X-Git-Url: http://git.onelab.eu/?a=blobdiff_plain;f=fs%2Fnfs%2Finode.c;h=2d8cf645f8bbca096215a4b2c1d3b09eebfd72d5;hb=16c70f8c1b54b61c3b951b6fb220df250fe09b32;hp=60b30747801240cc624dd555e717971908577911;hpb=c7b5ebbddf7bcd3651947760f423e3783bbe6573;p=linux-2.6.git

diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c
index 60b307478..2d8cf645f 100644
--- a/fs/nfs/inode.c
+++ b/fs/nfs/inode.c
@@ -13,7 +13,6 @@
  *
  */
 
-#include <linux/config.h>
 #include <linux/module.h>
 #include <linux/init.h>
 
@@ -26,6 +25,7 @@
 #include <linux/unistd.h>
 #include <linux/sunrpc/clnt.h>
 #include <linux/sunrpc/stats.h>
+#include <linux/sunrpc/metrics.h>
 #include <linux/nfs_fs.h>
 #include <linux/nfs_mount.h>
 #include <linux/nfs4_mount.h>
@@ -35,558 +35,126 @@
 #include <linux/mount.h>
 #include <linux/nfs_idmap.h>
 #include <linux/vfs.h>
+#include <linux/inet.h>
+#include <linux/nfs_xdr.h>
 #include <linux/vserver/xid.h>
 
 #include <asm/system.h>
 #include <asm/uaccess.h>
 
+#include "nfs4_fs.h"
+#include "callback.h"
 #include "delegation.h"
+#include "iostat.h"
+#include "internal.h"
 
 #define NFSDBG_FACILITY		NFSDBG_VFS
 #define NFS_PARANOIA 1
 
-/* Maximum number of readahead requests
- * FIXME: this should really be a sysctl so that users may tune it to suit
- *        their needs. People that do NFS over a slow network, might for
- *        instance want to reduce it to something closer to 1 for improved
- *        interactive response.
- */
-#define NFS_MAX_READAHEAD	(RPC_DEF_SLOT_TABLE - 1)
-
 static void nfs_invalidate_inode(struct inode *);
-static int nfs_update_inode(struct inode *, struct nfs_fattr *, unsigned long);
-
-static struct inode *nfs_alloc_inode(struct super_block *sb);
-static void nfs_destroy_inode(struct inode *);
-static int nfs_write_inode(struct inode *,int);
-static void nfs_delete_inode(struct inode *);
-static void nfs_clear_inode(struct inode *);
-static void nfs_umount_begin(struct super_block *);
-static int  nfs_statfs(struct super_block *, struct kstatfs *);
-static int  nfs_show_options(struct seq_file *, struct vfsmount *);
-
-static struct super_operations nfs_sops = { 
-	.alloc_inode	= nfs_alloc_inode,
-	.destroy_inode	= nfs_destroy_inode,
-	.write_inode	= nfs_write_inode,
-	.delete_inode	= nfs_delete_inode,
-	.statfs		= nfs_statfs,
-	.clear_inode	= nfs_clear_inode,
-	.umount_begin	= nfs_umount_begin,
-	.show_options	= nfs_show_options,
-};
+static int nfs_update_inode(struct inode *, struct nfs_fattr *);
 
-/*
- * RPC cruft for NFS
- */
-struct rpc_stat			nfs_rpcstat = {
-	.program		= &nfs_program
-};
-static struct rpc_version *	nfs_version[] = {
-	NULL,
-	NULL,
-	&nfs_version2,
-#if defined(CONFIG_NFS_V3)
-	&nfs_version3,
-#elif defined(CONFIG_NFS_V4)
-	NULL,
-#endif
-#if defined(CONFIG_NFS_V4)
-	&nfs_version4,
-#endif
-};
-
-struct rpc_program		nfs_program = {
-	.name			= "nfs",
-	.number			= NFS_PROGRAM,
-	.nrvers			= sizeof(nfs_version) / sizeof(nfs_version[0]),
-	.version		= nfs_version,
-	.stats			= &nfs_rpcstat,
-	.pipe_dir_name		= "/nfs",
-};
+static void nfs_zap_acl_cache(struct inode *);
 
-static inline unsigned long
-nfs_fattr_to_ino_t(struct nfs_fattr *fattr)
-{
-	return nfs_fileid_to_ino_t(fattr->fileid);
-}
+static kmem_cache_t * nfs_inode_cachep;
 
-static int
-nfs_write_inode(struct inode *inode, int sync)
+int nfs_write_inode(struct inode *inode, int sync)
 {
-	int flags = sync ? FLUSH_WAIT : 0;
+	int flags = sync ? FLUSH_SYNC : 0;
 	int ret;
 
-	ret = nfs_commit_inode(inode, 0, 0, flags);
+	ret = nfs_commit_inode(inode, flags);
 	if (ret < 0)
 		return ret;
 	return 0;
 }
 
-static void
-nfs_delete_inode(struct inode * inode)
+void nfs_clear_inode(struct inode *inode)
 {
-	dprintk("NFS: delete_inode(%s/%ld)\n", inode->i_sb->s_id, inode->i_ino);
+	struct nfs_inode *nfsi = NFS_I(inode);
+	struct rpc_cred *cred;
 
-	nfs_wb_all(inode);
 	/*
 	 * The following should never happen...
 	 */
-	if (nfs_have_writebacks(inode)) {
-		printk(KERN_ERR "nfs_delete_inode: inode %ld has pending RPC requests\n", inode->i_ino);
-	}
-
-	clear_inode(inode);
-}
-
-/*
- * For the moment, the only task for the NFS clear_inode method is to
- * release the mmap credential
- */
-static void
-nfs_clear_inode(struct inode *inode)
-{
-	struct nfs_inode *nfsi = NFS_I(inode);
-	struct rpc_cred *cred;
-
-	nfs_wb_all(inode);
+	BUG_ON(nfs_have_writebacks(inode));
 	BUG_ON (!list_empty(&nfsi->open_files));
+	nfs_zap_acl_cache(inode);
 	cred = nfsi->cache_access.cred;
 	if (cred)
 		put_rpccred(cred);
-	BUG_ON(atomic_read(&nfsi->data_updates) != 0);
-}
-
-void
-nfs_umount_begin(struct super_block *sb)
-{
-	struct nfs_server *server = NFS_SB(sb);
-	struct rpc_clnt	*rpc;
-
-	/* -EIO all pending I/O */
-	if ((rpc = server->client) != NULL)
-		rpc_killall_tasks(rpc);
-}
-
-
-static inline unsigned long
-nfs_block_bits(unsigned long bsize, unsigned char *nrbitsp)
-{
-	/* make sure blocksize is a power of two */
-	if ((bsize & (bsize - 1)) || nrbitsp) {
-		unsigned char	nrbits;
-
-		for (nrbits = 31; nrbits && !(bsize & (1 << nrbits)); nrbits--)
-			;
-		bsize = 1 << nrbits;
-		if (nrbitsp)
-			*nrbitsp = nrbits;
-	}
-
-	return bsize;
-}
-
-/*
- * Calculate the number of 512byte blocks used.
- */
-static inline unsigned long
-nfs_calc_block_size(u64 tsize)
-{
-	loff_t used = (tsize + 511) >> 9;
-	return (used > ULONG_MAX) ? ULONG_MAX : used;
-}
-
-/*
- * Compute and set NFS server blocksize
- */
-static inline unsigned long
-nfs_block_size(unsigned long bsize, unsigned char *nrbitsp)
-{
-	if (bsize < 1024)
-		bsize = NFS_DEF_FILE_IO_BUFFER_SIZE;
-	else if (bsize >= NFS_MAX_FILE_IO_BUFFER_SIZE)
-		bsize = NFS_MAX_FILE_IO_BUFFER_SIZE;
-
-	return nfs_block_bits(bsize, nrbitsp);
-}
 
-/*
- * Obtain the root inode of the file system.
- */
-static struct inode *
-nfs_get_root(struct super_block *sb, struct nfs_fh *rootfh, struct nfs_fsinfo *fsinfo)
-{
-	struct nfs_server	*server = NFS_SB(sb);
-	struct inode *rooti;
-	int			error;
-
-	error = server->rpc_ops->getroot(server, rootfh, fsinfo);
-	if (error < 0) {
-		dprintk("nfs_get_root: getattr error = %d\n", -error);
-		return ERR_PTR(error);
-	}
-
-	rooti = nfs_fhget(sb, rootfh, fsinfo->fattr);
-	if (!rooti)
-		return ERR_PTR(-ENOMEM);
-	return rooti;
-}
-
-/*
- * Do NFS version-independent mount processing, and sanity checking
- */
-static int
-nfs_sb_init(struct super_block *sb, rpc_authflavor_t authflavor)
-{
-	struct nfs_server	*server;
-	struct inode		*root_inode;
-	struct nfs_fattr	fattr;
-	struct nfs_fsinfo	fsinfo = {
-					.fattr = &fattr,
-				};
-	struct nfs_pathconf pathinfo = {
-			.fattr = &fattr,
-	};
-	int no_root_error = 0;
-
-	/* We probably want something more informative here */
-	snprintf(sb->s_id, sizeof(sb->s_id), "%x:%x", MAJOR(sb->s_dev), MINOR(sb->s_dev));
-
-	server = NFS_SB(sb);
-
-	sb->s_magic      = NFS_SUPER_MAGIC;
-
-	root_inode = nfs_get_root(sb, &server->fh, &fsinfo);
-	/* Did getting the root inode fail? */
-	if (IS_ERR(root_inode)) {
-		no_root_error = PTR_ERR(root_inode);
-		goto out_no_root;
-	}
-	sb->s_root = d_alloc_root(root_inode);
-	if (!sb->s_root) {
-		no_root_error = -ENOMEM;
-		goto out_no_root;
-	}
-	sb->s_root->d_op = server->rpc_ops->dentry_ops;
-
-	/* Get some general file system info */
-	if (server->namelen == 0 &&
-	    server->rpc_ops->pathconf(server, &server->fh, &pathinfo) >= 0)
-		server->namelen = pathinfo.max_namelen;
-	/* Work out a lot of parameters */
-	if (server->rsize == 0)
-		server->rsize = nfs_block_size(fsinfo.rtpref, NULL);
-	if (server->wsize == 0)
-		server->wsize = nfs_block_size(fsinfo.wtpref, NULL);
-
-	if (fsinfo.rtmax >= 512 && server->rsize > fsinfo.rtmax)
-		server->rsize = nfs_block_size(fsinfo.rtmax, NULL);
-	if (fsinfo.wtmax >= 512 && server->wsize > fsinfo.wtmax)
-		server->wsize = nfs_block_size(fsinfo.wtmax, NULL);
-
-	server->rpages = (server->rsize + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
-	if (server->rpages > NFS_READ_MAXIOV) {
-		server->rpages = NFS_READ_MAXIOV;
-		server->rsize = server->rpages << PAGE_CACHE_SHIFT;
-	}
-
-	server->wpages = (server->wsize + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
-        if (server->wpages > NFS_WRITE_MAXIOV) {
-		server->wpages = NFS_WRITE_MAXIOV;
-                server->wsize = server->wpages << PAGE_CACHE_SHIFT;
-	}
-
-	if (sb->s_blocksize == 0)
-		sb->s_blocksize = nfs_block_bits(server->wsize,
-							 &sb->s_blocksize_bits);
-	server->wtmult = nfs_block_bits(fsinfo.wtmult, NULL);
-
-	server->dtsize = nfs_block_size(fsinfo.dtpref, NULL);
-	if (server->dtsize > PAGE_CACHE_SIZE)
-		server->dtsize = PAGE_CACHE_SIZE;
-	if (server->dtsize > server->rsize)
-		server->dtsize = server->rsize;
-
-	if (server->flags & NFS_MOUNT_NOAC) {
-		server->acregmin = server->acregmax = 0;
-		server->acdirmin = server->acdirmax = 0;
-		sb->s_flags |= MS_SYNCHRONOUS;
-	}
-	server->backing_dev_info.ra_pages = server->rpages * NFS_MAX_READAHEAD;
-
-	if (server->flags & NFS_MOUNT_TAGXID)
-		sb->s_flags |= MS_TAGXID;
-
-	sb->s_maxbytes = fsinfo.maxfilesize;
-	if (sb->s_maxbytes > MAX_LFS_FILESIZE) 
-		sb->s_maxbytes = MAX_LFS_FILESIZE; 
-
-	/* We're airborne Set socket buffersize */
-	rpc_setbufsize(server->client, server->wsize + 100, server->rsize + 100);
-	return 0;
-	/* Yargs. It didn't work out. */
-out_no_root:
-	dprintk("nfs_sb_init: get root inode failed: errno %d\n", -no_root_error);
-	if (!IS_ERR(root_inode))
-		iput(root_inode);
-	return no_root_error;
+	nfs_fscache_release_fh_cookie(NFS_SERVER(inode), nfsi);
+	BUG_ON(atomic_read(&nfsi->data_updates) != 0);
 }
 
-/*
- * Create an RPC client handle.
+/**
+ * nfs_sync_mapping - helper to flush all mmapped dirty data to disk
  */
-static struct rpc_clnt *
-nfs_create_client(struct nfs_server *server, const struct nfs_mount_data *data)
+int nfs_sync_mapping(struct address_space *mapping)
 {
-	struct rpc_timeout	timeparms;
-	struct rpc_xprt		*xprt = NULL;
-	struct rpc_clnt		*clnt = NULL;
-	int			tcp   = (data->flags & NFS_MOUNT_TCP);
-
-	/* Initialize timeout values */
-	timeparms.to_initval = data->timeo * HZ / 10;
-	timeparms.to_retries = data->retrans;
-	timeparms.to_maxval  = tcp ? RPC_MAX_TCP_TIMEOUT : RPC_MAX_UDP_TIMEOUT;
-	timeparms.to_exponential = 1;
-
-	if (!timeparms.to_initval)
-		timeparms.to_initval = (tcp ? 600 : 11) * HZ / 10;
-	if (!timeparms.to_retries)
-		timeparms.to_retries = 5;
-
-	/* create transport and client */
-	xprt = xprt_create_proto(tcp ? IPPROTO_TCP : IPPROTO_UDP,
-				 &server->addr, &timeparms);
-	if (IS_ERR(xprt)) {
-		printk(KERN_WARNING "NFS: cannot create RPC transport.\n");
-		return (struct rpc_clnt *)xprt;
-	}
-	clnt = rpc_create_client(xprt, server->hostname, &nfs_program,
-				 server->rpc_ops->version, data->pseudoflavor);
-	if (IS_ERR(clnt)) {
-		printk(KERN_WARNING "NFS: cannot create RPC client.\n");
-		goto out_fail;
-	}
-
-	clnt->cl_intr     = (server->flags & NFS_MOUNT_INTR) ? 1 : 0;
-	clnt->cl_softrtry = (server->flags & NFS_MOUNT_SOFT) ? 1 : 0;
-	clnt->cl_droppriv = (server->flags & NFS_MOUNT_BROKEN_SUID) ? 1 : 0;
-	clnt->cl_tagxid   = (server->flags & NFS_MOUNT_TAGXID) ? 1 : 0;
-	clnt->cl_chatty   = 1;
-
-	return clnt;
+	int ret;
 
-out_fail:
-	xprt_destroy(xprt);
-	return clnt;
+	if (mapping->nrpages == 0)
+		return 0;
+	unmap_mapping_range(mapping, 0, 0, 0);
+	ret = filemap_write_and_wait(mapping);
+	if (ret != 0)
+		goto out;
+	ret = nfs_wb_all(mapping->host);
+out:
+	return ret;
 }
 
 /*
- * The way this works is that the mount process passes a structure
- * in the data argument which contains the server's IP address
- * and the root file handle obtained from the server's mount
- * daemon. We stash these away in the private superblock fields.
+ * Invalidate the local caches
  */
-static int
-nfs_fill_super(struct super_block *sb, struct nfs_mount_data *data, int silent)
+static void nfs_zap_caches_locked(struct inode *inode)
 {
-	struct nfs_server	*server;
-	rpc_authflavor_t	authflavor;
-
-	server           = NFS_SB(sb);
-	sb->s_blocksize_bits = 0;
-	sb->s_blocksize = 0;
-	if (data->bsize)
-		sb->s_blocksize = nfs_block_size(data->bsize, &sb->s_blocksize_bits);
-	if (data->rsize)
-		server->rsize = nfs_block_size(data->rsize, NULL);
-	if (data->wsize)
-		server->wsize = nfs_block_size(data->wsize, NULL);
-	server->flags    = data->flags & NFS_MOUNT_FLAGMASK;
-
-	server->acregmin = data->acregmin*HZ;
-	server->acregmax = data->acregmax*HZ;
-	server->acdirmin = data->acdirmin*HZ;
-	server->acdirmax = data->acdirmax*HZ;
-
-	/* Start lockd here, before we might error out */
-	if (!(server->flags & NFS_MOUNT_NONLM))
-		lockd_up();
-
-	server->namelen  = data->namlen;
-	server->hostname = kmalloc(strlen(data->hostname) + 1, GFP_KERNEL);
-	if (!server->hostname)
-		return -ENOMEM;
-	strcpy(server->hostname, data->hostname);
-
-	/* Check NFS protocol revision and initialize RPC op vector
-	 * and file handle pool. */
-	if (server->flags & NFS_MOUNT_VER3) {
-#ifdef CONFIG_NFS_V3
-		server->rpc_ops = &nfs_v3_clientops;
-		server->caps |= NFS_CAP_READDIRPLUS;
-		if (data->version < 4) {
-			printk(KERN_NOTICE "NFS: NFSv3 not supported by mount program.\n");
-			return -EIO;
-		}
-#else
-		printk(KERN_NOTICE "NFS: NFSv3 not supported.\n");
-		return -EIO;
-#endif
-	} else {
-		server->rpc_ops = &nfs_v2_clientops;
-	}
+	struct nfs_inode *nfsi = NFS_I(inode);
+	int mode = inode->i_mode;
 
-	/* Fill in pseudoflavor for mount version < 5 */
-	if (!(data->flags & NFS_MOUNT_SECFLAVOUR))
-		data->pseudoflavor = RPC_AUTH_UNIX;
-	authflavor = data->pseudoflavor;	/* save for sb_init() */
-	/* XXX maybe we want to add a server->pseudoflavor field */
-
-	/* Create RPC client handles */
-	server->client = nfs_create_client(server, data);
-	if (IS_ERR(server->client))
-		return PTR_ERR(server->client);
-	/* RFC 2623, sec 2.3.2 */
-	if (authflavor != RPC_AUTH_UNIX) {
-		server->client_sys = rpc_clone_client(server->client);
-		if (IS_ERR(server->client_sys))
-			return PTR_ERR(server->client_sys);
-		if (!rpcauth_create(RPC_AUTH_UNIX, server->client_sys))
-			return -ENOMEM;
-	} else {
-		atomic_inc(&server->client->cl_count);
-		server->client_sys = server->client;
-	}
+	nfs_inc_stats(inode, NFSIOS_ATTRINVALIDATE);
 
-	if (server->flags & NFS_MOUNT_VER3) {
-		if (server->namelen == 0 || server->namelen > NFS3_MAXNAMLEN)
-			server->namelen = NFS3_MAXNAMLEN;
-	} else {
-		if (server->namelen == 0 || server->namelen > NFS2_MAXNAMLEN)
-			server->namelen = NFS2_MAXNAMLEN;
-	}
+	NFS_ATTRTIMEO(inode) = NFS_MINATTRTIMEO(inode);
+	NFS_ATTRTIMEO_UPDATE(inode) = jiffies;
 
-	sb->s_op = &nfs_sops;
-	return nfs_sb_init(sb, authflavor);
+	memset(NFS_COOKIEVERF(inode), 0, sizeof(NFS_COOKIEVERF(inode)));
+	if (S_ISREG(mode) || S_ISDIR(mode) || S_ISLNK(mode))
+		nfsi->cache_validity |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA|NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL|NFS_INO_REVAL_PAGECACHE;
+	else
+		nfsi->cache_validity |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL|NFS_INO_REVAL_PAGECACHE;
 }
 
-static int
-nfs_statfs(struct super_block *sb, struct kstatfs *buf)
+void nfs_zap_caches(struct inode *inode)
 {
-	struct nfs_server *server = NFS_SB(sb);
-	unsigned char blockbits;
-	unsigned long blockres;
-	struct nfs_fh *rootfh = NFS_FH(sb->s_root->d_inode);
-	struct nfs_fattr fattr;
-	struct nfs_fsstat res = {
-			.fattr = &fattr,
-	};
-	int error;
-
-	lock_kernel();
-
-	error = server->rpc_ops->statfs(server, rootfh, &res);
-	buf->f_type = NFS_SUPER_MAGIC;
-	if (error < 0)
-		goto out_err;
-
-	buf->f_frsize = server->wtmult;
-	buf->f_bsize = sb->s_blocksize;
-	blockbits = sb->s_blocksize_bits;
-	blockres = (1 << blockbits) - 1;
-	buf->f_blocks = (res.tbytes + blockres) >> blockbits;
-	buf->f_bfree = (res.fbytes + blockres) >> blockbits;
-	buf->f_bavail = (res.abytes + blockres) >> blockbits;
-	buf->f_files = res.tfiles;
-	buf->f_ffree = res.afiles;
-
-	buf->f_namelen = server->namelen;
- out:
-	unlock_kernel();
-
-	return 0;
-
- out_err:
-	printk(KERN_WARNING "nfs_statfs: statfs error = %d\n", -error);
-	buf->f_bsize = buf->f_blocks = buf->f_bfree = buf->f_bavail = -1;
-	goto out;
+	spin_lock(&inode->i_lock);
+	nfs_zap_caches_locked(inode);
+	spin_unlock(&inode->i_lock);
 
+	nfs_fscache_zap_fh_cookie(NFS_SERVER(inode), NFS_I(inode));
 }
 
-static int nfs_show_options(struct seq_file *m, struct vfsmount *mnt)
+static void nfs_zap_acl_cache(struct inode *inode)
 {
-	static struct proc_nfs_info {
-		int flag;
-		char *str;
-		char *nostr;
-	} nfs_info[] = {
-		{ NFS_MOUNT_SOFT, ",soft", ",hard" },
-		{ NFS_MOUNT_INTR, ",intr", "" },
-		{ NFS_MOUNT_POSIX, ",posix", "" },
-		{ NFS_MOUNT_TCP, ",tcp", ",udp" },
-		{ NFS_MOUNT_NOCTO, ",nocto", "" },
-		{ NFS_MOUNT_NOAC, ",noac", "" },
-		{ NFS_MOUNT_NONLM, ",nolock", ",lock" },
-		{ NFS_MOUNT_BROKEN_SUID, ",broken_suid", "" },
-		{ NFS_MOUNT_TAGXID, ",tagxid", "" },
-		{ 0, NULL, NULL }
-	};
-	struct proc_nfs_info *nfs_infop;
-	struct nfs_server *nfss = NFS_SB(mnt->mnt_sb);
-
-	seq_printf(m, ",v%d", nfss->rpc_ops->version);
-	seq_printf(m, ",rsize=%d", nfss->rsize);
-	seq_printf(m, ",wsize=%d", nfss->wsize);
-	if (nfss->acregmin != 3*HZ)
-		seq_printf(m, ",acregmin=%d", nfss->acregmin/HZ);
-	if (nfss->acregmax != 60*HZ)
-		seq_printf(m, ",acregmax=%d", nfss->acregmax/HZ);
-	if (nfss->acdirmin != 30*HZ)
-		seq_printf(m, ",acdirmin=%d", nfss->acdirmin/HZ);
-	if (nfss->acdirmax != 60*HZ)
-		seq_printf(m, ",acdirmax=%d", nfss->acdirmax/HZ);
-	for (nfs_infop = nfs_info; nfs_infop->flag; nfs_infop++) {
-		if (nfss->flags & nfs_infop->flag)
-			seq_puts(m, nfs_infop->str);
-		else
-			seq_puts(m, nfs_infop->nostr);
-	}
-	seq_puts(m, ",addr=");
-	seq_escape(m, nfss->hostname, " \t\n\\");
-	return 0;
-}
-
-/*
- * Invalidate the local caches
- */
-void
-nfs_zap_caches(struct inode *inode)
-{
-	struct nfs_inode *nfsi = NFS_I(inode);
-	int mode = inode->i_mode;
+	void (*clear_acl_cache)(struct inode *);
 
-	NFS_ATTRTIMEO(inode) = NFS_MINATTRTIMEO(inode);
-	NFS_ATTRTIMEO_UPDATE(inode) = jiffies;
-
-	memset(NFS_COOKIEVERF(inode), 0, sizeof(NFS_COOKIEVERF(inode)));
-	if (S_ISREG(mode) || S_ISDIR(mode) || S_ISLNK(mode))
-		nfsi->flags |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA;
-	else
-		nfsi->flags |= NFS_INO_INVALID_ATTR;
+	clear_acl_cache = NFS_PROTO(inode)->clear_acl_cache;
+	if (clear_acl_cache != NULL)
+		clear_acl_cache(inode);
+	spin_lock(&inode->i_lock);
+	NFS_I(inode)->cache_validity &= ~NFS_INO_INVALID_ACL;
+	spin_unlock(&inode->i_lock);
 }
 
 /*
- * Invalidate, but do not unhash, the inode
+ * Invalidate, but do not unhash, the inode.
+ * NB: must be called with inode->i_lock held!
  */
-static void
-nfs_invalidate_inode(struct inode *inode)
+static void nfs_invalidate_inode(struct inode *inode)
 {
-	umode_t save_mode = inode->i_mode;
-
-	make_bad_inode(inode);
-	inode->i_mode = save_mode;
-	nfs_zap_caches(inode);
+	set_bit(NFS_INO_STALE, &NFS_FLAGS(inode));
+	nfs_zap_caches_locked(inode);
 }
 
 struct nfs_find_desc {
@@ -611,7 +179,7 @@ nfs_find_actor(struct inode *inode, void *opaque)
 		return 0;
 	if (nfs_compare_fh(NFS_FH(inode), fh))
 		return 0;
-	if (is_bad_inode(inode))
+	if (is_bad_inode(inode) || NFS_STALE(inode))
 		return 0;
 	return 1;
 }
@@ -641,8 +209,9 @@ nfs_fhget(struct super_block *sb, struct nfs_fh *fh, struct nfs_fattr *fattr)
 		.fh	= fh,
 		.fattr	= fattr
 	};
-	struct inode *inode = NULL;
+	struct inode *inode = ERR_PTR(-ENOENT);
 	unsigned long hash;
+	int maycache = 1;
 
 	if ((fattr->valid & NFS_ATTR_FATTR) == 0)
 		goto out_no_inode;
@@ -652,17 +221,23 @@ nfs_fhget(struct super_block *sb, struct nfs_fh *fh, struct nfs_fattr *fattr)
 		goto out_no_inode;
 	}
 
-	hash = nfs_fattr_to_ino_t(fattr);
+	hash = fattr->fileid;
+	if (sizeof(hash) < sizeof(u64))
+		hash ^= fattr->fileid >> (sizeof(u64) - sizeof(ino_t)) * 8;
 
-	if (!(inode = iget5_locked(sb, hash, nfs_find_actor, nfs_init_locked, &desc)))
+	inode = iget5_locked(sb, hash, nfs_find_actor, nfs_init_locked, &desc);
+	if (inode == NULL) {
+		inode = ERR_PTR(-ENOMEM);
 		goto out_no_inode;
+	}
 
 	if (inode->i_state & I_NEW) {
 		struct nfs_inode *nfsi = NFS_I(inode);
 
-		/* We set i_ino for the few things that still rely on it,
-		 * such as stat(2) */
-		inode->i_ino = hash;
+		/* We set i_ino for the few things that still rely on it, such
+		 * as printing messages; stat and filldir use the fileid
+		 * directly since i_ino may not be large enough */
+		inode->i_ino = fattr->fileid;
 
 		/* We can't support update_atime(), since the server will reset it */
 		inode->i_flags |= S_NOATIME|S_NOCMTIME;
@@ -670,23 +245,33 @@ nfs_fhget(struct super_block *sb, struct nfs_fh *fh, struct nfs_fattr *fattr)
 		/* Why so? Because we want revalidate for devices/FIFOs, and
 		 * that's precisely what we have in nfs_file_inode_operations.
 		 */
-		inode->i_op = &nfs_file_inode_operations;
+		inode->i_op = NFS_SB(sb)->nfs_client->rpc_ops->file_inode_ops;
 		if (S_ISREG(inode->i_mode)) {
 			inode->i_fop = &nfs_file_operations;
 			inode->i_data.a_ops = &nfs_file_aops;
 			inode->i_data.backing_dev_info = &NFS_SB(sb)->backing_dev_info;
 		} else if (S_ISDIR(inode->i_mode)) {
-			inode->i_op = NFS_SB(sb)->rpc_ops->dir_inode_ops;
+			inode->i_op = NFS_SB(sb)->nfs_client->rpc_ops->dir_inode_ops;
 			inode->i_fop = &nfs_dir_operations;
 			if (nfs_server_capable(inode, NFS_CAP_READDIRPLUS)
 			    && fattr->size <= NFS_LIMIT_READDIRPLUS)
-				NFS_FLAGS(inode) |= NFS_INO_ADVISE_RDPLUS;
+				set_bit(NFS_INO_ADVISE_RDPLUS, &NFS_FLAGS(inode));
+			/* Deal with crossing mountpoints */
+			if (!nfs_fsid_equal(&NFS_SB(sb)->fsid, &fattr->fsid)) {
+				if (fattr->valid & NFS_ATTR_FATTR_V4_REFERRAL)
+					inode->i_op = &nfs_referral_inode_operations;
+				else
+					inode->i_op = &nfs_mountpoint_inode_operations;
+				inode->i_fop = NULL;
+				maycache = 0;
+			}
 		} else if (S_ISLNK(inode->i_mode))
 			inode->i_op = &nfs_symlink_inode_operations;
 		else
 			init_special_inode(inode, inode->i_mode, fattr->rdev);
 
-		nfsi->read_cache_jiffies = fattr->timestamp;
+		nfsi->read_cache_jiffies = fattr->time_start;
+		nfsi->last_updated = jiffies;
 		inode->i_atime = fattr->atime;
 		inode->i_mtime = fattr->mtime;
 		inode->i_ctime = fattr->ctime;
@@ -703,16 +288,16 @@ nfs_fhget(struct super_block *sb, struct nfs_fh *fh, struct nfs_fattr *fattr)
 			 * report the blocks in 512byte units
 			 */
 			inode->i_blocks = nfs_calc_block_size(fattr->du.nfs3.used);
-			inode->i_blksize = inode->i_sb->s_blocksize;
 		} else {
 			inode->i_blocks = fattr->du.nfs2.blocks;
-			inode->i_blksize = fattr->du.nfs2.blocksize;
 		}
 		nfsi->attrtimeo = NFS_MINATTRTIMEO(inode);
 		nfsi->attrtimeo_timestamp = jiffies;
 		memset(nfsi->cookieverf, 0, sizeof(nfsi->cookieverf));
 		nfsi->cache_access.cred = NULL;
 
+		nfs_fscache_get_fh_cookie(sb, nfsi, maycache);
+
 		unlock_new_inode(inode);
 	} else
 		nfs_refresh_inode(inode, fattr);
@@ -723,14 +308,9 @@ nfs_fhget(struct super_block *sb, struct nfs_fh *fh, struct nfs_fattr *fattr)
 
 out:
 	return inode;
-/*
-fail_dlim:
-	make_bad_inode(inode);
-	iput(inode);
-	inode = NULL;
-*/
+
 out_no_inode:
-	printk("nfs_fhget: iget failed\n");
+	dprintk("nfs_fhget: iget failed with error %ld\n", PTR_ERR(inode));
 	goto out;
 }
 
@@ -743,6 +323,8 @@ nfs_setattr(struct dentry *dentry, struct iattr *attr)
 	struct nfs_fattr fattr;
 	int error;
 
+	nfs_inc_stats(inode, NFSIOS_VFSSETATTR);
+
 	if (attr->ia_valid & ATTR_SIZE) {
 		if (!S_ISREG(inode->i_mode) || attr->ia_size == i_size_read(inode))
 			attr->ia_valid &= ~ATTR_SIZE;
@@ -755,86 +337,125 @@ nfs_setattr(struct dentry *dentry, struct iattr *attr)
 
 	lock_kernel();
 	nfs_begin_data_update(inode);
-	/* Write all dirty data if we're changing file permissions or size */
-	if ((attr->ia_valid & (ATTR_MODE|ATTR_UID|ATTR_GID|ATTR_SIZE)) != 0) {
-		if (filemap_fdatawrite(inode->i_mapping) == 0)
-			filemap_fdatawait(inode->i_mapping);
-		nfs_wb_all(inode);
-	}
+	/* Write all dirty data */
+	filemap_write_and_wait(inode->i_mapping);
+	nfs_wb_all(inode);
+	/*
+	 * Return any delegations if we're going to change ACLs
+	 */
+	if ((attr->ia_valid & (ATTR_MODE|ATTR_UID|ATTR_GID)) != 0)
+		nfs_inode_return_delegation(inode);
 	error = NFS_PROTO(inode)->setattr(dentry, &fattr, attr);
-	if (error == 0) {
+	if (error == 0)
 		nfs_refresh_inode(inode, &fattr);
+	nfs_end_data_update(inode);
+	unlock_kernel();
+	return error;
+}
+
+/**
+ * nfs_setattr_update_inode - Update inode metadata after a setattr call.
+ * @inode: pointer to struct inode
+ * @attr: pointer to struct iattr
+ *
+ * Note: we do this in the *proc.c in order to ensure that
+ *       it works for things like exclusive creates too.
+ */
+void nfs_setattr_update_inode(struct inode *inode, struct iattr *attr)
+{
+	if ((attr->ia_valid & (ATTR_MODE|ATTR_UID|ATTR_GID)) != 0) {
 		if ((attr->ia_valid & ATTR_MODE) != 0) {
-			int mode;
-			mode = inode->i_mode & ~S_IALLUGO;
-			mode |= attr->ia_mode & S_IALLUGO;
+			int mode = attr->ia_mode & S_IALLUGO;
+			mode |= inode->i_mode & ~S_IALLUGO;
 			inode->i_mode = mode;
 		}
 		if ((attr->ia_valid & ATTR_UID) != 0)
 			inode->i_uid = attr->ia_uid;
 		if ((attr->ia_valid & ATTR_GID) != 0)
 			inode->i_gid = attr->ia_gid;
-		if ((attr->ia_valid & ATTR_XID) != 0)
+		if ((attr->ia_valid & ATTR_XID) && IS_TAGXID(inode))
 			inode->i_xid = attr->ia_xid;
-		if ((attr->ia_valid & ATTR_SIZE) != 0) {
-			inode->i_size = attr->ia_size;
-			vmtruncate(inode, attr->ia_size);
-		}
+		spin_lock(&inode->i_lock);
+		NFS_I(inode)->cache_validity |= NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL;
+		spin_unlock(&inode->i_lock);
 	}
-	if ((attr->ia_valid & (ATTR_MODE|ATTR_UID|ATTR_GID)) != 0) {
-		struct rpc_cred **cred = &NFS_I(inode)->cache_access.cred;
-		if (*cred) {
-			put_rpccred(*cred);
-			*cred = NULL;
-		}
+	if ((attr->ia_valid & ATTR_SIZE) != 0) {
+		nfs_inc_stats(inode, NFSIOS_SETATTRTRUNC);
+		inode->i_size = attr->ia_size;
+		nfs_fscache_set_size(NFS_SERVER(inode), NFS_I(inode), inode->i_size);
+		vmtruncate(inode, attr->ia_size);
 	}
-	nfs_end_data_update(inode);
-	unlock_kernel();
-	return error;
+}
+
+static int nfs_wait_schedule(void *word)
+{
+	if (signal_pending(current))
+		return -ERESTARTSYS;
+	schedule();
+	return 0;
 }
 
 /*
  * Wait for the inode to get unlocked.
- * (Used for NFS_INO_LOCKED and NFS_INO_REVALIDATING).
  */
-int
-nfs_wait_on_inode(struct inode *inode, int flag)
+static int nfs_wait_on_inode(struct inode *inode)
 {
 	struct rpc_clnt	*clnt = NFS_CLIENT(inode);
 	struct nfs_inode *nfsi = NFS_I(inode);
-
+	sigset_t oldmask;
 	int error;
-	if (!(NFS_FLAGS(inode) & flag))
-		return 0;
-	atomic_inc(&inode->i_count);
-	error = nfs_wait_event(clnt, nfsi->nfs_i_wait,
-				!(NFS_FLAGS(inode) & flag));
-	iput(inode);
+
+	rpc_clnt_sigmask(clnt, &oldmask);
+	error = wait_on_bit_lock(&nfsi->flags, NFS_INO_REVALIDATING,
+					nfs_wait_schedule, TASK_INTERRUPTIBLE);
+	rpc_clnt_sigunmask(clnt, &oldmask);
+
 	return error;
 }
 
+static void nfs_wake_up_inode(struct inode *inode)
+{
+	struct nfs_inode *nfsi = NFS_I(inode);
+
+	clear_bit(NFS_INO_REVALIDATING, &nfsi->flags);
+	smp_mb__after_clear_bit();
+	wake_up_bit(&nfsi->flags, NFS_INO_REVALIDATING);
+}
+
 int nfs_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat)
 {
 	struct inode *inode = dentry->d_inode;
-	struct nfs_inode *nfsi = NFS_I(inode);
-	int need_atime = nfsi->flags & NFS_INO_INVALID_ATIME;
+	int need_atime = NFS_I(inode)->cache_validity & NFS_INO_INVALID_ATIME;
 	int err;
 
-	if (__IS_FLG(inode, MS_NOATIME))
-		need_atime = 0;
-	else if (__IS_FLG(inode, MS_NODIRATIME) && S_ISDIR(inode->i_mode))
+	/* Flush out writes to the server in order to update c/mtime */
+	nfs_sync_inode_wait(inode, 0, 0, FLUSH_NOCOMMIT);
+
+	/*
+	 * We may force a getattr if the user cares about atime.
+	 *
+	 * Note that we only have to check the vfsmount flags here:
+	 *  - NFS always sets S_NOATIME by so checking it would give a
+	 *    bogus result
+	 *  - NFS never sets MS_NOATIME or MS_NODIRATIME so there is
+	 *    no point in checking those.
+	 */
+ 	if ((mnt->mnt_flags & MNT_NOATIME) ||
+ 	    ((mnt->mnt_flags & MNT_NODIRATIME) && S_ISDIR(inode->i_mode)))
 		need_atime = 0;
-	/* We may force a getattr if the user cares about atime */
+
 	if (need_atime)
 		err = __nfs_revalidate_inode(NFS_SERVER(inode), inode);
 	else
 		err = nfs_revalidate_inode(NFS_SERVER(inode), inode);
-	if (!err)
+	if (!err) {
 		generic_fillattr(inode, stat);
+		stat->ino = NFS_FILEID(inode);
+	}
 	return err;
 }
 
-struct nfs_open_context *alloc_nfs_open_context(struct dentry *dentry, struct rpc_cred *cred)
+static struct nfs_open_context *alloc_nfs_open_context(struct vfsmount *mnt, struct dentry *dentry, struct rpc_cred *cred)
 {
 	struct nfs_open_context *ctx;
 
@@ -842,11 +463,12 @@ struct nfs_open_context *alloc_nfs_open_context(struct dentry *dentry, struct rp
 	if (ctx != NULL) {
 		atomic_set(&ctx->count, 1);
 		ctx->dentry = dget(dentry);
+		ctx->vfsmnt = mntget(mnt);
 		ctx->cred = get_rpccred(cred);
 		ctx->state = NULL;
 		ctx->lockowner = current->files;
 		ctx->error = 0;
-		init_waitqueue_head(&ctx->waitq);
+		ctx->dir_cookie = 0;
 	}
 	return ctx;
 }
@@ -861,11 +483,18 @@ struct nfs_open_context *get_nfs_open_context(struct nfs_open_context *ctx)
 void put_nfs_open_context(struct nfs_open_context *ctx)
 {
 	if (atomic_dec_and_test(&ctx->count)) {
+		if (!list_empty(&ctx->list)) {
+			struct inode *inode = ctx->dentry->d_inode;
+			spin_lock(&inode->i_lock);
+			list_del(&ctx->list);
+			spin_unlock(&inode->i_lock);
+		}
 		if (ctx->state != NULL)
 			nfs4_close_state(ctx->state, ctx->mode);
 		if (ctx->cred != NULL)
 			put_rpccred(ctx->cred);
 		dput(ctx->dentry);
+		mntput(ctx->vfsmnt);
 		kfree(ctx);
 	}
 }
@@ -874,7 +503,7 @@ void put_nfs_open_context(struct nfs_open_context *ctx)
  * Ensure that mmap has a recent RPC credential for use when writing out
  * shared pages
  */
-void nfs_file_set_open_context(struct file *filp, struct nfs_open_context *ctx)
+static void nfs_file_set_open_context(struct file *filp, struct nfs_open_context *ctx)
 {
 	struct inode *inode = filp->f_dentry->d_inode;
 	struct nfs_inode *nfsi = NFS_I(inode);
@@ -885,13 +514,18 @@ void nfs_file_set_open_context(struct file *filp, struct nfs_open_context *ctx)
 	spin_unlock(&inode->i_lock);
 }
 
-struct nfs_open_context *nfs_find_open_context(struct inode *inode, int mode)
+/*
+ * Given an inode, search for an open context with the desired characteristics
+ */
+struct nfs_open_context *nfs_find_open_context(struct inode *inode, struct rpc_cred *cred, int mode)
 {
 	struct nfs_inode *nfsi = NFS_I(inode);
 	struct nfs_open_context *pos, *ctx = NULL;
 
 	spin_lock(&inode->i_lock);
 	list_for_each_entry(pos, &nfsi->open_files, list) {
+		if (cred != NULL && pos->cred != cred)
+			continue;
 		if ((pos->mode & mode) == mode) {
 			ctx = get_nfs_open_context(pos);
 			break;
@@ -901,7 +535,7 @@ struct nfs_open_context *nfs_find_open_context(struct inode *inode, int mode)
 	return ctx;
 }
 
-void nfs_file_clear_open_context(struct file *filp)
+static void nfs_file_clear_open_context(struct file *filp)
 {
 	struct inode *inode = filp->f_dentry->d_inode;
 	struct nfs_open_context *ctx = (struct nfs_open_context *)filp->private_data;
@@ -909,7 +543,7 @@ void nfs_file_clear_open_context(struct file *filp)
 	if (ctx) {
 		filp->private_data = NULL;
 		spin_lock(&inode->i_lock);
-		list_del(&ctx->list);
+		list_move_tail(&ctx->list, &NFS_I(inode)->open_files);
 		spin_unlock(&inode->i_lock);
 		put_nfs_open_context(ctx);
 	}
@@ -923,24 +557,23 @@ int nfs_open(struct inode *inode, struct file *filp)
 	struct nfs_open_context *ctx;
 	struct rpc_cred *cred;
 
-	if ((cred = rpcauth_lookupcred(NFS_CLIENT(inode)->cl_auth, 0)) == NULL)
-		return -ENOMEM;
-	ctx = alloc_nfs_open_context(filp->f_dentry, cred);
+	cred = rpcauth_lookupcred(NFS_CLIENT(inode)->cl_auth, 0);
+	if (IS_ERR(cred))
+		return PTR_ERR(cred);
+	ctx = alloc_nfs_open_context(filp->f_vfsmnt, filp->f_dentry, cred);
 	put_rpccred(cred);
 	if (ctx == NULL)
 		return -ENOMEM;
 	ctx->mode = filp->f_mode;
 	nfs_file_set_open_context(filp, ctx);
 	put_nfs_open_context(ctx);
-	if ((filp->f_mode & FMODE_WRITE) != 0)
-		nfs_begin_data_update(inode);
+	if ((filp->f_flags & O_ACCMODE) != O_RDONLY)
+		nfs_fscache_disable_fh_cookie(inode);
 	return 0;
 }
 
 int nfs_release(struct inode *inode, struct file *filp)
 {
-	if ((filp->f_mode & FMODE_WRITE) != 0)
-		nfs_end_data_update(inode);
 	nfs_file_clear_open_context(filp);
 	return 0;
 }
@@ -955,84 +588,65 @@ __nfs_revalidate_inode(struct nfs_server *server, struct inode *inode)
 	int		 status = -ESTALE;
 	struct nfs_fattr fattr;
 	struct nfs_inode *nfsi = NFS_I(inode);
-	unsigned long verifier;
-	unsigned int flags;
 
 	dfprintk(PAGECACHE, "NFS: revalidating (%s/%Ld)\n",
 		inode->i_sb->s_id, (long long)NFS_FILEID(inode));
 
+	nfs_inc_stats(inode, NFSIOS_INODEREVALIDATE);
 	lock_kernel();
 	if (!inode || is_bad_inode(inode))
  		goto out_nowait;
-	if (NFS_STALE(inode) && inode != inode->i_sb->s_root->d_inode)
+	if (NFS_STALE(inode))
  		goto out_nowait;
 
-	while (NFS_REVALIDATING(inode)) {
-		status = nfs_wait_on_inode(inode, NFS_INO_REVALIDATING);
-		if (status < 0)
-			goto out_nowait;
-		if (NFS_SERVER(inode)->flags & NFS_MOUNT_NOAC)
-			continue;
-		if (NFS_FLAGS(inode) & (NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA|NFS_INO_INVALID_ATIME))
-			continue;
-		status = NFS_STALE(inode) ? -ESTALE : 0;
-		goto out_nowait;
+	status = nfs_wait_on_inode(inode);
+	if (status < 0)
+		goto out;
+	if (NFS_STALE(inode)) {
+		status = -ESTALE;
+		/* Do we trust the cached ESTALE? */
+		if (NFS_ATTRTIMEO(inode) != 0) {
+			if (nfsi->cache_validity & (NFS_INO_INVALID_ATTR|NFS_INO_INVALID_ATIME)) {
+				/* no */
+			} else
+				goto out;
+		}
 	}
-	NFS_FLAGS(inode) |= NFS_INO_REVALIDATING;
 
-	/* Protect against RPC races by saving the change attribute */
-	verifier = nfs_save_change_attribute(inode);
 	status = NFS_PROTO(inode)->getattr(server, NFS_FH(inode), &fattr);
-	if (status) {
+	if (status != 0) {
 		dfprintk(PAGECACHE, "nfs_revalidate_inode: (%s/%Ld) getattr failed, error=%d\n",
 			 inode->i_sb->s_id,
 			 (long long)NFS_FILEID(inode), status);
 		if (status == -ESTALE) {
-			NFS_FLAGS(inode) |= NFS_INO_STALE;
-			if (inode != inode->i_sb->s_root->d_inode)
-				remove_inode_hash(inode);
+			nfs_zap_caches(inode);
+			if (!S_ISDIR(inode->i_mode))
+				set_bit(NFS_INO_STALE, &NFS_FLAGS(inode));
 		}
 		goto out;
 	}
 
-	status = nfs_update_inode(inode, &fattr, verifier);
+	spin_lock(&inode->i_lock);
+	status = nfs_update_inode(inode, &fattr);
 	if (status) {
+		spin_unlock(&inode->i_lock);
 		dfprintk(PAGECACHE, "nfs_revalidate_inode: (%s/%Ld) refresh failed, error=%d\n",
 			 inode->i_sb->s_id,
 			 (long long)NFS_FILEID(inode), status);
 		goto out;
 	}
-	flags = nfsi->flags;
-	/*
-	 * We may need to keep the attributes marked as invalid if
-	 * we raced with nfs_end_attr_update().
-	 */
-	if (verifier == nfsi->cache_change_attribute)
-		nfsi->flags &= ~(NFS_INO_INVALID_ATTR|NFS_INO_INVALID_ATIME);
-	/* Do the page cache invalidation */
-	if (flags & NFS_INO_INVALID_DATA) {
-		if (S_ISREG(inode->i_mode)) {
-			if (filemap_fdatawrite(inode->i_mapping) == 0)
-				filemap_fdatawait(inode->i_mapping);
-			nfs_wb_all(inode);
-		}
-		nfsi->flags &= ~NFS_INO_INVALID_DATA;
-		invalidate_inode_pages2(inode->i_mapping);
-		memset(NFS_COOKIEVERF(inode), 0, sizeof(NFS_COOKIEVERF(inode)));
-		dfprintk(PAGECACHE, "NFS: (%s/%Ld) data cache invalidated\n",
-				inode->i_sb->s_id,
-				(long long)NFS_FILEID(inode));
-		/* This ensures we revalidate dentries */
-		nfsi->cache_change_attribute++;
-	}
+	spin_unlock(&inode->i_lock);
+
+	if (nfsi->cache_validity & NFS_INO_INVALID_ACL)
+		nfs_zap_acl_cache(inode);
+
 	dfprintk(PAGECACHE, "NFS: (%s/%Ld) revalidation complete\n",
 		inode->i_sb->s_id,
 		(long long)NFS_FILEID(inode));
 
-	NFS_FLAGS(inode) &= ~NFS_INO_STALE;
-out:
-	NFS_FLAGS(inode) &= ~NFS_INO_REVALIDATING;
-	wake_up(&nfsi->nfs_i_wait);
+ out:
+	nfs_wake_up_inode(inode);
+
  out_nowait:
 	unlock_kernel();
 	return status;
@@ -1056,12 +670,52 @@ int nfs_attribute_timeout(struct inode *inode)
  */
 int nfs_revalidate_inode(struct nfs_server *server, struct inode *inode)
 {
-	if (!(NFS_FLAGS(inode) & (NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA))
+	if (!(NFS_I(inode)->cache_validity & NFS_INO_INVALID_ATTR)
 			&& !nfs_attribute_timeout(inode))
 		return NFS_STALE(inode) ? -ESTALE : 0;
 	return __nfs_revalidate_inode(server, inode);
 }
 
+/**
+ * nfs_revalidate_mapping - Revalidate the pagecache
+ * @inode - pointer to host inode
+ * @mapping - pointer to mapping
+ */
+int nfs_revalidate_mapping(struct inode *inode, struct address_space *mapping)
+{
+	struct nfs_inode *nfsi = NFS_I(inode);
+	int ret = 0;
+
+	if (NFS_STALE(inode))
+		ret = -ESTALE;
+	if ((nfsi->cache_validity & NFS_INO_REVAL_PAGECACHE)
+			|| nfs_attribute_timeout(inode))
+		ret = __nfs_revalidate_inode(NFS_SERVER(inode), inode);
+
+	if (nfsi->cache_validity & NFS_INO_INVALID_DATA) {
+		nfs_inc_stats(inode, NFSIOS_DATAINVALIDATE);
+		if (S_ISREG(inode->i_mode))
+			nfs_sync_mapping(mapping);
+		invalidate_inode_pages2(mapping);
+
+		spin_lock(&inode->i_lock);
+		nfsi->cache_validity &= ~NFS_INO_INVALID_DATA;
+		if (S_ISDIR(inode->i_mode)) {
+			memset(nfsi->cookieverf, 0, sizeof(nfsi->cookieverf));
+			/* This ensures we revalidate child dentries */
+			nfsi->cache_change_attribute = jiffies;
+		}
+		spin_unlock(&inode->i_lock);
+
+		nfs_fscache_renew_fh_cookie(NFS_SERVER(inode), nfsi);
+
+		dfprintk(PAGECACHE, "NFS: (%s/%Ld) data cache invalidated\n",
+				inode->i_sb->s_id,
+				(long long)NFS_FILEID(inode));
+	}
+	return ret;
+}
+
 /**
  * nfs_begin_data_update
  * @inode - pointer to inode
@@ -1084,39 +738,40 @@ void nfs_end_data_update(struct inode *inode)
 	struct nfs_inode *nfsi = NFS_I(inode);
 
 	if (!nfs_have_delegation(inode, FMODE_READ)) {
-		/* Mark the attribute cache for revalidation */
-		nfsi->flags |= NFS_INO_INVALID_ATTR;
-		/* Directories and symlinks: invalidate page cache too */
-		if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode))
-			nfsi->flags |= NFS_INO_INVALID_DATA;
+		/* Directories and symlinks: invalidate page cache */
+		if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode)) {
+			spin_lock(&inode->i_lock);
+			nfsi->cache_validity |= NFS_INO_INVALID_DATA;
+			spin_unlock(&inode->i_lock);
+		}
 	}
-	nfsi->cache_change_attribute ++;
+	nfsi->cache_change_attribute = jiffies;
 	atomic_dec(&nfsi->data_updates);
 }
 
-/**
- * nfs_end_data_update_defer
- * @inode - pointer to inode
- * Declare end of the operations that will update file data
- * This will defer marking the inode as needing revalidation
- * unless there are no other pending updates.
- */
-void nfs_end_data_update_defer(struct inode *inode)
+static void nfs_wcc_update_inode(struct inode *inode, struct nfs_fattr *fattr)
 {
 	struct nfs_inode *nfsi = NFS_I(inode);
 
-	if (atomic_dec_and_test(&nfsi->data_updates)) {
-		/* Mark the attribute cache for revalidation */
-		nfsi->flags |= NFS_INO_INVALID_ATTR;
-		/* Directories and symlinks: invalidate page cache too */
-		if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode))
-			nfsi->flags |= NFS_INO_INVALID_DATA;
-		nfsi->cache_change_attribute ++;
+	/* If we have atomic WCC data, we may update some attributes */
+	if ((fattr->valid & NFS_ATTR_WCC) != 0) {
+		if (timespec_equal(&inode->i_ctime, &fattr->pre_ctime)) {
+			memcpy(&inode->i_ctime, &fattr->ctime, sizeof(inode->i_ctime));
+			nfsi->cache_change_attribute = jiffies;
+		}
+		if (timespec_equal(&inode->i_mtime, &fattr->pre_mtime)) {
+			memcpy(&inode->i_mtime, &fattr->mtime, sizeof(inode->i_mtime));
+			nfsi->cache_change_attribute = jiffies;
+		}
+		if (inode->i_size == fattr->pre_size && nfsi->npages == 0) {
+			inode->i_size = fattr->size;
+			nfsi->cache_change_attribute = jiffies;
+		}
 	}
 }
 
 /**
- * nfs_refresh_inode - verify consistency of the inode attribute cache
+ * nfs_check_inode_attributes - verify consistency of the inode attribute cache
  * @inode - pointer to inode
  * @fattr - updated attributes
  *
@@ -1124,56 +779,40 @@ void nfs_end_data_update_defer(struct inode *inode)
  * so that fattr carries weak cache consistency data, then it may
  * also update the ctime/mtime/change_attribute.
  */
-int nfs_refresh_inode(struct inode *inode, struct nfs_fattr *fattr)
+static int nfs_check_inode_attributes(struct inode *inode, struct nfs_fattr *fattr)
 {
 	struct nfs_inode *nfsi = NFS_I(inode);
 	loff_t cur_size, new_isize;
 	int data_unstable;
 	uid_t uid;
 	gid_t gid;
-	xid_t xid = 0;
+	xid_t xid;
 
-	/* Do we hold a delegation? */
-	if (nfs_have_delegation(inode, FMODE_READ))
-		return 0;
+
+	/* Has the inode gone and changed behind our back? */
+	if (nfsi->fileid != fattr->fileid
+			|| (inode->i_mode & S_IFMT) != (fattr->mode & S_IFMT)) {
+		return -EIO;
+	}
 
 	/* Are we in the process of updating data on the server? */
 	data_unstable = nfs_caches_unstable(inode);
 
-	if (fattr->valid & NFS_ATTR_FATTR_V4) {
-		if ((fattr->valid & NFS_ATTR_PRE_CHANGE) != 0
-				&& nfsi->change_attr == fattr->pre_change_attr)
-			nfsi->change_attr = fattr->change_attr;
-		if (!data_unstable && nfsi->change_attr != fattr->change_attr)
-			nfsi->flags |= NFS_INO_INVALID_ATTR;
-	}
+	/* Do atomic weak cache consistency updates */
+	nfs_wcc_update_inode(inode, fattr);
 
-	if ((fattr->valid & NFS_ATTR_FATTR) == 0)
-		return 0;
+	if ((fattr->valid & NFS_ATTR_FATTR_V4) != 0 &&
+			nfsi->change_attr != fattr->change_attr)
+		nfsi->cache_validity |= NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE;
 
-	/* Has the inode gone and changed behind our back? */
-	if (nfsi->fileid != fattr->fileid
-			|| (inode->i_mode & S_IFMT) != (fattr->mode & S_IFMT))
-		return -EIO;
+	/* Verify a few of the more important attributes */
+	if (!timespec_equal(&inode->i_mtime, &fattr->mtime))
+		nfsi->cache_validity |= NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE;
 
 	cur_size = i_size_read(inode);
  	new_isize = nfs_size_to_loff_t(fattr->size);
-
-	/* If we have atomic WCC data, we may update some attributes */
-	if ((fattr->valid & NFS_ATTR_WCC) != 0) {
-		if (timespec_equal(&inode->i_ctime, &fattr->pre_ctime))
-			memcpy(&inode->i_ctime, &fattr->ctime, sizeof(inode->i_ctime));
-		if (timespec_equal(&inode->i_mtime, &fattr->pre_mtime))
-			memcpy(&inode->i_mtime, &fattr->mtime, sizeof(inode->i_mtime));
-	}
-
-	/* Verify a few of the more important attributes */
-	if (!data_unstable) {
-		if (!timespec_equal(&inode->i_mtime, &fattr->mtime)
-				|| cur_size != new_isize)
-			nfsi->flags |= NFS_INO_INVALID_ATTR;
-	} else if (S_ISREG(inode->i_mode) && new_isize > cur_size)
-			nfsi->flags |= NFS_INO_INVALID_ATTR;
+	if (cur_size != new_isize && nfsi->npages == 0)
+		nfsi->cache_validity |= NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE;
 
 	uid = INOXID_UID(XID_TAG(inode), fattr->uid, fattr->gid);
 	gid = INOXID_GID(XID_TAG(inode), fattr->uid, fattr->gid);
@@ -1184,19 +823,70 @@ int nfs_refresh_inode(struct inode *inode, struct nfs_fattr *fattr)
 			|| inode->i_uid != uid
 			|| inode->i_gid != gid
 			|| inode->i_xid != xid)
-		nfsi->flags |= NFS_INO_INVALID_ATTR;
+		nfsi->cache_validity |= NFS_INO_INVALID_ATTR | NFS_INO_INVALID_ACCESS | NFS_INO_INVALID_ACL;
 
 	/* Has the link count changed? */
 	if (inode->i_nlink != fattr->nlink)
-		nfsi->flags |= NFS_INO_INVALID_ATTR;
+		nfsi->cache_validity |= NFS_INO_INVALID_ATTR;
 
 	if (!timespec_equal(&inode->i_atime, &fattr->atime))
-		nfsi->flags |= NFS_INO_INVALID_ATIME;
+		nfsi->cache_validity |= NFS_INO_INVALID_ATIME;
 
-	nfsi->read_cache_jiffies = fattr->timestamp;
+	nfsi->read_cache_jiffies = fattr->time_start;
 	return 0;
 }
 
+/**
+ * nfs_refresh_inode - try to update the inode attribute cache
+ * @inode - pointer to inode
+ * @fattr - updated attributes
+ *
+ * Check that an RPC call that returned attributes has not overlapped with
+ * other recent updates of the inode metadata, then decide whether it is
+ * safe to do a full update of the inode attributes, or whether just to
+ * call nfs_check_inode_attributes.
+ */
+int nfs_refresh_inode(struct inode *inode, struct nfs_fattr *fattr)
+{
+	struct nfs_inode *nfsi = NFS_I(inode);
+	int status;
+
+	if ((fattr->valid & NFS_ATTR_FATTR) == 0)
+		return 0;
+	spin_lock(&inode->i_lock);
+	if (time_after(fattr->time_start, nfsi->last_updated))
+		status = nfs_update_inode(inode, fattr);
+	else
+		status = nfs_check_inode_attributes(inode, fattr);
+
+	spin_unlock(&inode->i_lock);
+	return status;
+}
+
+/**
+ * nfs_post_op_update_inode - try to update the inode attribute cache
+ * @inode - pointer to inode
+ * @fattr - updated attributes
+ *
+ * After an operation that has changed the inode metadata, mark the
+ * attribute cache as being invalid, then try to update it.
+ */
+int nfs_post_op_update_inode(struct inode *inode, struct nfs_fattr *fattr)
+{
+	struct nfs_inode *nfsi = NFS_I(inode);
+	int status = 0;
+
+	spin_lock(&inode->i_lock);
+	if (unlikely((fattr->valid & NFS_ATTR_FATTR) == 0)) {
+		nfsi->cache_validity |= NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE;
+		goto out;
+	}
+	status = nfs_update_inode(inode, fattr);
+out:
+	spin_unlock(&inode->i_lock);
+	return status;
+}
+
 /*
  * Many nfs protocol calls return the new file attributes after
  * an operation.  Here we update the inode to reflect the state
@@ -1209,33 +899,23 @@ int nfs_refresh_inode(struct inode *inode, struct nfs_fattr *fattr)
  *
  * A very similar scenario holds for the dir cache.
  */
-static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr, unsigned long verifier)
+static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr)
 {
+	struct nfs_server *server;
 	struct nfs_inode *nfsi = NFS_I(inode);
-	__u64		new_size;
-	loff_t		new_isize;
+	loff_t cur_isize, new_isize;
 	unsigned int	invalid = 0;
-	loff_t		cur_isize;
-	int data_unstable;
-	uid_t		uid;
-	gid_t		gid;
-	xid_t		xid = 0;
+	int data_stable;
+	uid_t uid;
+	gid_t gid;
+	xid_t xid;
 
 	dfprintk(VFS, "NFS: %s(%s/%ld ct=%d info=0x%x)\n",
 			__FUNCTION__, inode->i_sb->s_id, inode->i_ino,
 			atomic_read(&inode->i_count), fattr->valid);
 
-	if ((fattr->valid & NFS_ATTR_FATTR) == 0)
-		return 0;
-
-	if (nfsi->fileid != fattr->fileid) {
-		printk(KERN_ERR "%s: inode number mismatch\n"
-		       "expected (%s/0x%Lx), got (%s/0x%Lx)\n",
-		       __FUNCTION__,
-		       inode->i_sb->s_id, (long long)nfsi->fileid,
-		       inode->i_sb->s_id, (long long)fattr->fileid);
-		goto out_err;
-	}
+	if (nfsi->fileid != fattr->fileid)
+		goto out_fileid;
 
 	/*
 	 * Make sure the inode's type hasn't changed.
@@ -1243,63 +923,64 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr, unsign
 	if ((inode->i_mode & S_IFMT) != (fattr->mode & S_IFMT))
 		goto out_changed;
 
+	server = NFS_SERVER(inode);
+	/* Update the fsid if and only if this is the root directory */
+	if (inode == inode->i_sb->s_root->d_inode
+			&& !nfs_fsid_equal(&server->fsid, &fattr->fsid))
+		server->fsid = fattr->fsid;
+
 	/*
 	 * Update the read time so we don't revalidate too often.
 	 */
-	nfsi->read_cache_jiffies = fattr->timestamp;
+	nfsi->read_cache_jiffies = fattr->time_start;
+	nfsi->last_updated = jiffies;
 
 	/* Are we racing with known updates of the metadata on the server? */
-	data_unstable = ! nfs_verify_change_attribute(inode, verifier);
+	data_stable = nfs_verify_change_attribute(inode, fattr->time_start);
+	if (data_stable)
+		nfsi->cache_validity &= ~(NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE|NFS_INO_INVALID_ATIME);
+
+	/* Do atomic weak cache consistency updates */
+	nfs_wcc_update_inode(inode, fattr);
 
-	/* Check if the file size agrees */
-	new_size = fattr->size;
+	/* Check if our cached file size is stale */
  	new_isize = nfs_size_to_loff_t(fattr->size);
 	cur_isize = i_size_read(inode);
-	if (cur_isize != new_size) {
-#ifdef NFS_DEBUG_VERBOSE
-		printk(KERN_DEBUG "NFS: isize change on %s/%ld\n", inode->i_sb->s_id, inode->i_ino);
-#endif
-		/*
-		 * If we have pending writebacks, things can get
-		 * messy.
-		 */
-		if (S_ISREG(inode->i_mode) && data_unstable) {
-			if (new_isize > cur_isize) {
+	if (new_isize != cur_isize) {
+		/* Do we perhaps have any outstanding writes? */
+		if (nfsi->npages == 0) {
+			/* No, but did we race with nfs_end_data_update()? */
+			if (data_stable) {
 				inode->i_size = new_isize;
-				invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA;
+				invalid |= NFS_INO_INVALID_DATA;
+				nfs_fscache_set_size(NFS_SERVER(inode), nfsi, inode->i_size);
 			}
-		} else {
+			invalid |= NFS_INO_INVALID_ATTR;
+		} else if (new_isize > cur_isize) {
 			inode->i_size = new_isize;
 			invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA;
+			nfs_fscache_set_size(NFS_SERVER(inode), nfsi, inode->i_size);
 		}
+		nfsi->cache_change_attribute = jiffies;
+		dprintk("NFS: isize change on server for file %s/%ld\n",
+				inode->i_sb->s_id, inode->i_ino);
 	}
 
-	/*
-	 * Note: we don't check inode->i_mtime since pipes etc.
-	 *       can change this value in VFS without requiring a
-	 *	 cache revalidation.
-	 */
+	/* Check if the mtime agrees */
 	if (!timespec_equal(&inode->i_mtime, &fattr->mtime)) {
 		memcpy(&inode->i_mtime, &fattr->mtime, sizeof(inode->i_mtime));
-#ifdef NFS_DEBUG_VERBOSE
-		printk(KERN_DEBUG "NFS: mtime change on %s/%ld\n", inode->i_sb->s_id, inode->i_ino);
-#endif
-		if (!data_unstable)
-			invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA;
+		dprintk("NFS: mtime change on server for file %s/%ld\n",
+				inode->i_sb->s_id, inode->i_ino);
+		invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA;
+		nfsi->cache_change_attribute = jiffies;
 	}
 
-	if ((fattr->valid & NFS_ATTR_FATTR_V4)
-	    && nfsi->change_attr != fattr->change_attr) {
-#ifdef NFS_DEBUG_VERBOSE
-		printk(KERN_DEBUG "NFS: change_attr change on %s/%ld\n",
-		       inode->i_sb->s_id, inode->i_ino);
-#endif
-		nfsi->change_attr = fattr->change_attr;
-		if (!data_unstable)
-			invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA;
+	/* If ctime has changed we should definitely clear access+acl caches */
+	if (!timespec_equal(&inode->i_ctime, &fattr->ctime)) {
+		invalid |= NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL;
+		memcpy(&inode->i_ctime, &fattr->ctime, sizeof(inode->i_ctime));
+		nfsi->cache_change_attribute = jiffies;
 	}
-
-	memcpy(&inode->i_ctime, &fattr->ctime, sizeof(inode->i_ctime));
 	memcpy(&inode->i_atime, &fattr->atime, sizeof(inode->i_atime));
 
 	uid = INOXID_UID(XID_TAG(inode), fattr->uid, fattr->gid);
@@ -1309,14 +990,8 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr, unsign
 	if ((inode->i_mode & S_IALLUGO) != (fattr->mode & S_IALLUGO) ||
 	    inode->i_uid != uid ||
 	    inode->i_gid != gid ||
-	    inode->i_xid != xid) {
-		struct rpc_cred **cred = &NFS_I(inode)->cache_access.cred;
-		if (*cred) {
-			put_rpccred(*cred);
-			*cred = NULL;
-		}
-		invalid |= NFS_INO_INVALID_ATTR;
-	}
+	    inode->i_xid != xid)
+		invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL;
 
 	inode->i_mode = fattr->mode;
 	inode->i_nlink = fattr->nlink;
@@ -1329,14 +1004,22 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr, unsign
 		 * report the blocks in 512byte units
 		 */
 		inode->i_blocks = nfs_calc_block_size(fattr->du.nfs3.used);
-		inode->i_blksize = inode->i_sb->s_blocksize;
  	} else {
  		inode->i_blocks = fattr->du.nfs2.blocks;
- 		inode->i_blksize = fattr->du.nfs2.blocksize;
  	}
 
+	if ((fattr->valid & NFS_ATTR_FATTR_V4) != 0 &&
+			nfsi->change_attr != fattr->change_attr) {
+		dprintk("NFS: change_attr change on server for file %s/%ld\n",
+				inode->i_sb->s_id, inode->i_ino);
+		nfsi->change_attr = fattr->change_attr;
+		invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA|NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL;
+		nfsi->cache_change_attribute = jiffies;
+	}
+
 	/* Update attrtimeo value if we're out of the unstable period */
 	if (invalid & NFS_INO_INVALID_ATTR) {
+		nfs_inc_stats(inode, NFSIOS_ATTRINVALIDATE);
 		nfsi->attrtimeo = NFS_MINATTRTIMEO(inode);
 		nfsi->attrtimeo_timestamp = jiffies;
 	} else if (time_after(jiffies, nfsi->attrtimeo_timestamp+nfsi->attrtimeo)) {
@@ -1348,8 +1031,10 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr, unsign
 	if (!(S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode)
 				|| S_ISLNK(inode->i_mode)))
 		invalid &= ~NFS_INO_INVALID_DATA;
+	if (data_stable)
+		invalid &= ~(NFS_INO_INVALID_ATTR|NFS_INO_INVALID_ATIME|NFS_INO_REVAL_PAGECACHE);
 	if (!nfs_have_delegation(inode, FMODE_READ))
-		nfsi->flags |= invalid;
+		nfsi->cache_validity |= invalid;
 
 	return 0;
  out_changed:
@@ -1360,182 +1045,37 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr, unsign
 	printk(KERN_DEBUG "%s: inode %ld mode changed, %07o to %07o\n",
 			__FUNCTION__, inode->i_ino, inode->i_mode, fattr->mode);
 #endif
+ out_err:
 	/*
 	 * No need to worry about unhashing the dentry, as the
 	 * lookup validation will know that the inode is bad.
 	 * (But we fall through to invalidate the caches.)
 	 */
 	nfs_invalidate_inode(inode);
- out_err:
-	return -EIO;
-}
-
-/*
- * File system information
- */
-
-static int nfs_set_super(struct super_block *s, void *data)
-{
-	s->s_fs_info = data;
-	return set_anon_super(s, data);
-}
- 
-static int nfs_compare_super(struct super_block *sb, void *data)
-{
-	struct nfs_server *server = data;
-	struct nfs_server *old = NFS_SB(sb);
-
-	if (old->addr.sin_addr.s_addr != server->addr.sin_addr.s_addr)
-		return 0;
-	if (old->addr.sin_port != server->addr.sin_port)
-		return 0;
-	return !nfs_compare_fh(&old->fh, &server->fh);
-}
-
-static struct super_block *nfs_get_sb(struct file_system_type *fs_type,
-	int flags, const char *dev_name, void *raw_data)
-{
-	int error;
-	struct nfs_server *server;
-	struct super_block *s;
-	struct nfs_fh *root;
-	struct nfs_mount_data *data = raw_data;
-
-	if (!data) {
-		printk("nfs_read_super: missing data argument\n");
-		return ERR_PTR(-EINVAL);
-	}
-
-	server = kmalloc(sizeof(struct nfs_server), GFP_KERNEL);
-	if (!server)
-		return ERR_PTR(-ENOMEM);
-	memset(server, 0, sizeof(struct nfs_server));
-	/* Zero out the NFS state stuff */
-	init_nfsv4_state(server);
-
-	if (data->version != NFS_MOUNT_VERSION) {
-		printk("nfs warning: mount version %s than kernel\n",
-			data->version < NFS_MOUNT_VERSION ? "older" : "newer");
-		if (data->version < 2)
-			data->namlen = 0;
-		if (data->version < 3)
-			data->bsize  = 0;
-		if (data->version < 4) {
-			data->flags &= ~NFS_MOUNT_VER3;
-			data->root.size = NFS2_FHSIZE;
-			memcpy(data->root.data, data->old_root.data, NFS2_FHSIZE);
-		}
-		if (data->version < 5)
-			data->flags &= ~NFS_MOUNT_SECFLAVOUR;
-	}
-
-	root = &server->fh;
-	if (data->flags & NFS_MOUNT_VER3)
-		root->size = data->root.size;
-	else
-		root->size = NFS2_FHSIZE;
-	if (root->size > sizeof(root->data)) {
-		printk("nfs_get_sb: invalid root filehandle\n");
-		kfree(server);
-		return ERR_PTR(-EINVAL);
-	}
-	memcpy(root->data, data->root.data, root->size);
-
-	/* We now require that the mount process passes the remote address */
-	memcpy(&server->addr, &data->addr, sizeof(server->addr));
-	if (server->addr.sin_addr.s_addr == INADDR_ANY) {
-		printk("NFS: mount program didn't pass remote address!\n");
-		kfree(server);
-		return ERR_PTR(-EINVAL);
-	}
-
-	s = sget(fs_type, nfs_compare_super, nfs_set_super, server);
-
-	if (IS_ERR(s) || s->s_root) {
-		kfree(server);
-		return s;
-	}
-
-	s->s_flags = flags;
-
-	/* Fire up rpciod if not yet running */
-	if (rpciod_up() != 0) {
-		printk(KERN_WARNING "NFS: couldn't start rpciod!\n");
-		kfree(server);
-		return ERR_PTR(-EIO);
-	}
-
-	error = nfs_fill_super(s, data, flags & MS_VERBOSE ? 1 : 0);
-	if (error) {
-		up_write(&s->s_umount);
-		deactivate_super(s);
-		return ERR_PTR(error);
-	}
-	s->s_flags |= MS_ACTIVE;
-	return s;
+	return -ESTALE;
+
+ out_fileid:
+	printk(KERN_ERR "NFS: server %s error: fileid changed\n"
+		"fsid %s: expected fileid 0x%Lx, got 0x%Lx\n",
+		NFS_SERVER(inode)->nfs_client->cl_hostname, inode->i_sb->s_id,
+		(long long)nfsi->fileid, (long long)fattr->fileid);
+	goto out_err;
 }
 
-static void nfs_kill_super(struct super_block *s)
-{
-	struct nfs_server *server = NFS_SB(s);
-
-	kill_anon_super(s);
-
-	nfs4_renewd_prepare_shutdown(server);
-
-	if (server->client != NULL && !IS_ERR(server->client))
-		rpc_shutdown_client(server->client);
-	if (server->client_sys != NULL && !IS_ERR(server->client_sys))
-		rpc_shutdown_client(server->client_sys);
-
-	if (!(server->flags & NFS_MOUNT_NONLM))
-		lockd_down();	/* release rpc.lockd */
-
-	rpciod_down();		/* release rpciod */
-
-	destroy_nfsv4_state(server);
-
-	if (server->hostname != NULL)
-		kfree(server->hostname);
-	kfree(server);
-}
-
-static struct file_system_type nfs_fs_type = {
-	.owner		= THIS_MODULE,
-	.name		= "nfs",
-	.get_sb		= nfs_get_sb,
-	.kill_sb	= nfs_kill_super,
-	.fs_flags	= FS_ODD_RENAME|FS_REVAL_DOT|FS_BINARY_MOUNTDATA,
-};
 
 #ifdef CONFIG_NFS_V4
 
-static void nfs4_clear_inode(struct inode *);
-
-
-static struct super_operations nfs4_sops = { 
-	.alloc_inode	= nfs_alloc_inode,
-	.destroy_inode	= nfs_destroy_inode,
-	.write_inode	= nfs_write_inode,
-	.delete_inode	= nfs_delete_inode,
-	.statfs		= nfs_statfs,
-	.clear_inode	= nfs4_clear_inode,
-	.umount_begin	= nfs_umount_begin,
-	.show_options	= nfs_show_options,
-};
-
 /*
  * Clean out any remaining NFSv4 state that might be left over due
  * to open() calls that passed nfs_atomic_lookup, but failed to call
  * nfs_open().
  */
-static void nfs4_clear_inode(struct inode *inode)
+void nfs4_clear_inode(struct inode *inode)
 {
 	struct nfs_inode *nfsi = NFS_I(inode);
 
 	/* If we are holding a delegation, return it! */
-	if (nfsi->delegation != NULL)
-		nfs_inode_return_delegation(inode);
+	nfs_inode_return_delegation(inode);
 	/* First call standard NFS clear_inode() code */
 	nfs_clear_inode(inode);
 	/* Now clear out any remaining state */
@@ -1554,326 +1094,42 @@ static void nfs4_clear_inode(struct inode *inode)
 		nfs4_close_state(state, state->state);
 	}
 }
-
-
-static int nfs4_fill_super(struct super_block *sb, struct nfs4_mount_data *data, int silent)
-{
-	struct nfs_server *server;
-	struct nfs4_client *clp = NULL;
-	struct rpc_xprt *xprt = NULL;
-	struct rpc_clnt *clnt = NULL;
-	struct rpc_timeout timeparms;
-	rpc_authflavor_t authflavour;
-	int proto, err = -EIO;
-
-	sb->s_blocksize_bits = 0;
-	sb->s_blocksize = 0;
-	server = NFS_SB(sb);
-	if (data->rsize != 0)
-		server->rsize = nfs_block_size(data->rsize, NULL);
-	if (data->wsize != 0)
-		server->wsize = nfs_block_size(data->wsize, NULL);
-	server->flags = data->flags & NFS_MOUNT_FLAGMASK;
-
-	/* NFSv4 doesn't use NLM locking */
-	server->flags |= NFS_MOUNT_NONLM;
-
-	server->acregmin = data->acregmin*HZ;
-	server->acregmax = data->acregmax*HZ;
-	server->acdirmin = data->acdirmin*HZ;
-	server->acdirmax = data->acdirmax*HZ;
-
-	server->rpc_ops = &nfs_v4_clientops;
-	/* Initialize timeout values */
-
-	timeparms.to_initval = data->timeo * HZ / 10;
-	timeparms.to_retries = data->retrans;
-	timeparms.to_exponential = 1;
-	if (!timeparms.to_retries)
-		timeparms.to_retries = 5;
-
-	proto = data->proto;
-	/* Which IP protocol do we use? */
-	switch (proto) {
-	case IPPROTO_TCP:
-		timeparms.to_maxval  = RPC_MAX_TCP_TIMEOUT;
-		if (!timeparms.to_initval)
-			timeparms.to_initval = 600 * HZ / 10;
-		break;
-	case IPPROTO_UDP:
-		timeparms.to_maxval  = RPC_MAX_UDP_TIMEOUT;
-		if (!timeparms.to_initval)
-			timeparms.to_initval = 11 * HZ / 10;
-		break;
-	default:
-		return -EINVAL;
-	}
-
-	clp = nfs4_get_client(&server->addr.sin_addr);
-	if (!clp) {
-		printk(KERN_WARNING "NFS: failed to create NFS4 client.\n");
-		return -EIO;
-	}
-
-	/* Now create transport and client */
-	authflavour = RPC_AUTH_UNIX;
-	if (data->auth_flavourlen != 0) {
-		if (data->auth_flavourlen > 1)
-			printk(KERN_INFO "NFS: cannot yet deal with multiple auth flavours.\n");
-		if (copy_from_user(&authflavour, data->auth_flavours, sizeof(authflavour))) {
-			err = -EFAULT;
-			goto out_fail;
-		}
-	}
-
-	down_write(&clp->cl_sem);
-	if (clp->cl_rpcclient == NULL) {
-		xprt = xprt_create_proto(proto, &server->addr, &timeparms);
-		if (IS_ERR(xprt)) {
-			up_write(&clp->cl_sem);
-			printk(KERN_WARNING "NFS: cannot create RPC transport.\n");
-			err = PTR_ERR(xprt);
-			goto out_fail;
-		}
-		clnt = rpc_create_client(xprt, server->hostname, &nfs_program,
-				server->rpc_ops->version, authflavour);
-		if (IS_ERR(clnt)) {
-			up_write(&clp->cl_sem);
-			printk(KERN_WARNING "NFS: cannot create RPC client.\n");
-			xprt_destroy(xprt);
-			err = PTR_ERR(clnt);
-			goto out_fail;
-		}
-		clnt->cl_chatty   = 1;
-		clp->cl_rpcclient = clnt;
-		clp->cl_cred = rpcauth_lookupcred(clnt->cl_auth, 0);
-		memcpy(clp->cl_ipaddr, server->ip_addr, sizeof(clp->cl_ipaddr));
-		nfs_idmap_new(clp);
-	}
-	if (list_empty(&clp->cl_superblocks)) {
-		err = nfs4_init_client(clp);
-		if (err != 0) {
-			up_write(&clp->cl_sem);
-			goto out_fail;
-		}
-	}
-	list_add_tail(&server->nfs4_siblings, &clp->cl_superblocks);
-	clnt = rpc_clone_client(clp->cl_rpcclient);
-	if (!IS_ERR(clnt))
-			server->nfs4_state = clp;
-	up_write(&clp->cl_sem);
-	clp = NULL;
-
-	if (IS_ERR(clnt)) {
-		printk(KERN_WARNING "NFS: cannot create RPC client.\n");
-		return PTR_ERR(clnt);
-	}
-
-	clnt->cl_intr     = (server->flags & NFS4_MOUNT_INTR) ? 1 : 0;
-	clnt->cl_softrtry = (server->flags & NFS4_MOUNT_SOFT) ? 1 : 0;
-	server->client    = clnt;
-
-	if (server->nfs4_state->cl_idmap == NULL) {
-		printk(KERN_WARNING "NFS: failed to create idmapper.\n");
-		return -ENOMEM;
-	}
-
-	if (clnt->cl_auth->au_flavor != authflavour) {
-		if (rpcauth_create(authflavour, clnt) == NULL) {
-			printk(KERN_WARNING "NFS: couldn't create credcache!\n");
-			return -ENOMEM;
-		}
-	}
-
-	sb->s_op = &nfs4_sops;
-	err = nfs_sb_init(sb, authflavour);
-	if (err == 0)
-		return 0;
-out_fail:
-	if (clp)
-		nfs4_put_client(clp);
-	return err;
-}
-
-static int nfs4_compare_super(struct super_block *sb, void *data)
-{
-	struct nfs_server *server = data;
-	struct nfs_server *old = NFS_SB(sb);
-
-	if (strcmp(server->hostname, old->hostname) != 0)
-		return 0;
-	if (strcmp(server->mnt_path, old->mnt_path) != 0)
-		return 0;
-	return 1;
-}
-
-static void *
-nfs_copy_user_string(char *dst, struct nfs_string *src, int maxlen)
-{
-	void *p = NULL;
-
-	if (!src->len)
-		return ERR_PTR(-EINVAL);
-	if (src->len < maxlen)
-		maxlen = src->len;
-	if (dst == NULL) {
-		p = dst = kmalloc(maxlen + 1, GFP_KERNEL);
-		if (p == NULL)
-			return ERR_PTR(-ENOMEM);
-	}
-	if (copy_from_user(dst, src->data, maxlen)) {
-		if (p != NULL)
-			kfree(p);
-		return ERR_PTR(-EFAULT);
-	}
-	dst[maxlen] = '\0';
-	return dst;
-}
-
-static struct super_block *nfs4_get_sb(struct file_system_type *fs_type,
-	int flags, const char *dev_name, void *raw_data)
-{
-	int error;
-	struct nfs_server *server;
-	struct super_block *s;
-	struct nfs4_mount_data *data = raw_data;
-	void *p;
-
-	if (!data) {
-		printk("nfs_read_super: missing data argument\n");
-		return ERR_PTR(-EINVAL);
-	}
-
-	server = kmalloc(sizeof(struct nfs_server), GFP_KERNEL);
-	if (!server)
-		return ERR_PTR(-ENOMEM);
-	memset(server, 0, sizeof(struct nfs_server));
-	/* Zero out the NFS state stuff */
-	init_nfsv4_state(server);
-
-	if (data->version != NFS4_MOUNT_VERSION) {
-		printk("nfs warning: mount version %s than kernel\n",
-			data->version < NFS4_MOUNT_VERSION ? "older" : "newer");
-	}
-
-	p = nfs_copy_user_string(NULL, &data->hostname, 256);
-	if (IS_ERR(p))
-		goto out_err;
-	server->hostname = p;
-
-	p = nfs_copy_user_string(NULL, &data->mnt_path, 1024);
-	if (IS_ERR(p))
-		goto out_err;
-	server->mnt_path = p;
-
-	p = nfs_copy_user_string(server->ip_addr, &data->client_addr,
-			sizeof(server->ip_addr) - 1);
-	if (IS_ERR(p))
-		goto out_err;
-
-	/* We now require that the mount process passes the remote address */
-	if (data->host_addrlen != sizeof(server->addr)) {
-		s = ERR_PTR(-EINVAL);
-		goto out_free;
-	}
-	if (copy_from_user(&server->addr, data->host_addr, sizeof(server->addr))) {
-		s = ERR_PTR(-EFAULT);
-		goto out_free;
-	}
-	if (server->addr.sin_family != AF_INET ||
-	    server->addr.sin_addr.s_addr == INADDR_ANY) {
-		printk("NFS: mount program didn't pass remote IP address!\n");
-		s = ERR_PTR(-EINVAL);
-		goto out_free;
-	}
-
-	s = sget(fs_type, nfs4_compare_super, nfs_set_super, server);
-
-	if (IS_ERR(s) || s->s_root)
-		goto out_free;
-
-	s->s_flags = flags;
-
-	/* Fire up rpciod if not yet running */
-	if (rpciod_up() != 0) {
-		printk(KERN_WARNING "NFS: couldn't start rpciod!\n");
-		s = ERR_PTR(-EIO);
-		goto out_free;
-	}
-
-	error = nfs4_fill_super(s, data, flags & MS_VERBOSE ? 1 : 0);
-	if (error) {
-		up_write(&s->s_umount);
-		deactivate_super(s);
-		return ERR_PTR(error);
-	}
-	s->s_flags |= MS_ACTIVE;
-	return s;
-out_err:
-	s = (struct super_block *)p;
-out_free:
-	if (server->mnt_path)
-		kfree(server->mnt_path);
-	if (server->hostname)
-		kfree(server->hostname);
-	kfree(server);
-	return s;
-}
-
-static void nfs4_kill_super(struct super_block *sb)
-{
-	nfs_return_all_delegations(sb);
-	nfs_kill_super(sb);
-}
-
-static struct file_system_type nfs4_fs_type = {
-	.owner		= THIS_MODULE,
-	.name		= "nfs4",
-	.get_sb		= nfs4_get_sb,
-	.kill_sb	= nfs4_kill_super,
-	.fs_flags	= FS_ODD_RENAME|FS_REVAL_DOT|FS_BINARY_MOUNTDATA,
-};
-
-#define nfs4_init_once(nfsi) \
-	do { \
-		INIT_LIST_HEAD(&(nfsi)->open_states); \
-		nfsi->delegation = NULL; \
-		nfsi->delegation_state = 0; \
-		init_rwsem(&nfsi->rwsem); \
-	} while(0)
-#define register_nfs4fs() register_filesystem(&nfs4_fs_type)
-#define unregister_nfs4fs() unregister_filesystem(&nfs4_fs_type)
-#else
-#define nfs4_init_once(nfsi) \
-	do { } while (0)
-#define register_nfs4fs() (0)
-#define unregister_nfs4fs()
 #endif
 
-extern int nfs_init_nfspagecache(void);
-extern void nfs_destroy_nfspagecache(void);
-extern int nfs_init_readpagecache(void);
-extern int nfs_destroy_readpagecache(void);
-extern int nfs_init_writepagecache(void);
-extern int nfs_destroy_writepagecache(void);
-
-static kmem_cache_t * nfs_inode_cachep;
-
-static struct inode *nfs_alloc_inode(struct super_block *sb)
+struct inode *nfs_alloc_inode(struct super_block *sb)
 {
 	struct nfs_inode *nfsi;
 	nfsi = (struct nfs_inode *)kmem_cache_alloc(nfs_inode_cachep, SLAB_KERNEL);
 	if (!nfsi)
 		return NULL;
-	nfsi->flags = 0;
+	nfsi->flags = 0UL;
+	nfsi->cache_validity = 0UL;
+	nfsi->cache_change_attribute = jiffies;
+#ifdef CONFIG_NFS_V3_ACL
+	nfsi->acl_access = ERR_PTR(-EAGAIN);
+	nfsi->acl_default = ERR_PTR(-EAGAIN);
+#endif
+#ifdef CONFIG_NFS_V4
+	nfsi->nfs4_acl = NULL;
+#endif /* CONFIG_NFS_V4 */
 	return &nfsi->vfs_inode;
 }
 
-static void nfs_destroy_inode(struct inode *inode)
+void nfs_destroy_inode(struct inode *inode)
 {
 	kmem_cache_free(nfs_inode_cachep, NFS_I(inode));
 }
 
+static inline void nfs4_init_once(struct nfs_inode *nfsi)
+{
+#ifdef CONFIG_NFS_V4
+	INIT_LIST_HEAD(&nfsi->open_states);
+	nfsi->delegation = NULL;
+	nfsi->delegation_state = 0;
+	init_rwsem(&nfsi->rwsem);
+#endif
+}
+
 static void init_once(void * foo, kmem_cache_t * cachep, unsigned long flags)
 {
 	struct nfs_inode *nfsi = (struct nfs_inode *) foo;
@@ -1890,16 +1146,16 @@ static void init_once(void * foo, kmem_cache_t * cachep, unsigned long flags)
 		nfsi->ndirty = 0;
 		nfsi->ncommit = 0;
 		nfsi->npages = 0;
-		init_waitqueue_head(&nfsi->nfs_i_wait);
 		nfs4_init_once(nfsi);
 	}
 }
  
-int nfs_init_inodecache(void)
+static int __init nfs_init_inodecache(void)
 {
 	nfs_inode_cachep = kmem_cache_create("nfs_inode_cache",
 					     sizeof(struct nfs_inode),
-					     0, SLAB_RECLAIM_ACCOUNT,
+					     0, (SLAB_RECLAIM_ACCOUNT|
+						SLAB_MEM_SPREAD),
 					     init_once, NULL);
 	if (nfs_inode_cachep == NULL)
 		return -ENOMEM;
@@ -1907,7 +1163,7 @@ int nfs_init_inodecache(void)
 	return 0;
 }
 
-void nfs_destroy_inodecache(void)
+static void nfs_destroy_inodecache(void)
 {
 	if (kmem_cache_destroy(nfs_inode_cachep))
 		printk(KERN_INFO "nfs_inode_cache: not all structures were freed\n");
@@ -1920,6 +1176,14 @@ static int __init init_nfs_fs(void)
 {
 	int err;
 
+	err = nfs_fscache_register();
+	if (err < 0)
+		goto out6;
+
+	err = nfs_fs_proc_init();
+	if (err)
+		goto out5;
+
 	err = nfs_init_nfspagecache();
 	if (err)
 		goto out4;
@@ -1936,17 +1200,22 @@ static int __init init_nfs_fs(void)
 	if (err)
 		goto out1;
 
+	err = nfs_init_directcache();
+	if (err)
+		goto out0;
+
 #ifdef CONFIG_PROC_FS
 	rpc_proc_register(&nfs_rpcstat);
 #endif
-        err = register_filesystem(&nfs_fs_type);
-	if (err)
-		goto out;
-	if ((err = register_nfs4fs()) != 0)
+	if ((err = register_nfs_fs()) != 0)
 		goto out;
 	return 0;
 out:
+#ifdef CONFIG_PROC_FS
 	rpc_proc_unregister("nfs");
+#endif
+	nfs_destroy_directcache();
+out0:
 	nfs_destroy_writepagecache();
 out1:
 	nfs_destroy_readpagecache();
@@ -1955,20 +1224,26 @@ out2:
 out3:
 	nfs_destroy_nfspagecache();
 out4:
+	nfs_fs_proc_exit();
+out5:
+	nfs_fscache_unregister();
+out6:
 	return err;
 }
 
 static void __exit exit_nfs_fs(void)
 {
+	nfs_destroy_directcache();
 	nfs_destroy_writepagecache();
 	nfs_destroy_readpagecache();
 	nfs_destroy_inodecache();
 	nfs_destroy_nfspagecache();
+	nfs_fscache_unregister();
 #ifdef CONFIG_PROC_FS
 	rpc_proc_unregister("nfs");
 #endif
-	unregister_filesystem(&nfs_fs_type);
-	unregister_nfs4fs();
+	unregister_nfs_fs();
+	nfs_fs_proc_exit();
 }
 
 /* Not quite true; I just maintain it */