 		return 0;
-	if (fl->fl_flags & FL_POSIX)
-		lsp = nfs4_get_lock_state(state, fl->fl_owner, 0, NFS4_POSIX_LOCK_TYPE);
-	else if (fl->fl_flags & FL_FLOCK)
-		lsp = nfs4_get_lock_state(state, NULL, fl->fl_pid,
-				NFS4_FLOCK_LOCK_TYPE);
-	else
-		return -EINVAL;
+	lsp = nfs4_get_lock_state(state, fl->fl_owner);
 	if (lsp == NULL)
 		return -ENOMEM;
 	fl->fl_u.nfs4_fl.owner = lsp;
@@ -955,7 +942,6 @@ static int nfs4_copy_lock_stateid(nfs4_stateid *dst,
 {
 	struct nfs4_lock_state *lsp;
 	fl_owner_t fl_owner;
-	pid_t fl_pid;
 	int ret = -ENOENT;
 
 
@@ -966,9 +952,8 @@ static int nfs4_copy_lock_stateid(nfs4_stateid *dst,
 		goto out;
 
 	fl_owner = lockowner->l_owner;
-	fl_pid = lockowner->l_pid;
 	spin_lock(&state->state_lock);
-	lsp = __nfs4_find_lock_state(state, fl_owner, fl_pid, NFS4_ANY_LOCK_TYPE);
+	lsp = __nfs4_find_lock_state(state, fl_owner);
 	if (lsp && test_bit(NFS_LOCK_LOST, &lsp->ls_flags))
 		ret = -EIO;
 	else if (lsp != NULL && test_bit(NFS_LOCK_INITIALIZED, &lsp->ls_flags) != 0) {
@@ -1251,8 +1236,8 @@ int nfs4_wait_clnt_recover(struct nfs_client *clp)
 	might_sleep();
 
 	atomic_inc(&clp->cl_count);
-	res = wait_on_bit(&clp->cl_state, NFS4CLNT_MANAGER_RUNNING,
-			nfs_wait_bit_killable, TASK_KILLABLE);
+	res = wait_on_bit_action(&clp->cl_state, NFS4CLNT_MANAGER_RUNNING,
+				 nfs_wait_bit_killable, TASK_KILLABLE);
 	if (res)
 		goto out;
 	if (clp->cl_cons_state < 0)
diff --git a/fs/nfs/nfs4trace.h b/fs/nfs/nfs4trace.h
index 0a744f3a86f6..1c32adbe728d 100644
--- a/fs/nfs/nfs4trace.h
+++ b/fs/nfs/nfs4trace.h
@@ -932,11 +932,11 @@ DEFINE_NFS4_IDMAP_EVENT(nfs4_map_gid_to_group);
 
 DECLARE_EVENT_CLASS(nfs4_read_event,
 		TP_PROTO(
-			const struct nfs_pgio_data *data,
+			const struct nfs_pgio_header *hdr,
 			int error
 		),
 
-		TP_ARGS(data, error),
+		TP_ARGS(hdr, error),
 
 		TP_STRUCT__entry(
 			__field(dev_t, dev)
@@ -948,12 +948,12 @@ DECLARE_EVENT_CLASS(nfs4_read_event,
 		),
 
 		TP_fast_assign(
-			const struct inode *inode = data->header->inode;
+			const struct inode *inode = hdr->inode;
 			__entry->dev = inode->i_sb->s_dev;
 			__entry->fileid = NFS_FILEID(inode);
 			__entry->fhandle = nfs_fhandle_hash(NFS_FH(inode));
-			__entry->offset = data->args.offset;
-			__entry->count = data->args.count;
+			__entry->offset = hdr->args.offset;
+			__entry->count = hdr->args.count;
 			__entry->error = error;
 		),
 
@@ -972,10 +972,10 @@ DECLARE_EVENT_CLASS(nfs4_read_event,
 #define DEFINE_NFS4_READ_EVENT(name) \
 	DEFINE_EVENT(nfs4_read_event, name, \
 			TP_PROTO( \
-				const struct nfs_pgio_data *data, \
+				const struct nfs_pgio_header *hdr, \
 				int error \
 			), \
-			TP_ARGS(data, error))
+			TP_ARGS(hdr, error))
 DEFINE_NFS4_READ_EVENT(nfs4_read);
 #ifdef CONFIG_NFS_V4_1
 DEFINE_NFS4_READ_EVENT(nfs4_pnfs_read);
@@ -983,11 +983,11 @@ DEFINE_NFS4_READ_EVENT(nfs4_pnfs_read);
 
 DECLARE_EVENT_CLASS(nfs4_write_event,
 		TP_PROTO(
-			const struct nfs_pgio_data *data,
+			const struct nfs_pgio_header *hdr,
 			int error
 		),
 
-		TP_ARGS(data, error),
+		TP_ARGS(hdr, error),
 
 		TP_STRUCT__entry(
 			__field(dev_t, dev)
@@ -999,12 +999,12 @@ DECLARE_EVENT_CLASS(nfs4_write_event,
 		),
 
 		TP_fast_assign(
-			const struct inode *inode = data->header->inode;
+			const struct inode *inode = hdr->inode;
 			__entry->dev = inode->i_sb->s_dev;
 			__entry->fileid = NFS_FILEID(inode);
 			__entry->fhandle = nfs_fhandle_hash(NFS_FH(inode));
-			__entry->offset = data->args.offset;
-			__entry->count = data->args.count;
+			__entry->offset = hdr->args.offset;
+			__entry->count = hdr->args.count;
 			__entry->error = error;
 		),
 
@@ -1024,10 +1024,10 @@ DECLARE_EVENT_CLASS(nfs4_write_event,
 #define DEFINE_NFS4_WRITE_EVENT(name) \
 	DEFINE_EVENT(nfs4_write_event, name, \
 			TP_PROTO( \
-				const struct nfs_pgio_data *data, \
+				const struct nfs_pgio_header *hdr, \
 				int error \
 			), \
-			TP_ARGS(data, error))
+			TP_ARGS(hdr, error))
 DEFINE_NFS4_WRITE_EVENT(nfs4_write);
 #ifdef CONFIG_NFS_V4_1
 DEFINE_NFS4_WRITE_EVENT(nfs4_pnfs_write);
diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c
index 939ae606cfa4..e13b59d8d9aa 100644
--- a/fs/nfs/nfs4xdr.c
+++ b/fs/nfs/nfs4xdr.c
@@ -7092,7 +7092,7 @@ static int nfs4_xdr_dec_reclaim_complete(struct rpc_rqst *rqstp,
 	if (!status)
 		status = decode_sequence(xdr, &res->seq_res, rqstp);
 	if (!status)
-		status = decode_reclaim_complete(xdr, (void *)NULL);
+		status = decode_reclaim_complete(xdr, NULL);
 	return status;
 }
 
diff --git a/fs/nfs/objlayout/objio_osd.c b/fs/nfs/objlayout/objio_osd.c
index 611320753db2..ae05278b3761 100644
--- a/fs/nfs/objlayout/objio_osd.c
+++ b/fs/nfs/objlayout/objio_osd.c
@@ -439,22 +439,21 @@ static void _read_done(struct ore_io_state *ios, void *private)
 	objlayout_read_done(&objios->oir, status, objios->sync);
 }
 
-int objio_read_pagelist(struct nfs_pgio_data *rdata)
+int objio_read_pagelist(struct nfs_pgio_header *hdr)
 {
-	struct nfs_pgio_header *hdr = rdata->header;
 	struct objio_state *objios;
 	int ret;
 
 	ret = objio_alloc_io_state(NFS_I(hdr->inode)->layout, true,
-			hdr->lseg, rdata->args.pages, rdata->args.pgbase,
-			rdata->args.offset, rdata->args.count, rdata,
+			hdr->lseg, hdr->args.pages, hdr->args.pgbase,
+			hdr->args.offset, hdr->args.count, hdr,
 			GFP_KERNEL, &objios);
 	if (unlikely(ret))
 		return ret;
 
 	objios->ios->done = _read_done;
 	dprintk("%s: offset=0x%llx length=0x%x\n", __func__,
-		rdata->args.offset, rdata->args.count);
+		hdr->args.offset, hdr->args.count);
 	ret = ore_read(objios->ios);
 	if (unlikely(ret))
 		objio_free_result(&objios->oir);
@@ -487,11 +486,11 @@ static void _write_done(struct ore_io_state *ios, void *private)
 static struct page *__r4w_get_page(void *priv, u64 offset, bool *uptodate)
 {
 	struct objio_state *objios = priv;
-	struct nfs_pgio_data *wdata = objios->oir.rpcdata;
-	struct address_space *mapping = wdata->header->inode->i_mapping;
+	struct nfs_pgio_header *hdr = objios->oir.rpcdata;
+	struct address_space *mapping = hdr->inode->i_mapping;
 	pgoff_t index = offset / PAGE_SIZE;
 	struct page *page;
-	loff_t i_size = i_size_read(wdata->header->inode);
+	loff_t i_size = i_size_read(hdr->inode);
 
 	if (offset >= i_size) {
 		*uptodate = true;
@@ -531,15 +530,14 @@ static const struct _ore_r4w_op _r4w_op = {
 	.put_page = &__r4w_put_page,
 };
 
-int objio_write_pagelist(struct nfs_pgio_data *wdata, int how)
+int objio_write_pagelist(struct nfs_pgio_header *hdr, int how)
 {
-	struct nfs_pgio_header *hdr = wdata->header;
 	struct objio_state *objios;
 	int ret;
 
 	ret = objio_alloc_io_state(NFS_I(hdr->inode)->layout, false,
-			hdr->lseg, wdata->args.pages, wdata->args.pgbase,
-			wdata->args.offset, wdata->args.count, wdata, GFP_NOFS,
+			hdr->lseg, hdr->args.pages, hdr->args.pgbase,
+			hdr->args.offset, hdr->args.count, hdr, GFP_NOFS,
 			&objios);
 	if (unlikely(ret))
 		return ret;
@@ -551,7 +549,7 @@ int objio_write_pagelist(struct nfs_pgio_data *wdata, int how)
 		objios->ios->done = _write_done;
 
 	dprintk("%s: offset=0x%llx length=0x%x\n", __func__,
-		wdata->args.offset, wdata->args.count);
+		hdr->args.offset, hdr->args.count);
 	ret = ore_write(objios->ios);
 	if (unlikely(ret)) {
 		objio_free_result(&objios->oir);
diff --git a/fs/nfs/objlayout/objlayout.c b/fs/nfs/objlayout/objlayout.c
index 765d3f54e986..697a16d11fac 100644
--- a/fs/nfs/objlayout/objlayout.c
+++ b/fs/nfs/objlayout/objlayout.c
@@ -229,36 +229,36 @@ objlayout_io_set_result(struct objlayout_io_res *oir, unsigned index,
 static void _rpc_read_complete(struct work_struct *work)
 {
 	struct rpc_task *task;
-	struct nfs_pgio_data *rdata;
+	struct nfs_pgio_header *hdr;
 
 	dprintk("%s enter\n", __func__);
 	task = container_of(work, struct rpc_task, u.tk_work);
-	rdata = container_of(task, struct nfs_pgio_data, task);
+	hdr = container_of(task, struct nfs_pgio_header, task);
 
-	pnfs_ld_read_done(rdata);
+	pnfs_ld_read_done(hdr);
 }
 
 void
 objlayout_read_done(struct objlayout_io_res *oir, ssize_t status, bool sync)
 {
-	struct nfs_pgio_data *rdata = oir->rpcdata;
+	struct nfs_pgio_header *hdr = oir->rpcdata;
 
-	oir->status = rdata->task.tk_status = status;
+	oir->status = hdr->task.tk_status = status;
 	if (status >= 0)
-		rdata->res.count = status;
+		hdr->res.count = status;
 	else
-		rdata->header->pnfs_error = status;
+		hdr->pnfs_error = status;
 	objlayout_iodone(oir);
 	/* must not use oir after this point */
 
 	dprintk("%s: Return status=%zd eof=%d sync=%d\n", __func__,
-		status, rdata->res.eof, sync);
+		status, hdr->res.eof, sync);
 
 	if (sync)
-		pnfs_ld_read_done(rdata);
+		pnfs_ld_read_done(hdr);
 	else {
-		INIT_WORK(&rdata->task.u.tk_work, _rpc_read_complete);
-		schedule_work(&rdata->task.u.tk_work);
+		INIT_WORK(&hdr->task.u.tk_work, _rpc_read_complete);
+		schedule_work(&hdr->task.u.tk_work);
 	}
 }
 
@@ -266,12 +266,11 @@ objlayout_read_done(struct objlayout_io_res *oir, ssize_t status, bool sync)
  * Perform sync or async reads.
  */
 enum pnfs_try_status
-objlayout_read_pagelist(struct nfs_pgio_data *rdata)
+objlayout_read_pagelist(struct nfs_pgio_header *hdr)
 {
-	struct nfs_pgio_header *hdr = rdata->header;
 	struct inode *inode = hdr->inode;
-	loff_t offset = rdata->args.offset;
-	size_t count = rdata->args.count;
+	loff_t offset = hdr->args.offset;
+	size_t count = hdr->args.count;
 	int err;
 	loff_t eof;
 
@@ -279,23 +278,23 @@ objlayout_read_pagelist(struct nfs_pgio_data *rdata)
 	if (unlikely(offset + count > eof)) {
 		if (offset >= eof) {
 			err = 0;
-			rdata->res.count = 0;
-			rdata->res.eof = 1;
+			hdr->res.count = 0;
+			hdr->res.eof = 1;
 			/*FIXME: do we need to call pnfs_ld_read_done() */
 			goto out;
 		}
 		count = eof - offset;
 	}
 
-	rdata->res.eof = (offset + count) >= eof;
-	_fix_verify_io_params(hdr->lseg, &rdata->args.pages,
-			      &rdata->args.pgbase,
-			      rdata->args.offset, rdata->args.count);
+	hdr->res.eof = (offset + count) >= eof;
+	_fix_verify_io_params(hdr->lseg, &hdr->args.pages,
+			      &hdr->args.pgbase,
+			      hdr->args.offset, hdr->args.count);
 
 	dprintk("%s: inode(%lx) offset 0x%llx count 0x%Zx eof=%d\n",
-		__func__, inode->i_ino, offset, count, rdata->res.eof);
+		__func__, inode->i_ino, offset, count, hdr->res.eof);
 
-	err = objio_read_pagelist(rdata);
+	err = objio_read_pagelist(hdr);
  out:
 	if (unlikely(err)) {
 		hdr->pnfs_error = err;
@@ -312,38 +311,38 @@ objlayout_read_pagelist(struct nfs_pgio_data *rdata)
 static void _rpc_write_complete(struct work_struct *work)
 {
 	struct rpc_task *task;
-	struct nfs_pgio_data *wdata;
+	struct nfs_pgio_header *hdr;
 
 	dprintk("%s enter\n", __func__);
 	task = container_of(work, struct rpc_task, u.tk_work);
-	wdata = container_of(task, struct nfs_pgio_data, task);
+	hdr = container_of(task, struct nfs_pgio_header, task);
 
-	pnfs_ld_write_done(wdata);
+	pnfs_ld_write_done(hdr);
 }
 
 void
 objlayout_write_done(struct objlayout_io_res *oir, ssize_t status, bool sync)
 {
-	struct nfs_pgio_data *wdata = oir->rpcdata;
+	struct nfs_pgio_header *hdr = oir->rpcdata;
 
-	oir->status = wdata->task.tk_status = status;
+	oir->status = hdr->task.tk_status = status;
 	if (status >= 0) {
-		wdata->res.count = status;
-		wdata->verf.committed = oir->committed;
+		hdr->res.count = status;
+		hdr->verf.committed = oir->committed;
 	} else {
-		wdata->header->pnfs_error = status;
+		hdr->pnfs_error = status;
 	}
 	objlayout_iodone(oir);
 	/* must not use oir after this point */
 
 	dprintk("%s: Return status %zd committed %d sync=%d\n", __func__,
-		status, wdata->verf.committed, sync);
+		status, hdr->verf.committed, sync);
 
 	if (sync)
-		pnfs_ld_write_done(wdata);
+		pnfs_ld_write_done(hdr);
 	else {
-		INIT_WORK(&wdata->task.u.tk_work, _rpc_write_complete);
-		schedule_work(&wdata->task.u.tk_work);
+		INIT_WORK(&hdr->task.u.tk_work, _rpc_write_complete);
+		schedule_work(&hdr->task.u.tk_work);
 	}
 }
 
@@ -351,17 +350,15 @@ objlayout_write_done(struct objlayout_io_res *oir, ssize_t status, bool sync)
  * Perform sync or async writes.
  */
 enum pnfs_try_status
-objlayout_write_pagelist(struct nfs_pgio_data *wdata,
-			 int how)
+objlayout_write_pagelist(struct nfs_pgio_header *hdr, int how)
 {
-	struct nfs_pgio_header *hdr = wdata->header;
 	int err;
 
-	_fix_verify_io_params(hdr->lseg, &wdata->args.pages,
-			      &wdata->args.pgbase,
-			      wdata->args.offset, wdata->args.count);
+	_fix_verify_io_params(hdr->lseg, &hdr->args.pages,
+			      &hdr->args.pgbase,
+			      hdr->args.offset, hdr->args.count);
 
-	err = objio_write_pagelist(wdata, how);
+	err = objio_write_pagelist(hdr, how);
 	if (unlikely(err)) {
 		hdr->pnfs_error = err;
 		dprintk("%s: Returned Error %d\n", __func__, err);
diff --git a/fs/nfs/objlayout/objlayout.h b/fs/nfs/objlayout/objlayout.h
index 01e041029a6c..fd13f1d2f136 100644
--- a/fs/nfs/objlayout/objlayout.h
+++ b/fs/nfs/objlayout/objlayout.h
@@ -119,8 +119,8 @@ extern void objio_free_lseg(struct pnfs_layout_segment *lseg);
  */
 extern void objio_free_result(struct objlayout_io_res *oir);
 
-extern int objio_read_pagelist(struct nfs_pgio_data *rdata);
-extern int objio_write_pagelist(struct nfs_pgio_data *wdata, int how);
+extern int objio_read_pagelist(struct nfs_pgio_header *rdata);
+extern int objio_write_pagelist(struct nfs_pgio_header *wdata, int how);
 
 /*
  * callback API
@@ -168,10 +168,10 @@ extern struct pnfs_layout_segment *objlayout_alloc_lseg(
 extern void objlayout_free_lseg(struct pnfs_layout_segment *);
 
 extern enum pnfs_try_status objlayout_read_pagelist(
-	struct nfs_pgio_data *);
+	struct nfs_pgio_header *);
 
 extern enum pnfs_try_status objlayout_write_pagelist(
-	struct nfs_pgio_data *,
+	struct nfs_pgio_header *,
 	int how);
 
 extern void objlayout_encode_layoutcommit(
diff --git a/fs/nfs/pagelist.c b/fs/nfs/pagelist.c
index b6ee3a6ee96d..ba491926df5f 100644
--- a/fs/nfs/pagelist.c
+++ b/fs/nfs/pagelist.c
@@ -29,8 +29,6 @@
 static struct kmem_cache *nfs_page_cachep;
 static const struct rpc_call_ops nfs_pgio_common_ops;
 
-static void nfs_free_request(struct nfs_page *);
-
 static bool nfs_pgarray_set(struct nfs_page_array *p, unsigned int pagecount)
 {
 	p->npages = pagecount;
@@ -117,7 +115,7 @@ __nfs_iocounter_wait(struct nfs_io_counter *c)
 		set_bit(NFS_IO_INPROGRESS, &c->flags);
 		if (atomic_read(&c->io_count) == 0)
 			break;
-		ret = nfs_wait_bit_killable(&c->flags);
+		ret = nfs_wait_bit_killable(&q.key);
 	} while (atomic_read(&c->io_count) != 0);
 	finish_wait(wq, &q.wait);
 	return ret;
@@ -138,28 +136,29 @@ nfs_iocounter_wait(struct nfs_io_counter *c)
 	return __nfs_iocounter_wait(c);
 }
 
-static int nfs_wait_bit_uninterruptible(void *word)
-{
-	io_schedule();
-	return 0;
-}
-
 /*
  * nfs_page_group_lock - lock the head of the page group
  * @req - request in group that is to be locked
  *
  * this lock must be held if modifying the page group list
+ *
+ * returns result from wait_on_bit_lock: 0 on success, < 0 on error
  */
-void
-nfs_page_group_lock(struct nfs_page *req)
+int
+nfs_page_group_lock(struct nfs_page *req, bool wait)
 {
 	struct nfs_page *head = req->wb_head;
+	int ret;
 
 	WARN_ON_ONCE(head != head->wb_head);
 
-	wait_on_bit_lock(&head->wb_flags, PG_HEADLOCK,
-			nfs_wait_bit_uninterruptible,
+	do {
+		ret = wait_on_bit_lock(&head->wb_flags, PG_HEADLOCK,
 			TASK_UNINTERRUPTIBLE);
+	} while (wait && ret != 0);
+
+	WARN_ON_ONCE(ret > 0);
+	return ret;
 }
 
 /*
@@ -220,7 +219,7 @@ bool nfs_page_group_sync_on_bit(struct nfs_page *req, unsigned int bit)
 {
 	bool ret;
 
-	nfs_page_group_lock(req);
+	nfs_page_group_lock(req, true);
 	ret = nfs_page_group_sync_on_bit_locked(req, bit);
 	nfs_page_group_unlock(req);
 
@@ -239,20 +238,28 @@ nfs_page_group_init(struct nfs_page *req, struct nfs_page *prev)
 	WARN_ON_ONCE(prev == req);
 
 	if (!prev) {
+		/* a head request */
 		req->wb_head = req;
 		req->wb_this_page = req;
 	} else {
+		/* a subrequest */
 		WARN_ON_ONCE(prev->wb_this_page != prev->wb_head);
 		WARN_ON_ONCE(!test_bit(PG_HEADLOCK, &prev->wb_head->wb_flags));
 		req->wb_head = prev->wb_head;
 		req->wb_this_page = prev->wb_this_page;
 		prev->wb_this_page = req;
 
+		/* All subrequests take a ref on the head request until
+		 * nfs_page_group_destroy is called */
+		kref_get(&req->wb_head->wb_kref);
+
 		/* grab extra ref if head request has extra ref from
 		 * the write/commit path to handle handoff between write
 		 * and commit lists */
-		if (test_bit(PG_INODE_REF, &prev->wb_head->wb_flags))
+		if (test_bit(PG_INODE_REF, &prev->wb_head->wb_flags)) {
+			set_bit(PG_INODE_REF, &req->wb_flags);
 			kref_get(&req->wb_kref);
+		}
 	}
 }
 
@@ -269,6 +276,10 @@ nfs_page_group_destroy(struct kref *kref)
 	struct nfs_page *req = container_of(kref, struct nfs_page, wb_kref);
 	struct nfs_page *tmp, *next;
 
+	/* subrequests must release the ref on the head request */
+	if (req->wb_head != req)
+		nfs_release_request(req->wb_head);
+
 	if (!nfs_page_group_sync_on_bit(req, PG_TEARDOWN))
 		return;
 
@@ -394,7 +405,7 @@ static void nfs_clear_request(struct nfs_page *req)
  *
  * Note: Should never be called with the spinlock held!
  */
-static void nfs_free_request(struct nfs_page *req)
+void nfs_free_request(struct nfs_page *req)
 {
 	WARN_ON_ONCE(req->wb_this_page != req);
 
@@ -425,9 +436,8 @@ void nfs_release_request(struct nfs_page *req)
 int
 nfs_wait_on_request(struct nfs_page *req)
 {
-	return wait_on_bit(&req->wb_flags, PG_BUSY,
-			nfs_wait_bit_uninterruptible,
-			TASK_UNINTERRUPTIBLE);
+	return wait_on_bit_io(&req->wb_flags, PG_BUSY,
+			      TASK_UNINTERRUPTIBLE);
 }
 
 /*
@@ -452,123 +462,72 @@ size_t nfs_generic_pg_test(struct nfs_pageio_descriptor *desc,
 }
 EXPORT_SYMBOL_GPL(nfs_generic_pg_test);
 
-static inline struct nfs_rw_header *NFS_RW_HEADER(struct nfs_pgio_header *hdr)
-{
-	return container_of(hdr, struct nfs_rw_header, header);
-}
-
-/**
- * nfs_rw_header_alloc - Allocate a header for a read or write
- * @ops: Read or write function vector
- */
-struct nfs_rw_header *nfs_rw_header_alloc(const struct nfs_rw_ops *ops)
+struct nfs_pgio_header *nfs_pgio_header_alloc(const struct nfs_rw_ops *ops)
 {
-	struct nfs_rw_header *header = ops->rw_alloc_header();
-
-	if (header) {
-		struct nfs_pgio_header *hdr = &header->header;
+	struct nfs_pgio_header *hdr = ops->rw_alloc_header();
 
+	if (hdr) {
 		INIT_LIST_HEAD(&hdr->pages);
 		spin_lock_init(&hdr->lock);
-		atomic_set(&hdr->refcnt, 0);
 		hdr->rw_ops = ops;
 	}
-	return header;
+	return hdr;
 }
-EXPORT_SYMBOL_GPL(nfs_rw_header_alloc);
+EXPORT_SYMBOL_GPL(nfs_pgio_header_alloc);
 
 /*
- * nfs_rw_header_free - Free a read or write header
+ * nfs_pgio_header_free - Free a read or write header
  * @hdr: The header to free
  */
-void nfs_rw_header_free(struct nfs_pgio_header *hdr)
-{
-	hdr->rw_ops->rw_free_header(NFS_RW_HEADER(hdr));
-}
-EXPORT_SYMBOL_GPL(nfs_rw_header_free);
-
-/**
- * nfs_pgio_data_alloc - Allocate pageio data
- * @hdr: The header making a request
- * @pagecount: Number of pages to create
- */
-static struct nfs_pgio_data *nfs_pgio_data_alloc(struct nfs_pgio_header *hdr,
-						 unsigned int pagecount)
+void nfs_pgio_header_free(struct nfs_pgio_header *hdr)
 {
-	struct nfs_pgio_data *data, *prealloc;
-
-	prealloc = &NFS_RW_HEADER(hdr)->rpc_data;
-	if (prealloc->header == NULL)
-		data = prealloc;
-	else
-		data = kzalloc(sizeof(*data), GFP_KERNEL);
-	if (!data)
-		goto out;
-
-	if (nfs_pgarray_set(&data->pages, pagecount)) {
-		data->header = hdr;
-		atomic_inc(&hdr->refcnt);
-	} else {
-		if (data != prealloc)
-			kfree(data);
-		data = NULL;
-	}
-out:
-	return data;
+	hdr->rw_ops->rw_free_header(hdr);
 }
+EXPORT_SYMBOL_GPL(nfs_pgio_header_free);
 
 /**
- * nfs_pgio_data_release - Properly free pageio data
- * @data: The data to release
+ * nfs_pgio_data_destroy - make @hdr suitable for reuse
+ *
+ * Frees memory and releases refs from nfs_generic_pgio, so that it may
+ * be called again.
+ *
+ * @hdr: A header that has had nfs_generic_pgio called
  */
-void nfs_pgio_data_release(struct nfs_pgio_data *data)
+void nfs_pgio_data_destroy(struct nfs_pgio_header *hdr)
 {
-	struct nfs_pgio_header *hdr = data->header;
-	struct nfs_rw_header *pageio_header = NFS_RW_HEADER(hdr);
-
-	put_nfs_open_context(data->args.context);
-	if (data->pages.pagevec != data->pages.page_array)
-		kfree(data->pages.pagevec);
-	if (data == &pageio_header->rpc_data) {
-		data->header = NULL;
-		data = NULL;
-	}
-	if (atomic_dec_and_test(&hdr->refcnt))
-		hdr->completion_ops->completion(hdr);
-	/* Note: we only free the rpc_task after callbacks are done.
-	 * See the comment in rpc_free_task() for why
-	 */
-	kfree(data);
+	put_nfs_open_context(hdr->args.context);
+	if (hdr->page_array.pagevec != hdr->page_array.page_array)
+		kfree(hdr->page_array.pagevec);
 }
-EXPORT_SYMBOL_GPL(nfs_pgio_data_release);
+EXPORT_SYMBOL_GPL(nfs_pgio_data_destroy);
 
 /**
  * nfs_pgio_rpcsetup - Set up arguments for a pageio call
- * @data: The pageio data
+ * @hdr: The pageio hdr
  * @count: Number of bytes to read
  * @offset: Initial offset
  * @how: How to commit data (writes only)
  * @cinfo: Commit information for the call (writes only)
  */
-static void nfs_pgio_rpcsetup(struct nfs_pgio_data *data,
+static void nfs_pgio_rpcsetup(struct nfs_pgio_header *hdr,
 			      unsigned int count, unsigned int offset,
 			      int how, struct nfs_commit_info *cinfo)
 {
-	struct nfs_page *req = data->header->req;
+	struct nfs_page *req = hdr->req;
 
 	/* Set up the RPC argument and reply structs
-	 * NB: take care not to mess about with data->commit et al. */
+	 * NB: take care not to mess about with hdr->commit et al. */
 
-	data->args.fh     = NFS_FH(data->header->inode);
-	data->args.offset = req_offset(req) + offset;
+	hdr->args.fh     = NFS_FH(hdr->inode);
+	hdr->args.offset = req_offset(req) + offset;
 	/* pnfs_set_layoutcommit needs this */
-	data->mds_offset = data->args.offset;
-	data->args.pgbase = req->wb_pgbase + offset;
-	data->args.pages  = data->pages.pagevec;
-	data->args.count  = count;
-	data->args.context = get_nfs_open_context(req->wb_context);
-	data->args.lock_context = req->wb_lock_context;
-	data->args.stable  = NFS_UNSTABLE;
+	hdr->mds_offset = hdr->args.offset;
+	hdr->args.pgbase = req->wb_pgbase + offset;
+	hdr->args.pages  = hdr->page_array.pagevec;
+	hdr->args.count  = count;
+	hdr->args.context = get_nfs_open_context(req->wb_context);
+	hdr->args.lock_context = req->wb_lock_context;
+	hdr->args.stable  = NFS_UNSTABLE;
 	switch (how & (FLUSH_STABLE | FLUSH_COND_STABLE)) {
 	case 0:
 		break;
@@ -576,59 +535,59 @@ static void nfs_pgio_rpcsetup(struct nfs_pgio_data *data,
 		if (nfs_reqs_to_commit(cinfo))
 			break;
 	default:
-		data->args.stable = NFS_FILE_SYNC;
+		hdr->args.stable = NFS_FILE_SYNC;
 	}
 
-	data->res.fattr   = &data->fattr;
-	data->res.count   = count;
-	data->res.eof     = 0;
-	data->res.verf    = &data->verf;
-	nfs_fattr_init(&data->fattr);
+	hdr->res.fattr   = &hdr->fattr;
+	hdr->res.count   = count;
+	hdr->res.eof     = 0;
+	hdr->res.verf    = &hdr->verf;
+	nfs_fattr_init(&hdr->fattr);
 }
 
 /**
- * nfs_pgio_prepare - Prepare pageio data to go over the wire
+ * nfs_pgio_prepare - Prepare pageio hdr to go over the wire
  * @task: The current task
- * @calldata: pageio data to prepare
+ * @calldata: pageio header to prepare
  */
 static void nfs_pgio_prepare(struct rpc_task *task, void *calldata)
 {
-	struct nfs_pgio_data *data = calldata;
+	struct nfs_pgio_header *hdr = calldata;
 	int err;
-	err = NFS_PROTO(data->header->inode)->pgio_rpc_prepare(task, data);
+	err = NFS_PROTO(hdr->inode)->pgio_rpc_prepare(task, hdr);
 	if (err)
 		rpc_exit(task, err);
 }
 
-int nfs_initiate_pgio(struct rpc_clnt *clnt, struct nfs_pgio_data *data,
+int nfs_initiate_pgio(struct rpc_clnt *clnt, struct nfs_pgio_header *hdr,
 		      const struct rpc_call_ops *call_ops, int how, int flags)
 {
 	struct rpc_task *task;
 	struct rpc_message msg = {
-		.rpc_argp = &data->args,
-		.rpc_resp = &data->res,
-		.rpc_cred = data->header->cred,
+		.rpc_argp = &hdr->args,
+		.rpc_resp = &hdr->res,
+		.rpc_cred = hdr->cred,
 	};
 	struct rpc_task_setup task_setup_data = {
 		.rpc_client = clnt,
-		.task = &data->task,
+		.task = &hdr->task,
 		.rpc_message = &msg,
 		.callback_ops = call_ops,
-		.callback_data = data,
+		.callback_data = hdr,
 		.workqueue = nfsiod_workqueue,
 		.flags = RPC_TASK_ASYNC | flags,
 	};
 	int ret = 0;
 
-	data->header->rw_ops->rw_initiate(data, &msg, &task_setup_data, how);
+	hdr->rw_ops->rw_initiate(hdr, &msg, &task_setup_data, how);
 
 	dprintk("NFS: %5u initiated pgio call "
 		"(req %s/%llu, %u bytes @ offset %llu)\n",
-		data->task.tk_pid,
-		data->header->inode->i_sb->s_id,
-		(unsigned long long)NFS_FILEID(data->header->inode),
-		data->args.count,
-		(unsigned long long)data->args.offset);
+		hdr->task.tk_pid,
+		hdr->inode->i_sb->s_id,
+		(unsigned long long)NFS_FILEID(hdr->inode),
+		hdr->args.count,
+		(unsigned long long)hdr->args.offset);
 
 	task = rpc_run_task(&task_setup_data);
 	if (IS_ERR(task)) {
@@ -655,22 +614,23 @@ static int nfs_pgio_error(struct nfs_pageio_descriptor *desc,
 			  struct nfs_pgio_header *hdr)
 {
 	set_bit(NFS_IOHDR_REDO, &hdr->flags);
-	nfs_pgio_data_release(hdr->data);
-	hdr->data = NULL;
+	nfs_pgio_data_destroy(hdr);
+	hdr->completion_ops->completion(hdr);
 	desc->pg_completion_ops->error_cleanup(&desc->pg_list);
 	return -ENOMEM;
 }
 
 /**
  * nfs_pgio_release - Release pageio data
- * @calldata: The pageio data to release
+ * @calldata: The pageio header to release
  */
 static void nfs_pgio_release(void *calldata)
 {
-	struct nfs_pgio_data *data = calldata;
-	if (data->header->rw_ops->rw_release)
-		data->header->rw_ops->rw_release(data);
-	nfs_pgio_data_release(data);
+	struct nfs_pgio_header *hdr = calldata;
+	if (hdr->rw_ops->rw_release)
+		hdr->rw_ops->rw_release(hdr);
+	nfs_pgio_data_destroy(hdr);
+	hdr->completion_ops->completion(hdr);
 }
 
 /**
@@ -711,22 +671,22 @@ EXPORT_SYMBOL_GPL(nfs_pageio_init);
 /**
  * nfs_pgio_result - Basic pageio error handling
  * @task: The task that ran
- * @calldata: Pageio data to check
+ * @calldata: Pageio header to check
  */
 static void nfs_pgio_result(struct rpc_task *task, void *calldata)
 {
-	struct nfs_pgio_data *data = calldata;
-	struct inode *inode = data->header->inode;
+	struct nfs_pgio_header *hdr = calldata;
+	struct inode *inode = hdr->inode;
 
 	dprintk("NFS: %s: %5u, (status %d)\n", __func__,
 		task->tk_pid, task->tk_status);
 
-	if (data->header->rw_ops->rw_done(task, data, inode) != 0)
+	if (hdr->rw_ops->rw_done(task, hdr, inode) != 0)
 		return;
 	if (task->tk_status < 0)
-		nfs_set_pgio_error(data->header, task->tk_status, data->args.offset);
+		nfs_set_pgio_error(hdr, task->tk_status, hdr->args.offset);
 	else
-		data->header->rw_ops->rw_result(task, data);
+		hdr->rw_ops->rw_result(task, hdr);
 }
 
 /*
@@ -742,17 +702,16 @@ int nfs_generic_pgio(struct nfs_pageio_descriptor *desc,
 {
 	struct nfs_page		*req;
 	struct page		**pages;
-	struct nfs_pgio_data	*data;
 	struct list_head *head = &desc->pg_list;
 	struct nfs_commit_info cinfo;
+	unsigned int pagecount;
 
-	data = nfs_pgio_data_alloc(hdr, nfs_page_array_len(desc->pg_base,
-							   desc->pg_count));
-	if (!data)
+	pagecount = nfs_page_array_len(desc->pg_base, desc->pg_count);
+	if (!nfs_pgarray_set(&hdr->page_array, pagecount))
 		return nfs_pgio_error(desc, hdr);
 
 	nfs_init_cinfo(&cinfo, desc->pg_inode, desc->pg_dreq);
-	pages = data->pages.pagevec;
+	pages = hdr->page_array.pagevec;
 	while (!list_empty(head)) {
 		req = nfs_list_entry(head->next);
 		nfs_list_remove_request(req);
@@ -765,8 +724,7 @@ int nfs_generic_pgio(struct nfs_pageio_descriptor *desc,
 		desc->pg_ioflags &= ~FLUSH_COND_STABLE;
 
 	/* Set up the argument struct */
-	nfs_pgio_rpcsetup(data, desc->pg_count, 0, desc->pg_ioflags, &cinfo);
-	hdr->data = data;
+	nfs_pgio_rpcsetup(hdr, desc->pg_count, 0, desc->pg_ioflags, &cinfo);
 	desc->pg_rpc_callops = &nfs_pgio_common_ops;
 	return 0;
 }
@@ -774,25 +732,20 @@ EXPORT_SYMBOL_GPL(nfs_generic_pgio);
 
 static int nfs_generic_pg_pgios(struct nfs_pageio_descriptor *desc)
 {
-	struct nfs_rw_header *rw_hdr;
 	struct nfs_pgio_header *hdr;
 	int ret;
 
-	rw_hdr = nfs_rw_header_alloc(desc->pg_rw_ops);
-	if (!rw_hdr) {
+	hdr = nfs_pgio_header_alloc(desc->pg_rw_ops);
+	if (!hdr) {
 		desc->pg_completion_ops->error_cleanup(&desc->pg_list);
 		return -ENOMEM;
 	}
-	hdr = &rw_hdr->header;
-	nfs_pgheader_init(desc, hdr, nfs_rw_header_free);
-	atomic_inc(&hdr->refcnt);
+	nfs_pgheader_init(desc, hdr, nfs_pgio_header_free);
 	ret = nfs_generic_pgio(desc, hdr);
 	if (ret == 0)
 		ret = nfs_initiate_pgio(NFS_CLIENT(hdr->inode),
-					hdr->data, desc->pg_rpc_callops,
+					hdr, desc->pg_rpc_callops,
 					desc->pg_ioflags, 0);
-	if (atomic_dec_and_test(&hdr->refcnt))
-		hdr->completion_ops->completion(hdr);
 	return ret;
 }
 
@@ -905,8 +858,13 @@ static int __nfs_pageio_add_request(struct nfs_pageio_descriptor *desc,
 	struct nfs_page *subreq;
 	unsigned int bytes_left = 0;
 	unsigned int offset, pgbase;
+	int ret;
 
-	nfs_page_group_lock(req);
+	ret = nfs_page_group_lock(req, false);
+	if (ret < 0) {
+		desc->pg_error = ret;
+		return 0;
+	}
 
 	subreq = req;
 	bytes_left = subreq->wb_bytes;
@@ -925,11 +883,14 @@ static int __nfs_pageio_add_request(struct nfs_pageio_descriptor *desc,
 			nfs_pageio_doio(desc);
 			if (desc->pg_error < 0)
 				return 0;
-			desc->pg_moreio = 0;
 			if (desc->pg_recoalesce)
 				return 0;
 			/* retry add_request for this subreq */
-			nfs_page_group_lock(req);
+			ret = nfs_page_group_lock(req, false);
+			if (ret < 0) {
+				desc->pg_error = ret;
+				return 0;
+			}
 			continue;
 		}
 
@@ -972,6 +933,7 @@ static int nfs_do_recoalesce(struct nfs_pageio_descriptor *desc)
 		desc->pg_count = 0;
 		desc->pg_base = 0;
 		desc->pg_recoalesce = 0;
+		desc->pg_moreio = 0;
 
 		while (!list_empty(&head)) {
 			struct nfs_page *req;
@@ -1003,7 +965,38 @@ int nfs_pageio_add_request(struct nfs_pageio_descriptor *desc,
 	} while (ret);
 	return ret;
 }
-EXPORT_SYMBOL_GPL(nfs_pageio_add_request);
+
+/*
+ * nfs_pageio_resend - Transfer requests to new descriptor and resend
+ * @hdr - the pgio header to move request from
+ * @desc - the pageio descriptor to add requests to
+ *
+ * Try to move each request (nfs_page) from @hdr to @desc then attempt
+ * to send them.
+ *
+ * Returns 0 on success and < 0 on error.
+ */
+int nfs_pageio_resend(struct nfs_pageio_descriptor *desc,
+		      struct nfs_pgio_header *hdr)
+{
+	LIST_HEAD(failed);
+
+	desc->pg_dreq = hdr->dreq;
+	while (!list_empty(&hdr->pages)) {
+		struct nfs_page *req = nfs_list_entry(hdr->pages.next);
+
+		nfs_list_remove_request(req);
+		if (!nfs_pageio_add_request(desc, req))
+			nfs_list_add_request(req, &failed);
+	}
+	nfs_pageio_complete(desc);
+	if (!list_empty(&failed)) {
+		list_move(&failed, &hdr->pages);
+		return -EIO;
+	}
+	return 0;
+}
+EXPORT_SYMBOL_GPL(nfs_pageio_resend);
 
 /**
  * nfs_pageio_complete - Complete I/O on an nfs_pageio_descriptor
@@ -1019,7 +1012,6 @@ void nfs_pageio_complete(struct nfs_pageio_descriptor *desc)
 			break;
 	}
 }
-EXPORT_SYMBOL_GPL(nfs_pageio_complete);
 
 /**
  * nfs_pageio_cond_complete - Conditional I/O completion
diff --git a/fs/nfs/pnfs.c b/fs/nfs/pnfs.c
index 6fdcd233d6f7..a3851debf8a2 100644
--- a/fs/nfs/pnfs.c
+++ b/fs/nfs/pnfs.c
@@ -361,6 +361,23 @@ pnfs_put_lseg(struct pnfs_layout_segment *lseg)
 }
 EXPORT_SYMBOL_GPL(pnfs_put_lseg);
 
+static void pnfs_put_lseg_async_work(struct work_struct *work)
+{
+	struct pnfs_layout_segment *lseg;
+
+	lseg = container_of(work, struct pnfs_layout_segment, pls_work);
+
+	pnfs_put_lseg(lseg);
+}
+
+void
+pnfs_put_lseg_async(struct pnfs_layout_segment *lseg)
+{
+	INIT_WORK(&lseg->pls_work, pnfs_put_lseg_async_work);
+	schedule_work(&lseg->pls_work);
+}
+EXPORT_SYMBOL_GPL(pnfs_put_lseg_async);
+
 static u64
 end_offset(u64 start, u64 len)
 {
@@ -1470,41 +1487,19 @@ pnfs_generic_pg_test(struct nfs_pageio_descriptor *pgio, struct nfs_page *prev,
 }
 EXPORT_SYMBOL_GPL(pnfs_generic_pg_test);
 
-int pnfs_write_done_resend_to_mds(struct inode *inode,
-				struct list_head *head,
-				const struct nfs_pgio_completion_ops *compl_ops,
-				struct nfs_direct_req *dreq)
+int pnfs_write_done_resend_to_mds(struct nfs_pgio_header *hdr)
 {
 	struct nfs_pageio_descriptor pgio;
-	LIST_HEAD(failed);
 
 	/* Resend all requests through the MDS */
-	nfs_pageio_init_write(&pgio, inode, FLUSH_STABLE, true, compl_ops);
-	pgio.pg_dreq = dreq;
-	while (!list_empty(head)) {
-		struct nfs_page *req = nfs_list_entry(head->next);
-
-		nfs_list_remove_request(req);
-		if (!nfs_pageio_add_request(&pgio, req))
-			nfs_list_add_request(req, &failed);
-	}
-	nfs_pageio_complete(&pgio);
-
-	if (!list_empty(&failed)) {
-		/* For some reason our attempt to resend pages. Mark the
-		 * overall send request as having failed, and let
-		 * nfs_writeback_release_full deal with the error.
-		 */
-		list_move(&failed, head);
-		return -EIO;
-	}
-	return 0;
+	nfs_pageio_init_write(&pgio, hdr->inode, FLUSH_STABLE, true,
+			      hdr->completion_ops);
+	return nfs_pageio_resend(&pgio, hdr);
 }
 EXPORT_SYMBOL_GPL(pnfs_write_done_resend_to_mds);
 
-static void pnfs_ld_handle_write_error(struct nfs_pgio_data *data)
+static void pnfs_ld_handle_write_error(struct nfs_pgio_header *hdr)
 {
-	struct nfs_pgio_header *hdr = data->header;
 
 	dprintk("pnfs write error = %d\n", hdr->pnfs_error);
 	if (NFS_SERVER(hdr->inode)->pnfs_curr_ld->flags &
@@ -1512,50 +1507,42 @@ static void pnfs_ld_handle_write_error(struct nfs_pgio_data *data)
 		pnfs_return_layout(hdr->inode);
 	}
 	if (!test_and_set_bit(NFS_IOHDR_REDO, &hdr->flags))
-		data->task.tk_status = pnfs_write_done_resend_to_mds(hdr->inode,
-							&hdr->pages,
-							hdr->completion_ops,
-							hdr->dreq);
+		hdr->task.tk_status = pnfs_write_done_resend_to_mds(hdr);
 }
 
 /*
  * Called by non rpc-based layout drivers
  */
-void pnfs_ld_write_done(struct nfs_pgio_data *data)
+void pnfs_ld_write_done(struct nfs_pgio_header *hdr)
 {
-	struct nfs_pgio_header *hdr = data->header;
-
-	trace_nfs4_pnfs_write(data, hdr->pnfs_error);
+	trace_nfs4_pnfs_write(hdr, hdr->pnfs_error);
 	if (!hdr->pnfs_error) {
-		pnfs_set_layoutcommit(data);
-		hdr->mds_ops->rpc_call_done(&data->task, data);
+		pnfs_set_layoutcommit(hdr);
+		hdr->mds_ops->rpc_call_done(&hdr->task, hdr);
 	} else
-		pnfs_ld_handle_write_error(data);
-	hdr->mds_ops->rpc_release(data);
+		pnfs_ld_handle_write_error(hdr);
+	hdr->mds_ops->rpc_release(hdr);
 }
 EXPORT_SYMBOL_GPL(pnfs_ld_write_done);
 
 static void
 pnfs_write_through_mds(struct nfs_pageio_descriptor *desc,
-		struct nfs_pgio_data *data)
+		struct nfs_pgio_header *hdr)
 {
-	struct nfs_pgio_header *hdr = data->header;
-
 	if (!test_and_set_bit(NFS_IOHDR_REDO, &hdr->flags)) {
 		list_splice_tail_init(&hdr->pages, &desc->pg_list);
 		nfs_pageio_reset_write_mds(desc);
 		desc->pg_recoalesce = 1;
 	}
-	nfs_pgio_data_release(data);
+	nfs_pgio_data_destroy(hdr);
 }
 
 static enum pnfs_try_status
-pnfs_try_to_write_data(struct nfs_pgio_data *wdata,
+pnfs_try_to_write_data(struct nfs_pgio_header *hdr,
 			const struct rpc_call_ops *call_ops,
 			struct pnfs_layout_segment *lseg,
 			int how)
 {
-	struct nfs_pgio_header *hdr = wdata->header;
 	struct inode *inode = hdr->inode;
 	enum pnfs_try_status trypnfs;
 	struct nfs_server *nfss = NFS_SERVER(inode);
@@ -1563,8 +1550,8 @@ pnfs_try_to_write_data(struct nfs_pgio_data *wdata,
 	hdr->mds_ops = call_ops;
 
 	dprintk("%s: Writing ino:%lu %u@%llu (how %d)\n", __func__,
-		inode->i_ino, wdata->args.count, wdata->args.offset, how);
-	trypnfs = nfss->pnfs_curr_ld->write_pagelist(wdata, how);
+		inode->i_ino, hdr->args.count, hdr->args.offset, how);
+	trypnfs = nfss->pnfs_curr_ld->write_pagelist(hdr, how);
 	if (trypnfs != PNFS_NOT_ATTEMPTED)
 		nfs_inc_stats(inode, NFSIOS_PNFS_WRITE);
 	dprintk("%s End (trypnfs:%d)\n", __func__, trypnfs);
@@ -1575,139 +1562,105 @@ static void
 pnfs_do_write(struct nfs_pageio_descriptor *desc,
 	      struct nfs_pgio_header *hdr, int how)
 {
-	struct nfs_pgio_data *data = hdr->data;
 	const struct rpc_call_ops *call_ops = desc->pg_rpc_callops;
 	struct pnfs_layout_segment *lseg = desc->pg_lseg;
 	enum pnfs_try_status trypnfs;
 
 	desc->pg_lseg = NULL;
-	trypnfs = pnfs_try_to_write_data(data, call_ops, lseg, how);
+	trypnfs = pnfs_try_to_write_data(hdr, call_ops, lseg, how);
 	if (trypnfs == PNFS_NOT_ATTEMPTED)
-		pnfs_write_through_mds(desc, data);
+		pnfs_write_through_mds(desc, hdr);
 	pnfs_put_lseg(lseg);
 }
 
 static void pnfs_writehdr_free(struct nfs_pgio_header *hdr)
 {
 	pnfs_put_lseg(hdr->lseg);
-	nfs_rw_header_free(hdr);
+	nfs_pgio_header_free(hdr);
 }
 EXPORT_SYMBOL_GPL(pnfs_writehdr_free);
 
 int
 pnfs_generic_pg_writepages(struct nfs_pageio_descriptor *desc)
 {
-	struct nfs_rw_header *whdr;
 	struct nfs_pgio_header *hdr;
 	int ret;
 
-	whdr = nfs_rw_header_alloc(desc->pg_rw_ops);
-	if (!whdr) {
+	hdr = nfs_pgio_header_alloc(desc->pg_rw_ops);
+	if (!hdr) {
 		desc->pg_completion_ops->error_cleanup(&desc->pg_list);
 		pnfs_put_lseg(desc->pg_lseg);
 		desc->pg_lseg = NULL;
 		return -ENOMEM;
 	}
-	hdr = &whdr->header;
 	nfs_pgheader_init(desc, hdr, pnfs_writehdr_free);
 	hdr->lseg = pnfs_get_lseg(desc->pg_lseg);
-	atomic_inc(&hdr->refcnt);
 	ret = nfs_generic_pgio(desc, hdr);
 	if (ret != 0) {
 		pnfs_put_lseg(desc->pg_lseg);
 		desc->pg_lseg = NULL;
 	} else
 		pnfs_do_write(desc, hdr, desc->pg_ioflags);
-	if (atomic_dec_and_test(&hdr->refcnt))
-		hdr->completion_ops->completion(hdr);
 	return ret;
 }
 EXPORT_SYMBOL_GPL(pnfs_generic_pg_writepages);
 
-int pnfs_read_done_resend_to_mds(struct inode *inode,
-				struct list_head *head,
-				const struct nfs_pgio_completion_ops *compl_ops,
-				struct nfs_direct_req *dreq)
+int pnfs_read_done_resend_to_mds(struct nfs_pgio_header *hdr)
 {
 	struct nfs_pageio_descriptor pgio;
-	LIST_HEAD(failed);
 
 	/* Resend all requests through the MDS */
-	nfs_pageio_init_read(&pgio, inode, true, compl_ops);
-	pgio.pg_dreq = dreq;
-	while (!list_empty(head)) {
-		struct nfs_page *req = nfs_list_entry(head->next);
-
-		nfs_list_remove_request(req);
-		if (!nfs_pageio_add_request(&pgio, req))
-			nfs_list_add_request(req, &failed);
-	}
-	nfs_pageio_complete(&pgio);
-
-	if (!list_empty(&failed)) {
-		list_move(&failed, head);
-		return -EIO;
-	}
-	return 0;
+	nfs_pageio_init_read(&pgio, hdr->inode, true, hdr->completion_ops);
+	return nfs_pageio_resend(&pgio, hdr);
 }
 EXPORT_SYMBOL_GPL(pnfs_read_done_resend_to_mds);
 
-static void pnfs_ld_handle_read_error(struct nfs_pgio_data *data)
+static void pnfs_ld_handle_read_error(struct nfs_pgio_header *hdr)
 {
-	struct nfs_pgio_header *hdr = data->header;
-
 	dprintk("pnfs read error = %d\n", hdr->pnfs_error);
 	if (NFS_SERVER(hdr->inode)->pnfs_curr_ld->flags &
 	    PNFS_LAYOUTRET_ON_ERROR) {
 		pnfs_return_layout(hdr->inode);
 	}
 	if (!test_and_set_bit(NFS_IOHDR_REDO, &hdr->flags))
-		data->task.tk_status = pnfs_read_done_resend_to_mds(hdr->inode,
-							&hdr->pages,
-							hdr->completion_ops,
-							hdr->dreq);
+		hdr->task.tk_status = pnfs_read_done_resend_to_mds(hdr);
 }
 
 /*
  * Called by non rpc-based layout drivers
  */
-void pnfs_ld_read_done(struct nfs_pgio_data *data)
+void pnfs_ld_read_done(struct nfs_pgio_header *hdr)
 {
-	struct nfs_pgio_header *hdr = data->header;
-
-	trace_nfs4_pnfs_read(data, hdr->pnfs_error);
+	trace_nfs4_pnfs_read(hdr, hdr->pnfs_error);
 	if (likely(!hdr->pnfs_error)) {
-		__nfs4_read_done_cb(data);
-		hdr->mds_ops->rpc_call_done(&data->task, data);
+		__nfs4_read_done_cb(hdr);
+		hdr->mds_ops->rpc_call_done(&hdr->task, hdr);
 	} else
-		pnfs_ld_handle_read_error(data);
-	hdr->mds_ops->rpc_release(data);
+		pnfs_ld_handle_read_error(hdr);
+	hdr->mds_ops->rpc_release(hdr);
 }
 EXPORT_SYMBOL_GPL(pnfs_ld_read_done);
 
 static void
 pnfs_read_through_mds(struct nfs_pageio_descriptor *desc,
-		struct nfs_pgio_data *data)
+		struct nfs_pgio_header *hdr)
 {
-	struct nfs_pgio_header *hdr = data->header;
-
 	if (!test_and_set_bit(NFS_IOHDR_REDO, &hdr->flags)) {
 		list_splice_tail_init(&hdr->pages, &desc->pg_list);
 		nfs_pageio_reset_read_mds(desc);
 		desc->pg_recoalesce = 1;
 	}
-	nfs_pgio_data_release(data);
+	nfs_pgio_data_destroy(hdr);
 }
 
 /*
  * Call the appropriate parallel I/O subsystem read function.
  */
 static enum pnfs_try_status
-pnfs_try_to_read_data(struct nfs_pgio_data *rdata,
+pnfs_try_to_read_data(struct nfs_pgio_header *hdr,
 		       const struct rpc_call_ops *call_ops,
 		       struct pnfs_layout_segment *lseg)
 {
-	struct nfs_pgio_header *hdr = rdata->header;
 	struct inode *inode = hdr->inode;
 	struct nfs_server *nfss = NFS_SERVER(inode);
 	enum pnfs_try_status trypnfs;
@@ -1715,9 +1668,9 @@ pnfs_try_to_read_data(struct nfs_pgio_data *rdata,
 	hdr->mds_ops = call_ops;
 
 	dprintk("%s: Reading ino:%lu %u@%llu\n",
-		__func__, inode->i_ino, rdata->args.count, rdata->args.offset);
+		__func__, inode->i_ino, hdr->args.count, hdr->args.offset);
 
-	trypnfs = nfss->pnfs_curr_ld->read_pagelist(rdata);
+	trypnfs = nfss->pnfs_curr_ld->read_pagelist(hdr);
 	if (trypnfs != PNFS_NOT_ATTEMPTED)
 		nfs_inc_stats(inode, NFSIOS_PNFS_READ);
 	dprintk("%s End (trypnfs:%d)\n", __func__, trypnfs);
@@ -1727,52 +1680,46 @@ pnfs_try_to_read_data(struct nfs_pgio_data *rdata,
 static void
 pnfs_do_read(struct nfs_pageio_descriptor *desc, struct nfs_pgio_header *hdr)
 {
-	struct nfs_pgio_data *data = hdr->data;
 	const struct rpc_call_ops *call_ops = desc->pg_rpc_callops;
 	struct pnfs_layout_segment *lseg = desc->pg_lseg;
 	enum pnfs_try_status trypnfs;
 
 	desc->pg_lseg = NULL;
-	trypnfs = pnfs_try_to_read_data(data, call_ops, lseg);
+	trypnfs = pnfs_try_to_read_data(hdr, call_ops, lseg);
 	if (trypnfs == PNFS_NOT_ATTEMPTED)
-		pnfs_read_through_mds(desc, data);
+		pnfs_read_through_mds(desc, hdr);
 	pnfs_put_lseg(lseg);
 }
 
 static void pnfs_readhdr_free(struct nfs_pgio_header *hdr)
 {
 	pnfs_put_lseg(hdr->lseg);
-	nfs_rw_header_free(hdr);
+	nfs_pgio_header_free(hdr);
 }
 EXPORT_SYMBOL_GPL(pnfs_readhdr_free);
 
 int
 pnfs_generic_pg_readpages(struct nfs_pageio_descriptor *desc)
 {
-	struct nfs_rw_header *rhdr;
 	struct nfs_pgio_header *hdr;
 	int ret;
 
-	rhdr = nfs_rw_header_alloc(desc->pg_rw_ops);
-	if (!rhdr) {
+	hdr = nfs_pgio_header_alloc(desc->pg_rw_ops);
+	if (!hdr) {
 		desc->pg_completion_ops->error_cleanup(&desc->pg_list);
 		ret = -ENOMEM;
 		pnfs_put_lseg(desc->pg_lseg);
 		desc->pg_lseg = NULL;
 		return ret;
 	}
-	hdr = &rhdr->header;
 	nfs_pgheader_init(desc, hdr, pnfs_readhdr_free);
 	hdr->lseg = pnfs_get_lseg(desc->pg_lseg);
-	atomic_inc(&hdr->refcnt);
 	ret = nfs_generic_pgio(desc, hdr);
 	if (ret != 0) {
 		pnfs_put_lseg(desc->pg_lseg);
 		desc->pg_lseg = NULL;
 	} else
 		pnfs_do_read(desc, hdr);
-	if (atomic_dec_and_test(&hdr->refcnt))
-		hdr->completion_ops->completion(hdr);
 	return ret;
 }
 EXPORT_SYMBOL_GPL(pnfs_generic_pg_readpages);
@@ -1820,12 +1767,11 @@ void pnfs_set_lo_fail(struct pnfs_layout_segment *lseg)
 EXPORT_SYMBOL_GPL(pnfs_set_lo_fail);
 
 void
-pnfs_set_layoutcommit(struct nfs_pgio_data *wdata)
+pnfs_set_layoutcommit(struct nfs_pgio_header *hdr)
 {
-	struct nfs_pgio_header *hdr = wdata->header;
 	struct inode *inode = hdr->inode;
 	struct nfs_inode *nfsi = NFS_I(inode);
-	loff_t end_pos = wdata->mds_offset + wdata->res.count;
+	loff_t end_pos = hdr->mds_offset + hdr->res.count;
 	bool mark_as_dirty = false;
 
 	spin_lock(&inode->i_lock);
@@ -1885,7 +1831,7 @@ pnfs_layoutcommit_inode(struct inode *inode, bool sync)
 	if (test_and_set_bit(NFS_INO_LAYOUTCOMMITTING, &nfsi->flags)) {
 		if (!sync)
 			goto out;
-		status = wait_on_bit_lock(&nfsi->flags,
+		status = wait_on_bit_lock_action(&nfsi->flags,
 				NFS_INO_LAYOUTCOMMITTING,
 				nfs_wait_bit_killable,
 				TASK_KILLABLE);
diff --git a/fs/nfs/pnfs.h b/fs/nfs/pnfs.h
index 4fb309a2b4c4..aca3dff5dae6 100644
--- a/fs/nfs/pnfs.h
+++ b/fs/nfs/pnfs.h
@@ -32,6 +32,7 @@
 
 #include <linux/nfs_fs.h>
 #include <linux/nfs_page.h>
+#include <linux/workqueue.h>
 
 enum {
 	NFS_LSEG_VALID = 0,	/* cleared when lseg is recalled/returned */
@@ -46,6 +47,7 @@ struct pnfs_layout_segment {
 	atomic_t pls_refcount;
 	unsigned long pls_flags;
 	struct pnfs_layout_hdr *pls_layout;
+	struct work_struct pls_work;
 };
 
 enum pnfs_try_status {
@@ -104,6 +106,8 @@ struct pnfs_layoutdriver_type {
 				  int max);
 	void (*recover_commit_reqs) (struct list_head *list,
 				     struct nfs_commit_info *cinfo);
+	struct nfs_page * (*search_commit_reqs)(struct nfs_commit_info *cinfo,
+						struct page *page);
 	int (*commit_pagelist)(struct inode *inode,
 			       struct list_head *mds_pages,
 			       int how,
@@ -113,8 +117,8 @@ struct pnfs_layoutdriver_type {
 	 * Return PNFS_ATTEMPTED to indicate the layout code has attempted
 	 * I/O, else return PNFS_NOT_ATTEMPTED to fall back to normal NFS
 	 */
-	enum pnfs_try_status (*read_pagelist) (struct nfs_pgio_data *nfs_data);
-	enum pnfs_try_status (*write_pagelist) (struct nfs_pgio_data *nfs_data, int how);
+	enum pnfs_try_status (*read_pagelist)(struct nfs_pgio_header *);
+	enum pnfs_try_status (*write_pagelist)(struct nfs_pgio_header *, int);
 
 	void (*free_deviceid_node) (struct nfs4_deviceid_node *);
 
@@ -179,6 +183,7 @@ extern int nfs4_proc_layoutreturn(struct nfs4_layoutreturn *lrp);
 /* pnfs.c */
 void pnfs_get_layout_hdr(struct pnfs_layout_hdr *lo);
 void pnfs_put_lseg(struct pnfs_layout_segment *lseg);
+void pnfs_put_lseg_async(struct pnfs_layout_segment *lseg);
 
 void set_pnfs_layoutdriver(struct nfs_server *, const struct nfs_fh *, u32);
 void unset_pnfs_layoutdriver(struct nfs_server *);
@@ -213,13 +218,13 @@ bool pnfs_roc(struct inode *ino);
 void pnfs_roc_release(struct inode *ino);
 void pnfs_roc_set_barrier(struct inode *ino, u32 barrier);
 bool pnfs_roc_drain(struct inode *ino, u32 *barrier, struct rpc_task *task);
-void pnfs_set_layoutcommit(struct nfs_pgio_data *wdata);
+void pnfs_set_layoutcommit(struct nfs_pgio_header *);
 void pnfs_cleanup_layoutcommit(struct nfs4_layoutcommit_data *data);
 int pnfs_layoutcommit_inode(struct inode *inode, bool sync);
 int _pnfs_return_layout(struct inode *);
 int pnfs_commit_and_return_layout(struct inode *);
-void pnfs_ld_write_done(struct nfs_pgio_data *);
-void pnfs_ld_read_done(struct nfs_pgio_data *);
+void pnfs_ld_write_done(struct nfs_pgio_header *);
+void pnfs_ld_read_done(struct nfs_pgio_header *);
 struct pnfs_layout_segment *pnfs_update_layout(struct inode *ino,
 					       struct nfs_open_context *ctx,
 					       loff_t pos,
@@ -228,12 +233,8 @@ struct pnfs_layout_segment *pnfs_update_layout(struct inode *ino,
 					       gfp_t gfp_flags);
 
 void nfs4_deviceid_mark_client_invalid(struct nfs_client *clp);
-int pnfs_read_done_resend_to_mds(struct inode *inode, struct list_head *head,
-			const struct nfs_pgio_completion_ops *compl_ops,
-			struct nfs_direct_req *dreq);
-int pnfs_write_done_resend_to_mds(struct inode *inode, struct list_head *head,
-			const struct nfs_pgio_completion_ops *compl_ops,
-			struct nfs_direct_req *dreq);
+int pnfs_read_done_resend_to_mds(struct nfs_pgio_header *);
+int pnfs_write_done_resend_to_mds(struct nfs_pgio_header *);
 struct nfs4_threshold *pnfs_mdsthreshold_alloc(void);
 
 /* nfs4_deviceid_flags */
@@ -345,6 +346,17 @@ pnfs_recover_commit_reqs(struct inode *inode, struct list_head *list,
 	NFS_SERVER(inode)->pnfs_curr_ld->recover_commit_reqs(list, cinfo);
 }
 
+static inline struct nfs_page *
+pnfs_search_commit_reqs(struct inode *inode, struct nfs_commit_info *cinfo,
+			struct page *page)
+{
+	struct pnfs_layoutdriver_type *ld = NFS_SERVER(inode)->pnfs_curr_ld;
+
+	if (ld == NULL || ld->search_commit_reqs == NULL)
+		return NULL;
+	return ld->search_commit_reqs(cinfo, page);
+}
+
 /* Should the pNFS client commit and return the layout upon a setattr */
 static inline bool
 pnfs_ld_layoutret_on_setattr(struct inode *inode)
@@ -410,6 +422,10 @@ static inline void pnfs_put_lseg(struct pnfs_layout_segment *lseg)
 {
 }
 
+static inline void pnfs_put_lseg_async(struct pnfs_layout_segment *lseg)
+{
+}
+
 static inline int pnfs_return_layout(struct inode *ino)
 {
 	return 0;
@@ -496,6 +512,13 @@ pnfs_recover_commit_reqs(struct inode *inode, struct list_head *list,
 {
 }
 
+static inline struct nfs_page *
+pnfs_search_commit_reqs(struct inode *inode, struct nfs_commit_info *cinfo,
+			struct page *page)
+{
+	return NULL;
+}
+
 static inline int pnfs_layoutcommit_inode(struct inode *inode, bool sync)
 {
 	return 0;
diff --git a/fs/nfs/proc.c b/fs/nfs/proc.c
index c171ce1a8a30..b09cc23d6f43 100644
--- a/fs/nfs/proc.c
+++ b/fs/nfs/proc.c
@@ -578,46 +578,49 @@ nfs_proc_pathconf(struct nfs_server *server, struct nfs_fh *fhandle,
 	return 0;
 }
 
-static int nfs_read_done(struct rpc_task *task, struct nfs_pgio_data *data)
+static int nfs_read_done(struct rpc_task *task, struct nfs_pgio_header *hdr)
 {
-	struct inode *inode = data->header->inode;
+	struct inode *inode = hdr->inode;
 
 	nfs_invalidate_atime(inode);
 	if (task->tk_status >= 0) {
-		nfs_refresh_inode(inode, data->res.fattr);
+		nfs_refresh_inode(inode, hdr->res.fattr);
 		/* Emulate the eof flag, which isn't normally needed in NFSv2
 		 * as it is guaranteed to always return the file attributes
 		 */
-		if (data->args.offset + data->res.count >= data->res.fattr->size)
-			data->res.eof = 1;
+		if (hdr->args.offset + hdr->res.count >= hdr->res.fattr->size)
+			hdr->res.eof = 1;
 	}
 	return 0;
 }
 
-static void nfs_proc_read_setup(struct nfs_pgio_data *data, struct rpc_message *msg)
+static void nfs_proc_read_setup(struct nfs_pgio_header *hdr,
+				struct rpc_message *msg)
 {
 	msg->rpc_proc = &nfs_procedures[NFSPROC_READ];
 }
 
-static int nfs_proc_pgio_rpc_prepare(struct rpc_task *task, struct nfs_pgio_data *data)
+static int nfs_proc_pgio_rpc_prepare(struct rpc_task *task,
+				     struct nfs_pgio_header *hdr)
 {
 	rpc_call_start(task);
 	return 0;
 }
 
-static int nfs_write_done(struct rpc_task *task, struct nfs_pgio_data *data)
+static int nfs_write_done(struct rpc_task *task, struct nfs_pgio_header *hdr)
 {
-	struct inode *inode = data->header->inode;
+	struct inode *inode = hdr->inode;
 
 	if (task->tk_status >= 0)
-		nfs_post_op_update_inode_force_wcc(inode, data->res.fattr);
+		nfs_post_op_update_inode_force_wcc(inode, hdr->res.fattr);
 	return 0;
 }
 
-static void nfs_proc_write_setup(struct nfs_pgio_data *data, struct rpc_message *msg)
+static void nfs_proc_write_setup(struct nfs_pgio_header *hdr,
+				 struct rpc_message *msg)
 {
 	/* Note: NFSv2 ignores @stable and always uses NFS_FILE_SYNC */
-	data->args.stable = NFS_FILE_SYNC;
+	hdr->args.stable = NFS_FILE_SYNC;
 	msg->rpc_proc = &nfs_procedures[NFSPROC_WRITE];
 }
 
diff --git a/fs/nfs/read.c b/fs/nfs/read.c
index e818a475ca64..beff2769c5c5 100644
--- a/fs/nfs/read.c
+++ b/fs/nfs/read.c
@@ -33,12 +33,12 @@ static const struct nfs_rw_ops nfs_rw_read_ops;
 
 static struct kmem_cache *nfs_rdata_cachep;
 
-static struct nfs_rw_header *nfs_readhdr_alloc(void)
+static struct nfs_pgio_header *nfs_readhdr_alloc(void)
 {
 	return kmem_cache_zalloc(nfs_rdata_cachep, GFP_KERNEL);
 }
 
-static void nfs_readhdr_free(struct nfs_rw_header *rhdr)
+static void nfs_readhdr_free(struct nfs_pgio_header *rhdr)
 {
 	kmem_cache_free(nfs_rdata_cachep, rhdr);
 }
@@ -115,12 +115,6 @@ static void nfs_readpage_release(struct nfs_page *req)
 
 		unlock_page(req->wb_page);
 	}
-
-	dprintk("NFS: read done (%s/%Lu %d@%Ld)\n",
-			req->wb_context->dentry->d_inode->i_sb->s_id,
-			(unsigned long long)NFS_FILEID(req->wb_context->dentry->d_inode),
-			req->wb_bytes,
-			(long long)req_offset(req));
 	nfs_release_request(req);
 }
 
@@ -172,14 +166,15 @@ out:
 	hdr->release(hdr);
 }
 
-static void nfs_initiate_read(struct nfs_pgio_data *data, struct rpc_message *msg,
+static void nfs_initiate_read(struct nfs_pgio_header *hdr,
+			      struct rpc_message *msg,
 			      struct rpc_task_setup *task_setup_data, int how)
 {
-	struct inode *inode = data->header->inode;
+	struct inode *inode = hdr->inode;
 	int swap_flags = IS_SWAPFILE(inode) ? NFS_RPC_SWAPFLAGS : 0;
 
 	task_setup_data->flags |= swap_flags;
-	NFS_PROTO(inode)->read_setup(data, msg);
+	NFS_PROTO(inode)->read_setup(hdr, msg);
 }
 
 static void
@@ -203,14 +198,15 @@ static const struct nfs_pgio_completion_ops nfs_async_read_completion_ops = {
  * This is the callback from RPC telling us whether a reply was
  * received or some error occurred (timeout or socket shutdown).
  */
-static int nfs_readpage_done(struct rpc_task *task, struct nfs_pgio_data *data,
+static int nfs_readpage_done(struct rpc_task *task,
+			     struct nfs_pgio_header *hdr,
 			     struct inode *inode)
 {
-	int status = NFS_PROTO(inode)->read_done(task, data);
+	int status = NFS_PROTO(inode)->read_done(task, hdr);
 	if (status != 0)
 		return status;
 
-	nfs_add_stats(inode, NFSIOS_SERVERREADBYTES, data->res.count);
+	nfs_add_stats(inode, NFSIOS_SERVERREADBYTES, hdr->res.count);
 
 	if (task->tk_status == -ESTALE) {
 		set_bit(NFS_INO_STALE, &NFS_I(inode)->flags);
@@ -219,34 +215,34 @@ static int nfs_readpage_done(struct rpc_task *task, struct nfs_pgio_data *data,
 	return 0;
 }
 
-static void nfs_readpage_retry(struct rpc_task *task, struct nfs_pgio_data *data)
+static void nfs_readpage_retry(struct rpc_task *task,
+			       struct nfs_pgio_header *hdr)
 {
-	struct nfs_pgio_args *argp = &data->args;
-	struct nfs_pgio_res  *resp = &data->res;
+	struct nfs_pgio_args *argp = &hdr->args;
+	struct nfs_pgio_res  *resp = &hdr->res;
 
 	/* This is a short read! */
-	nfs_inc_stats(data->header->inode, NFSIOS_SHORTREAD);
+	nfs_inc_stats(hdr->inode, NFSIOS_SHORTREAD);
 	/* Has the server at least made some progress? */
 	if (resp->count == 0) {
-		nfs_set_pgio_error(data->header, -EIO, argp->offset);
+		nfs_set_pgio_error(hdr, -EIO, argp->offset);
 		return;
 	}
-	/* Yes, so retry the read at the end of the data */
-	data->mds_offset += resp->count;
+	/* Yes, so retry the read at the end of the hdr */
+	hdr->mds_offset += resp->count;
 	argp->offset += resp->count;
 	argp->pgbase += resp->count;
 	argp->count -= resp->count;
 	rpc_restart_call_prepare(task);
 }
 
-static void nfs_readpage_result(struct rpc_task *task, struct nfs_pgio_data *data)
+static void nfs_readpage_result(struct rpc_task *task,
+				struct nfs_pgio_header *hdr)
 {
-	struct nfs_pgio_header *hdr = data->header;
-
-	if (data->res.eof) {
+	if (hdr->res.eof) {
 		loff_t bound;
 
-		bound = data->args.offset + data->res.count;
+		bound = hdr->args.offset + hdr->res.count;
 		spin_lock(&hdr->lock);
 		if (bound < hdr->io_start + hdr->good_bytes) {
 			set_bit(NFS_IOHDR_EOF, &hdr->flags);
@@ -254,8 +250,8 @@ static void nfs_readpage_result(struct rpc_task *task, struct nfs_pgio_data *dat
 			hdr->good_bytes = bound - hdr->io_start;
 		}
 		spin_unlock(&hdr->lock);
-	} else if (data->res.count != data->args.count)
-		nfs_readpage_retry(task, data);
+	} else if (hdr->res.count != hdr->args.count)
+		nfs_readpage_retry(task, hdr);
 }
 
 /*
@@ -404,7 +400,7 @@ out:
 int __init nfs_init_readpagecache(void)
 {
 	nfs_rdata_cachep = kmem_cache_create("nfs_read_data",
-					     sizeof(struct nfs_rw_header),
+					     sizeof(struct nfs_pgio_header),
 					     0, SLAB_HWCACHE_ALIGN,
 					     NULL);
 	if (nfs_rdata_cachep == NULL)
diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index 084af1060d79..e4499d5b51e8 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -1027,8 +1027,7 @@ static bool nfs_auth_info_add(struct nfs_auth_info *auth_info,
 			      rpc_authflavor_t flavor)
 {
 	unsigned int i;
-	unsigned int max_flavor_len = (sizeof(auth_info->flavors) /
-				       sizeof(auth_info->flavors[0]));
+	unsigned int max_flavor_len = ARRAY_SIZE(auth_info->flavors);
 
 	/* make sure this flavor isn't already in the list */
 	for (i = 0; i < auth_info->flavor_len; i++) {
@@ -2180,7 +2179,7 @@ out_no_address:
 	return -EINVAL;
 }
 
-#define NFS_MOUNT_CMP_FLAGMASK ~(NFS_MOUNT_INTR \
+#define NFS_REMOUNT_CMP_FLAGMASK ~(NFS_MOUNT_INTR \
 		| NFS_MOUNT_SECURE \
 		| NFS_MOUNT_TCP \
 		| NFS_MOUNT_VER3 \
@@ -2188,15 +2187,16 @@ out_no_address:
 		| NFS_MOUNT_NONLM \
 		| NFS_MOUNT_BROKEN_SUID \
 		| NFS_MOUNT_STRICTLOCK \
-		| NFS_MOUNT_UNSHARED \
-		| NFS_MOUNT_NORESVPORT \
 		| NFS_MOUNT_LEGACY_INTERFACE)
 
+#define NFS_MOUNT_CMP_FLAGMASK (NFS_REMOUNT_CMP_FLAGMASK & \
+		~(NFS_MOUNT_UNSHARED | NFS_MOUNT_NORESVPORT))
+
 static int
 nfs_compare_remount_data(struct nfs_server *nfss,
 			 struct nfs_parsed_mount_data *data)
 {
-	if ((data->flags ^ nfss->flags) & NFS_MOUNT_CMP_FLAGMASK ||
+	if ((data->flags ^ nfss->flags) & NFS_REMOUNT_CMP_FLAGMASK ||
 	    data->rsize != nfss->rsize ||
 	    data->wsize != nfss->wsize ||
 	    data->version != nfss->nfs_client->rpc_ops->version ||
diff --git a/fs/nfs/write.c b/fs/nfs/write.c
index 3ee5af4e738e..e3b5cf28bdc5 100644
--- a/fs/nfs/write.c
+++ b/fs/nfs/write.c
@@ -46,6 +46,9 @@ static const struct rpc_call_ops nfs_commit_ops;
 static const struct nfs_pgio_completion_ops nfs_async_write_completion_ops;
 static const struct nfs_commit_completion_ops nfs_commit_completion_ops;
 static const struct nfs_rw_ops nfs_rw_write_ops;
+static void nfs_clear_request_commit(struct nfs_page *req);
+static void nfs_init_cinfo_from_inode(struct nfs_commit_info *cinfo,
+				      struct inode *inode);
 
 static struct kmem_cache *nfs_wdata_cachep;
 static mempool_t *nfs_wdata_mempool;
@@ -70,18 +73,18 @@ void nfs_commit_free(struct nfs_commit_data *p)
 }
 EXPORT_SYMBOL_GPL(nfs_commit_free);
 
-static struct nfs_rw_header *nfs_writehdr_alloc(void)
+static struct nfs_pgio_header *nfs_writehdr_alloc(void)
 {
-	struct nfs_rw_header *p = mempool_alloc(nfs_wdata_mempool, GFP_NOIO);
+	struct nfs_pgio_header *p = mempool_alloc(nfs_wdata_mempool, GFP_NOIO);
 
 	if (p)
 		memset(p, 0, sizeof(*p));
 	return p;
 }
 
-static void nfs_writehdr_free(struct nfs_rw_header *whdr)
+static void nfs_writehdr_free(struct nfs_pgio_header *hdr)
 {
-	mempool_free(whdr, nfs_wdata_mempool);
+	mempool_free(hdr, nfs_wdata_mempool);
 }
 
 static void nfs_context_set_write_error(struct nfs_open_context *ctx, int error)
@@ -91,38 +94,76 @@ static void nfs_context_set_write_error(struct nfs_open_context *ctx, int error)
 	set_bit(NFS_CONTEXT_ERROR_WRITE, &ctx->flags);
 }
 
+/*
+ * nfs_page_search_commits_for_head_request_locked
+ *
+ * Search through commit lists on @inode for the head request for @page.
+ * Must be called while holding the inode (which is cinfo) lock.
+ *
+ * Returns the head request if found, or NULL if not found.
+ */
+static struct nfs_page *
+nfs_page_search_commits_for_head_request_locked(struct nfs_inode *nfsi,
+						struct page *page)
+{
+	struct nfs_page *freq, *t;
+	struct nfs_commit_info cinfo;
+	struct inode *inode = &nfsi->vfs_inode;
+
+	nfs_init_cinfo_from_inode(&cinfo, inode);
+
+	/* search through pnfs commit lists */
+	freq = pnfs_search_commit_reqs(inode, &cinfo, page);
+	if (freq)
+		return freq->wb_head;
+
+	/* Linearly search the commit list for the correct request */
+	list_for_each_entry_safe(freq, t, &cinfo.mds->list, wb_list) {
+		if (freq->wb_page == page)
+			return freq->wb_head;
+	}
+
+	return NULL;
+}
+
+/*
+ * nfs_page_find_head_request_locked - find head request associated with @page
+ *
+ * must be called while holding the inode lock.
+ *
+ * returns matching head request with reference held, or NULL if not found.
+ */
 static struct nfs_page *
-nfs_page_find_request_locked(struct nfs_inode *nfsi, struct page *page)
+nfs_page_find_head_request_locked(struct nfs_inode *nfsi, struct page *page)
 {
 	struct nfs_page *req = NULL;
 
 	if (PagePrivate(page))
 		req = (struct nfs_page *)page_private(page);
-	else if (unlikely(PageSwapCache(page))) {
-		struct nfs_page *freq, *t;
-
-		/* Linearly search the commit list for the correct req */
-		list_for_each_entry_safe(freq, t, &nfsi->commit_info.list, wb_list) {
-			if (freq->wb_page == page) {
-				req = freq;
-				break;
-			}
-		}
-	}
+	else if (unlikely(PageSwapCache(page)))
+		req = nfs_page_search_commits_for_head_request_locked(nfsi,
+			page);
 
-	if (req)
+	if (req) {
+		WARN_ON_ONCE(req->wb_head != req);
 		kref_get(&req->wb_kref);
+	}
 
 	return req;
 }
 
-static struct nfs_page *nfs_page_find_request(struct page *page)
+/*
+ * nfs_page_find_head_request - find head request associated with @page
+ *
+ * returns matching head request with reference held, or NULL if not found.
+ */
+static struct nfs_page *nfs_page_find_head_request(struct page *page)
 {
 	struct inode *inode = page_file_mapping(page)->host;
 	struct nfs_page *req = NULL;
 
 	spin_lock(&inode->i_lock);
-	req = nfs_page_find_request_locked(NFS_I(inode), page);
+	req = nfs_page_find_head_request_locked(NFS_I(inode), page);
 	spin_unlock(&inode->i_lock);
 	return req;
 }
@@ -200,7 +241,7 @@ static bool nfs_page_group_covers_page(struct nfs_page *req)
 	unsigned int pos = 0;
 	unsigned int len = nfs_page_length(req->wb_page);
 
-	nfs_page_group_lock(req);
+	nfs_page_group_lock(req, true);
 
 	do {
 		tmp = nfs_page_group_search_locked(req->wb_head, pos);
@@ -274,36 +315,242 @@ static void nfs_end_page_writeback(struct nfs_page *req)
 		clear_bdi_congested(&nfss->backing_dev_info, BLK_RW_ASYNC);
 }
 
-static struct nfs_page *nfs_find_and_lock_request(struct page *page, bool nonblock)
+
+/* nfs_page_group_clear_bits
+ *   @req - an nfs request
+ * clears all page group related bits from @req
+ */
+static void
+nfs_page_group_clear_bits(struct nfs_page *req)
+{
+	clear_bit(PG_TEARDOWN, &req->wb_flags);
+	clear_bit(PG_UNLOCKPAGE, &req->wb_flags);
+	clear_bit(PG_UPTODATE, &req->wb_flags);
+	clear_bit(PG_WB_END, &req->wb_flags);
+	clear_bit(PG_REMOVE, &req->wb_flags);
+}
+
+
+/*
+ * nfs_unroll_locks_and_wait -  unlock all newly locked reqs and wait on @req
+ *
+ * this is a helper function for nfs_lock_and_join_requests
+ *
+ * @inode - inode associated with request page group, must be holding inode lock
+ * @head  - head request of page group, must be holding head lock
+ * @req   - request that couldn't lock and needs to wait on the req bit lock
+ * @nonblock - if true, don't actually wait
+ *
+ * NOTE: this must be called holding page_group bit lock and inode spin lock
+ *       and BOTH will be released before returning.
+ *
+ * returns 0 on success, < 0 on error.
+ */
+static int
+nfs_unroll_locks_and_wait(struct inode *inode, struct nfs_page *head,
+			  struct nfs_page *req, bool nonblock)
+	__releases(&inode->i_lock)
+{
+	struct nfs_page *tmp;
+	int ret;
+
+	/* relinquish all the locks successfully grabbed this run */
+	for (tmp = head ; tmp != req; tmp = tmp->wb_this_page)
+		nfs_unlock_request(tmp);
+
+	WARN_ON_ONCE(test_bit(PG_TEARDOWN, &req->wb_flags));
+
+	/* grab a ref on the request that will be waited on */
+	kref_get(&req->wb_kref);
+
+	nfs_page_group_unlock(head);
+	spin_unlock(&inode->i_lock);
+
+	/* release ref from nfs_page_find_head_request_locked */
+	nfs_release_request(head);
+
+	if (!nonblock)
+		ret = nfs_wait_on_request(req);
+	else
+		ret = -EAGAIN;
+	nfs_release_request(req);
+
+	return ret;
+}
+
+/*
+ * nfs_destroy_unlinked_subrequests - destroy recently unlinked subrequests
+ *
+ * @destroy_list - request list (using wb_this_page) terminated by @old_head
+ * @old_head - the old head of the list
+ *
+ * All subrequests must be locked and removed from all lists, so at this point
+ * they are only "active" in this function, and possibly in nfs_wait_on_request
+ * with a reference held by some other context.
+ */
+static void
+nfs_destroy_unlinked_subrequests(struct nfs_page *destroy_list,
+				 struct nfs_page *old_head)
+{
+	while (destroy_list) {
+		struct nfs_page *subreq = destroy_list;
+
+		destroy_list = (subreq->wb_this_page == old_head) ?
+				   NULL : subreq->wb_this_page;
+
+		WARN_ON_ONCE(old_head != subreq->wb_head);
+
+		/* make sure old group is not used */
+		subreq->wb_head = subreq;
+		subreq->wb_this_page = subreq;
+
+		/* subreq is now totally disconnected from page group or any
+		 * write / commit lists. last chance to wake any waiters */
+		nfs_unlock_request(subreq);
+
+		if (!test_bit(PG_TEARDOWN, &subreq->wb_flags)) {
+			/* release ref on old head request */
+			nfs_release_request(old_head);
+
+			nfs_page_group_clear_bits(subreq);
+
+			/* release the PG_INODE_REF reference */
+			if (test_and_clear_bit(PG_INODE_REF, &subreq->wb_flags))
+				nfs_release_request(subreq);
+			else
+				WARN_ON_ONCE(1);
+		} else {
+			WARN_ON_ONCE(test_bit(PG_CLEAN, &subreq->wb_flags));
+			/* zombie requests have already released the last
+			 * reference and were waiting on the rest of the
+			 * group to complete. Since it's no longer part of a
+			 * group, simply free the request */
+			nfs_page_group_clear_bits(subreq);
+			nfs_free_request(subreq);
+		}
+	}
+}
+
+/*
+ * nfs_lock_and_join_requests - join all subreqs to the head req and return
+ *                              a locked reference, cancelling any pending
+ *                              operations for this page.
+ *
+ * @page - the page used to lookup the "page group" of nfs_page structures
+ * @nonblock - if true, don't block waiting for request locks
+ *
+ * This function joins all sub requests to the head request by first
+ * locking all requests in the group, cancelling any pending operations
+ * and finally updating the head request to cover the whole range covered by
+ * the (former) group.  All subrequests are removed from any write or commit
+ * lists, unlinked from the group and destroyed.
+ *
+ * Returns a locked, referenced pointer to the head request - which after
+ * this call is guaranteed to be the only request associated with the page.
+ * Returns NULL if no requests are found for @page, or a ERR_PTR if an
+ * error was encountered.
+ */
+static struct nfs_page *
+nfs_lock_and_join_requests(struct page *page, bool nonblock)
 {
 	struct inode *inode = page_file_mapping(page)->host;
-	struct nfs_page *req;
+	struct nfs_page *head, *subreq;
+	struct nfs_page *destroy_list = NULL;
+	unsigned int total_bytes;
 	int ret;
 
+try_again:
+	total_bytes = 0;
+
+	WARN_ON_ONCE(destroy_list);
+
 	spin_lock(&inode->i_lock);
-	for (;;) {
-		req = nfs_page_find_request_locked(NFS_I(inode), page);
-		if (req == NULL)
-			break;
-		if (nfs_lock_request(req))
-			break;
-		/* Note: If we hold the page lock, as is the case in nfs_writepage,
-		 *	 then the call to nfs_lock_request() will always
-		 *	 succeed provided that someone hasn't already marked the
-		 *	 request as dirty (in which case we don't care).
-		 */
+
+	/*
+	 * A reference is taken only on the head request which acts as a
+	 * reference to the whole page group - the group will not be destroyed
+	 * until the head reference is released.
+	 */
+	head = nfs_page_find_head_request_locked(NFS_I(inode), page);
+
+	if (!head) {
 		spin_unlock(&inode->i_lock);
-		if (!nonblock)
-			ret = nfs_wait_on_request(req);
-		else
-			ret = -EAGAIN;
-		nfs_release_request(req);
-		if (ret != 0)
+		return NULL;
+	}
+
+	/* lock each request in the page group */
+	ret = nfs_page_group_lock(head, false);
+	if (ret < 0)
+		return ERR_PTR(ret);
+	subreq = head;
+	do {
+		/*
+		 * Subrequests are always contiguous, non overlapping
+		 * and in order. If not, it's a programming error.
+		 */
+		WARN_ON_ONCE(subreq->wb_offset !=
+		     (head->wb_offset + total_bytes));
+
+		/* keep track of how many bytes this group covers */
+		total_bytes += subreq->wb_bytes;
+
+		if (!nfs_lock_request(subreq)) {
+			/* releases page group bit lock and
+			 * inode spin lock and all references */
+			ret = nfs_unroll_locks_and_wait(inode, head,
+				subreq, nonblock);
+
+			if (ret == 0)
+				goto try_again;
+
 			return ERR_PTR(ret);
-		spin_lock(&inode->i_lock);
+		}
+
+		subreq = subreq->wb_this_page;
+	} while (subreq != head);
+
+	/* Now that all requests are locked, make sure they aren't on any list.
+	 * Commit list removal accounting is done after locks are dropped */
+	subreq = head;
+	do {
+		nfs_clear_request_commit(subreq);
+		subreq = subreq->wb_this_page;
+	} while (subreq != head);
+
+	/* unlink subrequests from head, destroy them later */
+	if (head->wb_this_page != head) {
+		/* destroy list will be terminated by head */
+		destroy_list = head->wb_this_page;
+		head->wb_this_page = head;
+
+		/* change head request to cover whole range that
+		 * the former page group covered */
+		head->wb_bytes = total_bytes;
 	}
+
+	/*
+	 * prepare head request to be added to new pgio descriptor
+	 */
+	nfs_page_group_clear_bits(head);
+
+	/*
+	 * some part of the group was still on the inode list - otherwise
+	 * the group wouldn't be involved in async write.
+	 * grab a reference for the head request, iff it needs one.
+	 */
+	if (!test_and_set_bit(PG_INODE_REF, &head->wb_flags))
+		kref_get(&head->wb_kref);
+
+	nfs_page_group_unlock(head);
+
+	/* drop lock to clean uprequests on destroy list */
 	spin_unlock(&inode->i_lock);
-	return req;
+
+	nfs_destroy_unlinked_subrequests(destroy_list, head);
+
+	/* still holds ref on head from nfs_page_find_head_request_locked
+	 * and still has lock on head from lock loop */
+	return head;
 }
 
 /*
@@ -316,7 +563,7 @@ static int nfs_page_async_flush(struct nfs_pageio_descriptor *pgio,
 	struct nfs_page *req;
 	int ret = 0;
 
-	req = nfs_find_and_lock_request(page, nonblock);
+	req = nfs_lock_and_join_requests(page, nonblock);
 	if (!req)
 		goto out;
 	ret = PTR_ERR(req);
@@ -397,7 +644,7 @@ int nfs_writepages(struct address_space *mapping, struct writeback_control *wbc)
 	int err;
 
 	/* Stop dirtying of new pages while we sync */
-	err = wait_on_bit_lock(bitlock, NFS_INO_FLUSHING,
+	err = wait_on_bit_lock_action(bitlock, NFS_INO_FLUSHING,
 			nfs_wait_bit_killable, TASK_KILLABLE);
 	if (err)
 		goto out_err;
@@ -448,7 +695,9 @@ static void nfs_inode_add_request(struct inode *inode, struct nfs_page *req)
 		set_page_private(req->wb_page, (unsigned long)req);
 	}
 	nfsi->npages++;
-	set_bit(PG_INODE_REF, &req->wb_flags);
+	/* this a head request for a page group - mark it as having an
+	 * extra reference so sub groups can follow suit */
+	WARN_ON(test_and_set_bit(PG_INODE_REF, &req->wb_flags));
 	kref_get(&req->wb_kref);
 	spin_unlock(&inode->i_lock);
 }
@@ -474,7 +723,11 @@ static void nfs_inode_remove_request(struct nfs_page *req)
 		nfsi->npages--;
 		spin_unlock(&inode->i_lock);
 	}
-	nfs_release_request(req);
+
+	if (test_and_clear_bit(PG_INODE_REF, &req->wb_flags))
+		nfs_release_request(req);
+	else
+		WARN_ON_ONCE(1);
 }
 
 static void
@@ -578,6 +831,7 @@ nfs_clear_page_commit(struct page *page)
 	dec_bdi_stat(page_file_mapping(page)->backing_dev_info, BDI_RECLAIMABLE);
 }
 
+/* Called holding inode (/cinfo) lock */
 static void
 nfs_clear_request_commit(struct nfs_page *req)
 {
@@ -587,20 +841,17 @@ nfs_clear_request_commit(struct nfs_page *req)
 
 		nfs_init_cinfo_from_inode(&cinfo, inode);
 		if (!pnfs_clear_request_commit(req, &cinfo)) {
-			spin_lock(cinfo.lock);
 			nfs_request_remove_commit_list(req, &cinfo);
-			spin_unlock(cinfo.lock);
 		}
 		nfs_clear_page_commit(req->wb_page);
 	}
 }
 
-static inline
-int nfs_write_need_commit(struct nfs_pgio_data *data)
+int nfs_write_need_commit(struct nfs_pgio_header *hdr)
 {
-	if (data->verf.committed == NFS_DATA_SYNC)
-		return data->header->lseg == NULL;
-	return data->verf.committed != NFS_FILE_SYNC;
+	if (hdr->verf.committed == NFS_DATA_SYNC)
+		return hdr->lseg == NULL;
+	return hdr->verf.committed != NFS_FILE_SYNC;
 }
 
 #else
@@ -626,8 +877,7 @@ nfs_clear_request_commit(struct nfs_page *req)
 {
 }
 
-static inline
-int nfs_write_need_commit(struct nfs_pgio_data *data)
+int nfs_write_need_commit(struct nfs_pgio_header *hdr)
 {
 	return 0;
 }
@@ -638,7 +888,6 @@ static void nfs_write_completion(struct nfs_pgio_header *hdr)
 {
 	struct nfs_commit_info cinfo;
 	unsigned long bytes = 0;
-	bool do_destroy;
 
 	if (test_bit(NFS_IOHDR_REDO, &hdr->flags))
 		goto out;
@@ -654,11 +903,7 @@ static void nfs_write_completion(struct nfs_pgio_header *hdr)
 			nfs_context_set_write_error(req->wb_context, hdr->error);
 			goto remove_req;
 		}
-		if (test_bit(NFS_IOHDR_NEED_RESCHED, &hdr->flags)) {
-			nfs_mark_request_dirty(req);
-			goto next;
-		}
-		if (test_bit(NFS_IOHDR_NEED_COMMIT, &hdr->flags)) {
+		if (nfs_write_need_commit(hdr)) {
 			memcpy(&req->wb_verf, &hdr->verf.verifier, sizeof(req->wb_verf));
 			nfs_mark_request_commit(req, hdr->lseg, &cinfo);
 			goto next;
@@ -668,7 +913,6 @@ remove_req:
 next:
 		nfs_unlock_request(req);
 		nfs_end_page_writeback(req);
-		do_destroy = !test_bit(NFS_IOHDR_NEED_COMMIT, &hdr->flags);
 		nfs_release_request(req);
 	}
 out:
@@ -769,7 +1013,7 @@ static struct nfs_page *nfs_try_to_update_request(struct inode *inode,
 	spin_lock(&inode->i_lock);
 
 	for (;;) {
-		req = nfs_page_find_request_locked(NFS_I(inode), page);
+		req = nfs_page_find_head_request_locked(NFS_I(inode), page);
 		if (req == NULL)
 			goto out_unlock;
 
@@ -810,9 +1054,9 @@ static struct nfs_page *nfs_try_to_update_request(struct inode *inode,
 	else
 		req->wb_bytes = rqend - req->wb_offset;
 out_unlock:
-	spin_unlock(&inode->i_lock);
 	if (req)
 		nfs_clear_request_commit(req);
+	spin_unlock(&inode->i_lock);
 	return req;
 out_flushme:
 	spin_unlock(&inode->i_lock);
@@ -877,7 +1121,7 @@ int nfs_flush_incompatible(struct file *file, struct page *page)
 	 * dropped page.
 	 */
 	do {
-		req = nfs_page_find_request(page);
+		req = nfs_page_find_head_request(page);
 		if (req == NULL)
 			return 0;
 		l_ctx = req->wb_lock_context;
@@ -934,12 +1178,14 @@ static bool nfs_write_pageuptodate(struct page *page, struct inode *inode)
 
 	if (nfs_have_delegated_attributes(inode))
 		goto out;
-	if (nfsi->cache_validity & (NFS_INO_INVALID_DATA|NFS_INO_REVAL_PAGECACHE))
+	if (nfsi->cache_validity & NFS_INO_REVAL_PAGECACHE)
 		return false;
 	smp_rmb();
 	if (test_bit(NFS_INO_INVALIDATING, &nfsi->flags))
 		return false;
 out:
+	if (nfsi->cache_validity & NFS_INO_INVALID_DATA)
+		return false;
 	return PageUptodate(page) != 0;
 }
 
@@ -1011,17 +1257,18 @@ static int flush_task_priority(int how)
 	return RPC_PRIORITY_NORMAL;
 }
 
-static void nfs_initiate_write(struct nfs_pgio_data *data, struct rpc_message *msg,
+static void nfs_initiate_write(struct nfs_pgio_header *hdr,
+			       struct rpc_message *msg,
 			       struct rpc_task_setup *task_setup_data, int how)
 {
-	struct inode *inode = data->header->inode;
+	struct inode *inode = hdr->inode;
 	int priority = flush_task_priority(how);
 
 	task_setup_data->priority = priority;
-	NFS_PROTO(inode)->write_setup(data, msg);
+	NFS_PROTO(inode)->write_setup(hdr, msg);
 
 	nfs4_state_protect_write(NFS_SERVER(inode)->nfs_client,
-				 &task_setup_data->rpc_client, msg, data);
+				 &task_setup_data->rpc_client, msg, hdr);
 }
 
 /* If a nfs_flush_* function fails, it should remove reqs from @head and
@@ -1083,21 +1330,9 @@ void nfs_commit_prepare(struct rpc_task *task, void *calldata)
 	NFS_PROTO(data->inode)->commit_rpc_prepare(task, data);
 }
 
-static void nfs_writeback_release_common(struct nfs_pgio_data *data)
+static void nfs_writeback_release_common(struct nfs_pgio_header *hdr)
 {
-	struct nfs_pgio_header *hdr = data->header;
-	int status = data->task.tk_status;
-
-	if ((status >= 0) && nfs_write_need_commit(data)) {
-		spin_lock(&hdr->lock);
-		if (test_bit(NFS_IOHDR_NEED_RESCHED, &hdr->flags))
-			; /* Do nothing */
-		else if (!test_and_set_bit(NFS_IOHDR_NEED_COMMIT, &hdr->flags))
-			memcpy(&hdr->verf, &data->verf, sizeof(hdr->verf));
-		else if (memcmp(&hdr->verf, &data->verf, sizeof(hdr->verf)))
-			set_bit(NFS_IOHDR_NEED_RESCHED, &hdr->flags);
-		spin_unlock(&hdr->lock);
-	}
+	/* do nothing! */
 }
 
 /*
@@ -1128,7 +1363,8 @@ static int nfs_should_remove_suid(const struct inode *inode)
 /*
  * This function is called when the WRITE call is complete.
  */
-static int nfs_writeback_done(struct rpc_task *task, struct nfs_pgio_data *data,
+static int nfs_writeback_done(struct rpc_task *task,
+			      struct nfs_pgio_header *hdr,
 			      struct inode *inode)
 {
 	int status;
@@ -1140,13 +1376,14 @@ static int nfs_writeback_done(struct rpc_task *task, struct nfs_pgio_data *data,
 	 * another writer had changed the file, but some applications
 	 * depend on tighter cache coherency when writing.
 	 */
-	status = NFS_PROTO(inode)->write_done(task, data);
+	status = NFS_PROTO(inode)->write_done(task, hdr);
 	if (status != 0)
 		return status;
-	nfs_add_stats(inode, NFSIOS_SERVERWRITTENBYTES, data->res.count);
+	nfs_add_stats(inode, NFSIOS_SERVERWRITTENBYTES, hdr->res.count);
 
 #if IS_ENABLED(CONFIG_NFS_V3) || IS_ENABLED(CONFIG_NFS_V4)
-	if (data->res.verf->committed < data->args.stable && task->tk_status >= 0) {
+	if (hdr->res.verf->committed < hdr->args.stable &&
+	    task->tk_status >= 0) {
 		/* We tried a write call, but the server did not
 		 * commit data to stable storage even though we
 		 * requested it.
@@ -1162,7 +1399,7 @@ static int nfs_writeback_done(struct rpc_task *task, struct nfs_pgio_data *data,
 			dprintk("NFS:       faulty NFS server %s:"
 				" (committed = %d) != (stable = %d)\n",
 				NFS_SERVER(inode)->nfs_client->cl_hostname,
-				data->res.verf->committed, data->args.stable);
+				hdr->res.verf->committed, hdr->args.stable);
 			complain = jiffies + 300 * HZ;
 		}
 	}
@@ -1177,16 +1414,17 @@ static int nfs_writeback_done(struct rpc_task *task, struct nfs_pgio_data *data,
 /*
  * This function is called when the WRITE call is complete.
  */
-static void nfs_writeback_result(struct rpc_task *task, struct nfs_pgio_data *data)
+static void nfs_writeback_result(struct rpc_task *task,
+				 struct nfs_pgio_header *hdr)
 {
-	struct nfs_pgio_args	*argp = &data->args;
-	struct nfs_pgio_res	*resp = &data->res;
+	struct nfs_pgio_args	*argp = &hdr->args;
+	struct nfs_pgio_res	*resp = &hdr->res;
 
 	if (resp->count < argp->count) {
 		static unsigned long    complain;
 
 		/* This a short write! */
-		nfs_inc_stats(data->header->inode, NFSIOS_SHORTWRITE);
+		nfs_inc_stats(hdr->inode, NFSIOS_SHORTWRITE);
 
 		/* Has the server at least made some progress? */
 		if (resp->count == 0) {
@@ -1196,14 +1434,14 @@ static void nfs_writeback_result(struct rpc_task *task, struct nfs_pgio_data *da
 				       argp->count);
 				complain = jiffies + 300 * HZ;
 			}
-			nfs_set_pgio_error(data->header, -EIO, argp->offset);
+			nfs_set_pgio_error(hdr, -EIO, argp->offset);
 			task->tk_status = -EIO;
 			return;
 		}
 		/* Was this an NFSv2 write or an NFSv3 stable write? */
 		if (resp->verf->committed != NFS_UNSTABLE) {
 			/* Resend from where the server left off */
-			data->mds_offset += resp->count;
+			hdr->mds_offset += resp->count;
 			argp->offset += resp->count;
 			argp->pgbase += resp->count;
 			argp->count -= resp->count;
@@ -1473,7 +1711,7 @@ int nfs_commit_inode(struct inode *inode, int how)
 			return error;
 		if (!may_wait)
 			goto out_mark_dirty;
-		error = wait_on_bit(&NFS_I(inode)->flags,
+		error = wait_on_bit_action(&NFS_I(inode)->flags,
 				NFS_INO_COMMIT,
 				nfs_wait_bit_killable,
 				TASK_KILLABLE);
@@ -1567,27 +1805,28 @@ int nfs_wb_page_cancel(struct inode *inode, struct page *page)
 	struct nfs_page *req;
 	int ret = 0;
 
-	for (;;) {
-		wait_on_page_writeback(page);
-		req = nfs_page_find_request(page);
-		if (req == NULL)
-			break;
-		if (nfs_lock_request(req)) {
-			nfs_clear_request_commit(req);
-			nfs_inode_remove_request(req);
-			/*
-			 * In case nfs_inode_remove_request has marked the
-			 * page as being dirty
-			 */
-			cancel_dirty_page(page, PAGE_CACHE_SIZE);
-			nfs_unlock_and_release_request(req);
-			break;
-		}
-		ret = nfs_wait_on_request(req);
-		nfs_release_request(req);
-		if (ret < 0)
-			break;
+	wait_on_page_writeback(page);
+
+	/* blocking call to cancel all requests and join to a single (head)
+	 * request */
+	req = nfs_lock_and_join_requests(page, false);
+
+	if (IS_ERR(req)) {
+		ret = PTR_ERR(req);
+	} else if (req) {
+		/* all requests from this page have been cancelled by
+		 * nfs_lock_and_join_requests, so just remove the head
+		 * request from the inode / page_private pointer and
+		 * release it */
+		nfs_inode_remove_request(req);
+		/*
+		 * In case nfs_inode_remove_request has marked the
+		 * page as being dirty
+		 */
+		cancel_dirty_page(page, PAGE_CACHE_SIZE);
+		nfs_unlock_and_release_request(req);
 	}
+
 	return ret;
 }
 
@@ -1653,7 +1892,7 @@ int nfs_migrate_page(struct address_space *mapping, struct page *newpage,
 int __init nfs_init_writepagecache(void)
 {
 	nfs_wdata_cachep = kmem_cache_create("nfs_write_data",
-					     sizeof(struct nfs_rw_header),
+					     sizeof(struct nfs_pgio_header),
 					     0, SLAB_HWCACHE_ALIGN,
 					     NULL);
 	if (nfs_wdata_cachep == NULL)