92 files changed, 5097 insertions, 2759 deletions
diff --git a/fs/Kconfig b/fs/Kconfig
index ddc7462ddb56..ca9affd676ae 100644
--- a/fs/Kconfig
+++ b/fs/Kconfig
@@ -326,8 +326,8 @@ source "fs/xfs/Kconfig"
 source "fs/gfs2/Kconfig"
 config OCFS2_FS
-        tristate "OCFS2 file system support (EXPERIMENTAL)"
+        tristate "OCFS2 file system support"
-        depends on NET && SYSFS && EXPERIMENTAL
+        depends on NET && SYSFS
        select CONFIGFS_FS
        select JBD
        select CRC32
@@ -1472,8 +1472,8 @@ config NFS_V4
          If unsure, say N.
 config NFS_DIRECTIO
-        bool "Allow direct I/O on NFS files (EXPERIMENTAL)"
+        bool "Allow direct I/O on NFS files"
-        depends on NFS_FS && EXPERIMENTAL
+        depends on NFS_FS
        help
          This option enables applications to perform uncached I/O on files
          in NFS file systems using the O_DIRECT open() flag.  When O_DIRECT
diff --git a/fs/affs/affs.h b/fs/affs/affs.h
index 0ddd4cc0d1a0..1dc8438ef389 100644
--- a/fs/affs/affs.h
+++ b/fs/affs/affs.h
@@ -1,7 +1,6 @@
 #include <linux/types.h>
 #include <linux/fs.h>
 #include <linux/buffer_head.h>
-#include <linux/affs_fs.h>
 #include <linux/amigaffs.h>
 /* AmigaOS allows file names with up to 30 characters length.
diff --git a/fs/affs/super.c b/fs/affs/super.c
index 5200f4938df0..17352011ab67 100644
--- a/fs/affs/super.c
+++ b/fs/affs/super.c
@@ -14,6 +14,7 @@
 #include <linux/init.h>
 #include <linux/statfs.h>
 #include <linux/parser.h>
+#include <linux/magic.h>
 #include "affs.h"
 extern struct timezone sys_tz;
diff --git a/fs/autofs/autofs_i.h b/fs/autofs/autofs_i.h
index a62327f1bdff..c7700d9b3f96 100644
--- a/fs/autofs/autofs_i.h
+++ b/fs/autofs/autofs_i.h
@@ -37,8 +37,6 @@
 #define DPRINTK(D) ((void)0)
 #endif
-#define AUTOFS_SUPER_MAGIC 0x0187
 /*
 * If the daemon returns a negative response (AUTOFS_IOC_FAIL) then the
 * kernel will keep the negative response cached for up to the time given
diff --git a/fs/autofs/inode.c b/fs/autofs/inode.c
index 65e5ed42190e..af2efbbb5d76 100644
--- a/fs/autofs/inode.c
+++ b/fs/autofs/inode.c
@@ -16,6 +16,7 @@
 #include <linux/file.h>
 #include <linux/parser.h>
 #include <linux/bitops.h>
+#include <linux/magic.h>
 #include "autofs_i.h"
 #include <linux/module.h>
diff --git a/fs/autofs4/autofs_i.h b/fs/autofs4/autofs_i.h
index d6603d02304c..480ab178cba5 100644
--- a/fs/autofs4/autofs_i.h
+++ b/fs/autofs4/autofs_i.h
@@ -40,8 +40,6 @@
 #define DPRINTK(fmt,args...) do {} while(0)
 #endif
-#define AUTOFS_SUPER_MAGIC 0x0187
 /* Unified info structure.  This is pointed to by both the dentry and
   inode structures.  Each file in the filesystem has an instance of this
   structure.  It holds a reference to the dentry, so dentries are never
diff --git a/fs/autofs4/inode.c b/fs/autofs4/inode.c
index fde78b110ddd..11a6a9ae51b7 100644
--- a/fs/autofs4/inode.c
+++ b/fs/autofs4/inode.c
@@ -19,6 +19,7 @@
 #include <linux/parser.h>
 #include <linux/bitops.h>
 #include <linux/smp_lock.h>
+#include <linux/magic.h>
 #include "autofs_i.h"
 #include <linux/module.h>
diff --git a/fs/cifs/CHANGES b/fs/cifs/CHANGES
index 0feb3bd49cb8..1eb9a2ec0a3b 100644
--- a/fs/cifs/CHANGES
+++ b/fs/cifs/CHANGES
@@ -1,3 +1,7 @@
+Version 1.46
+------------
+Support deep tree mounts.  Better support OS/2, Win9x (DOS) time stamps.
 Version 1.45
 ------------
 Do not time out lockw calls when using posix extensions. Do not
@@ -6,7 +10,8 @@ on requests on other threads.  Improve POSIX locking emulation,
 (lock cancel now works, and unlock of merged range works even
 to Windows servers now).  Fix oops on mount to lanman servers
 (win9x, os/2 etc.) when null password.  Do not send listxattr
-(SMB to query all EAs) if nouser_xattr specified.
+(SMB to query all EAs) if nouser_xattr specified.  Fix SE Linux
+problem (instantiate inodes/dentries in right order for readdir).
 Version 1.44
 ------------
diff --git a/fs/cifs/cifs_fs_sb.h b/fs/cifs/cifs_fs_sb.h
index ad58eb0c4d6d..fd1e52ebcee6 100644
--- a/fs/cifs/cifs_fs_sb.h
+++ b/fs/cifs/cifs_fs_sb.h
@@ -40,5 +40,7 @@ struct cifs_sb_info {
        mode_t  mnt_file_mode;
        mode_t  mnt_dir_mode;
        int     mnt_cifs_flags;
+        int     prepathlen;
+        char *  prepath;
 };
 #endif                          /* _CIFS_FS_SB_H */
diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c
index 3cd750029be2..c3ef1c0d0e68 100644
--- a/fs/cifs/cifsfs.c
+++ b/fs/cifs/cifsfs.c
@@ -189,7 +189,6 @@ cifs_statfs(struct dentry *dentry, struct kstatfs *buf)
        buf->f_files = 0;       /* undefined */
        buf->f_ffree = 0;       /* unlimited */
-#ifdef CONFIG_CIFS_EXPERIMENTAL
 /* BB we could add a second check for a QFS Unix capability bit */
 /* BB FIXME check CIFS_POSIX_EXTENSIONS Unix cap first FIXME BB */
    if ((pTcon->ses->capabilities & CAP_UNIX) && (CIFS_POSIX_EXTENSIONS &
@@ -199,7 +198,6 @@ cifs_statfs(struct dentry *dentry, struct kstatfs *buf)
    /* Only need to call the old QFSInfo if failed
    on newer one */
    if(rc)
-#endif /* CIFS_EXPERIMENTAL */
        rc = CIFSSMBQFSInfo(xid, pTcon, buf);
        /* Old Windows servers do not support level 103, retry with level 
diff --git a/fs/cifs/cifsfs.h b/fs/cifs/cifsfs.h
index 39ee8ef3bdeb..bea875d9a46a 100644
--- a/fs/cifs/cifsfs.h
+++ b/fs/cifs/cifsfs.h
@@ -100,5 +100,5 @@ extern ssize_t	cifs_getxattr(struct dentry *, const char *, void *, size_t);
 extern ssize_t  cifs_listxattr(struct dentry *, char *, size_t);
 extern int cifs_ioctl (struct inode * inode, struct file * filep,
                       unsigned int command, unsigned long arg);
-#define CIFS_VERSION   "1.45"
+#define CIFS_VERSION   "1.46"
 #endif                          /* _CIFSFS_H */
diff --git a/fs/cifs/cifspdu.h b/fs/cifs/cifspdu.h
index 86239023545b..81df2bf8e75a 100644
--- a/fs/cifs/cifspdu.h
+++ b/fs/cifs/cifspdu.h
@@ -1344,6 +1344,7 @@ struct smb_t2_rsp {
 #define SMB_QUERY_ATTR_FLAGS            0x206  /* append,immutable etc. */
 #define SMB_QUERY_POSIX_PERMISSION      0x207
 #define SMB_QUERY_POSIX_LOCK            0x208
+/* #define SMB_POSIX_OPEN               0x209 */
 #define SMB_QUERY_FILE_INTERNAL_INFO    0x3ee
 #define SMB_QUERY_FILE_ACCESS_INFO      0x3f0
 #define SMB_QUERY_FILE_NAME_INFO2       0x3f1 /* 0x30 bytes */
@@ -1363,6 +1364,7 @@ struct smb_t2_rsp {
 #define SMB_SET_XATTR                   0x205
 #define SMB_SET_ATTR_FLAGS              0x206  /* append, immutable etc. */
 #define SMB_SET_POSIX_LOCK              0x208
+#define SMB_POSIX_OPEN                  0x209
 #define SMB_SET_FILE_BASIC_INFO2        0x3ec
 #define SMB_SET_FILE_RENAME_INFORMATION 0x3f2 /* BB check if qpathinfo level too */
 #define SMB_FILE_ALL_INFO2              0x3fa
diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c
index 5d394c726860..0e9ba0b9d71e 100644
--- a/fs/cifs/connect.c
+++ b/fs/cifs/connect.c
@@ -89,6 +89,7 @@ struct smb_vol {
        unsigned int wsize;
        unsigned int sockopt;
        unsigned short int port;
+        char * prepath;
 };
 static int ipv4_connect(struct sockaddr_in *psin_server, 
@@ -993,6 +994,28 @@ cifs_parse_mount_options(char *options, const char *devname,struct smb_vol *vol)
                                printk(KERN_WARNING "CIFS: domain name too long\n");
                                return 1;
                        }
+                } else if (strnicmp(data, "prefixpath", 10) == 0) {
+                        if (!value || !*value) {
+                                printk(KERN_WARNING
+                                       "CIFS: invalid path prefix\n");
+                                return 1;       /* needs_arg; */
+                        }
+                        if ((temp_len = strnlen(value, 1024)) < 1024) {
+                                if(value[0] != '/')
+                                        temp_len++;  /* missing leading slash */
+                                vol->prepath = kmalloc(temp_len+1,GFP_KERNEL);
+                                if(vol->prepath == NULL)
+                                        return 1;
+                                if(value[0] != '/') {
+                                        vol->prepath[0] = '/';
+                                        strcpy(vol->prepath+1,value);
+                                } else
+                                        strcpy(vol->prepath,value);
+                                cFYI(1,("prefix path %s",vol->prepath));
+                        } else {
+                                printk(KERN_WARNING "CIFS: prefix too long\n");
+                                return 1;
+                        }
                } else if (strnicmp(data, "iocharset", 9) == 0) {
                        if (!value || !*value) {
                                printk(KERN_WARNING "CIFS: invalid iocharset specified\n");
@@ -1605,6 +1628,7 @@ cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb,
        if (cifs_parse_mount_options(mount_data, devname, &volume_info)) {
                kfree(volume_info.UNC);
                kfree(volume_info.password);
+                kfree(volume_info.prepath);
                FreeXid(xid);
                return -EINVAL;
        }
@@ -1619,6 +1643,7 @@ cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb,
           locations such as env variables and files on disk */
                kfree(volume_info.UNC);
                kfree(volume_info.password);
+                kfree(volume_info.prepath);
                FreeXid(xid);
                return -EINVAL;
        }
@@ -1639,6 +1664,7 @@ cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb,
                        /* we failed translating address */
                        kfree(volume_info.UNC);
                        kfree(volume_info.password);
+                        kfree(volume_info.prepath);
                        FreeXid(xid);
                        return -EINVAL;
                }
@@ -1651,6 +1677,7 @@ cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb,
                cERROR(1,("Connecting to DFS root not implemented yet"));
                kfree(volume_info.UNC);
                kfree(volume_info.password);
+                kfree(volume_info.prepath);
                FreeXid(xid);
                return -EINVAL;
        } else /* which servers DFS root would we conect to */ {
@@ -1658,6 +1685,7 @@ cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb,
                       ("CIFS mount error: No UNC path (e.g. -o unc=//192.168.1.100/public) specified"));
                kfree(volume_info.UNC);
                kfree(volume_info.password);
+                kfree(volume_info.prepath);
                FreeXid(xid);
                return -EINVAL;
        }
@@ -1672,6 +1700,7 @@ cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb,
                        cERROR(1,("CIFS mount error: iocharset %s not found",volume_info.iocharset));
                        kfree(volume_info.UNC);
                        kfree(volume_info.password);
+                        kfree(volume_info.prepath);
                        FreeXid(xid);
                        return -ELIBACC;
                }
@@ -1688,6 +1717,7 @@ cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb,
        else {
                kfree(volume_info.UNC);
                kfree(volume_info.password);
+                kfree(volume_info.prepath);
                FreeXid(xid);
                return -EINVAL;
        }
@@ -1710,6 +1740,7 @@ cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb,
                                sock_release(csocket);
                        kfree(volume_info.UNC);
                        kfree(volume_info.password);
+                        kfree(volume_info.prepath);
                        FreeXid(xid);
                        return rc;
                }
@@ -1720,6 +1751,7 @@ cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb,
                        sock_release(csocket);
                        kfree(volume_info.UNC);
                        kfree(volume_info.password);
+                        kfree(volume_info.prepath);
                        FreeXid(xid);
                        return rc;
                } else {
@@ -1744,6 +1776,7 @@ cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb,
                                sock_release(csocket);
                                kfree(volume_info.UNC);
                                kfree(volume_info.password);
+                                kfree(volume_info.prepath);
                                FreeXid(xid);
                                return rc;
                        }
@@ -1831,6 +1864,14 @@ cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb,
                        /* Windows ME may prefer this */
                        cFYI(1,("readsize set to minimum 2048"));
                }
+                /* calculate prepath */
+                cifs_sb->prepath = volume_info.prepath;
+                if(cifs_sb->prepath) {
+                        cifs_sb->prepathlen = strlen(cifs_sb->prepath);
+                        cifs_sb->prepath[0] = CIFS_DIR_SEP(cifs_sb);
+                        volume_info.prepath = NULL;
+                } else 
+                        cifs_sb->prepathlen = 0;
                cifs_sb->mnt_uid = volume_info.linux_uid;
                cifs_sb->mnt_gid = volume_info.linux_gid;
                cifs_sb->mnt_file_mode = volume_info.file_mode;
@@ -2008,6 +2049,7 @@ cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb,
        the password ptr is put in the new session structure (in which case the
        password will be freed at unmount time) */
        kfree(volume_info.UNC);
+        kfree(volume_info.prepath);
        FreeXid(xid);
        return rc;
 }
@@ -3195,6 +3237,7 @@ cifs_umount(struct super_block *sb, struct cifs_sb_info *cifs_sb)
        int xid;
        struct cifsSesInfo *ses = NULL;
        struct task_struct *cifsd_task;
+        char * tmp;
        xid = GetXid();
@@ -3228,6 +3271,10 @@ cifs_umount(struct super_block *sb, struct cifs_sb_info *cifs_sb)
        }
        
        cifs_sb->tcon = NULL;
+        tmp = cifs_sb->prepath;
+        cifs_sb->prepathlen = 0;
+        cifs_sb->prepath = NULL;
+        kfree(tmp);
        if (ses)
                schedule_timeout_interruptible(msecs_to_jiffies(500));
        if (ses)
diff --git a/fs/cifs/dir.c b/fs/cifs/dir.c
index 914239d53634..66b825ade3e1 100644
--- a/fs/cifs/dir.c
+++ b/fs/cifs/dir.c
@@ -46,7 +46,8 @@ char *
 build_path_from_dentry(struct dentry *direntry)
 {
        struct dentry *temp;
-        int namelen = 0;
+        int namelen;
+        int pplen;
        char *full_path;
        char dirsep;
@@ -56,7 +57,9 @@ build_path_from_dentry(struct dentry *direntry)
                when the server crashed */
        dirsep = CIFS_DIR_SEP(CIFS_SB(direntry->d_sb));
+        pplen = CIFS_SB(direntry->d_sb)->prepathlen;
 cifs_bp_rename_retry:
+        namelen = pplen; 
        for (temp = direntry; !IS_ROOT(temp);) {
                namelen += (1 + temp->d_name.len);
                temp = temp->d_parent;
@@ -70,7 +73,6 @@ cifs_bp_rename_retry:
        if(full_path == NULL)
                return full_path;
        full_path[namelen] = 0; /* trailing null */
        for (temp = direntry; !IS_ROOT(temp);) {
                namelen -= 1 + temp->d_name.len;
                if (namelen < 0) {
@@ -79,7 +81,7 @@ cifs_bp_rename_retry:
                        full_path[namelen] = dirsep;
                        strncpy(full_path + namelen + 1, temp->d_name.name,
                                temp->d_name.len);
-                        cFYI(0, (" name: %s ", full_path + namelen));
+                        cFYI(0, ("name: %s", full_path + namelen));
                }
                temp = temp->d_parent;
                if(temp == NULL) {
@@ -88,18 +90,23 @@ cifs_bp_rename_retry:
                        return NULL;
                }
        }
-        if (namelen != 0) {
+        if (namelen != pplen) {
                cERROR(1,
-                       ("We did not end path lookup where we expected namelen is %d",
+                       ("did not end path lookup where expected namelen is %d",
                        namelen));
-                /* presumably this is only possible if we were racing with a rename 
+                /* presumably this is only possible if racing with a rename 
                of one of the parent directories  (we can not lock the dentries
                above us to prevent this, but retrying should be harmless) */
                kfree(full_path);
-                namelen = 0;
                goto cifs_bp_rename_retry;
        }
+        /* DIR_SEP already set for byte  0 / vs \ but not for
+           subsequent slashes in prepath which currently must
+           be entered the right way - not sure if there is an alternative
+           since the '\' is a valid posix character so we can not switch
+           those safely to '/' if any are found in the middle of the prepath */
+        /* BB test paths to Windows with '/' in the midst of prepath */
+        strncpy(full_path,CIFS_SB(direntry->d_sb)->prepath,pplen);
        return full_path;
 }
diff --git a/fs/cifs/file.c b/fs/cifs/file.c
index e9c5ba9084fc..ddb012a68023 100644
--- a/fs/cifs/file.c
+++ b/fs/cifs/file.c
@@ -752,6 +752,7 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock)
                        int stored_rc = 0;
                        struct cifsLockInfo *li, *tmp;
+                        rc = 0;
                        down(&fid->lock_sem);
                        list_for_each_entry_safe(li, tmp, &fid->llist, llist) {
                                if (pfLock->fl_start <= li->offset &&
@@ -766,7 +767,7 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock)
                                        kfree(li);
                                }
                        }
-                up(&fid->lock_sem);
+                        up(&fid->lock_sem);
                }
        }
diff --git a/fs/cifs/xattr.c b/fs/cifs/xattr.c
index 067648b7179b..18fcec190f8b 100644
--- a/fs/cifs/xattr.c
+++ b/fs/cifs/xattr.c
@@ -269,7 +269,7 @@ ssize_t cifs_getxattr(struct dentry * direntry, const char * ea_name,
                                rc = CIFSSMBGetCIFSACL(xid, pTcon, fid,
                                        ea_value, buf_size,
                                        ACL_TYPE_ACCESS);
-                                CIFSSMBClose(xid, pTcon, fid)
+                                CIFSSMBClose(xid, pTcon, fid);
                        }
                } */  /* BB enable after fixing up return data */
                                
diff --git a/fs/configfs/dir.c b/fs/configfs/dir.c
index df025453dd97..816e8ef64560 100644
--- a/fs/configfs/dir.c
+++ b/fs/configfs/dir.c
@@ -86,6 +86,32 @@ static struct configfs_dirent *configfs_new_dirent(struct configfs_dirent * pare
        return sd;
 }
+/*
+ *
+ * Return -EEXIST if there is already a configfs element with the same
+ * name for the same parent.
+ *
+ * called with parent inode's i_mutex held
+ */
+int configfs_dirent_exists(struct configfs_dirent *parent_sd,
+                           const unsigned char *new)
+{
+        struct configfs_dirent * sd;
+        list_for_each_entry(sd, &parent_sd->s_children, s_sibling) {
+                if (sd->s_element) {
+                        const unsigned char *existing = configfs_get_name(sd);
+                        if (strcmp(existing, new))
+                                continue;
+                        else
+                                return -EEXIST;
+                }
+        }
+        return 0;
+}
 int configfs_make_dirent(struct configfs_dirent * parent_sd,
                         struct dentry * dentry, void * element,
                         umode_t mode, int type)
@@ -136,8 +162,10 @@ static int create_dir(struct config_item * k, struct dentry * p,
        int error;
        umode_t mode = S_IFDIR| S_IRWXU | S_IRUGO | S_IXUGO;
-        error = configfs_make_dirent(p->d_fsdata, d, k, mode,
+        error = configfs_dirent_exists(p->d_fsdata, d->d_name.name);
-                                     CONFIGFS_DIR);
+        if (!error)
+                error = configfs_make_dirent(p->d_fsdata, d, k, mode,
+                                             CONFIGFS_DIR);
        if (!error) {
                error = configfs_create(d, mode, init_dir);
                if (!error) {
diff --git a/fs/dcache.c b/fs/dcache.c
index 1b4a3a34ec57..17b392a2049e 100644
--- a/fs/dcache.c
+++ b/fs/dcache.c
@@ -828,17 +828,19 @@ void d_instantiate(struct dentry *entry, struct inode * inode)
 * (or otherwise set) by the caller to indicate that it is now
 * in use by the dcache.
 */
-struct dentry *d_instantiate_unique(struct dentry *entry, struct inode *inode)
+static struct dentry *__d_instantiate_unique(struct dentry *entry,
+                                             struct inode *inode)
 {
        struct dentry *alias;
        int len = entry->d_name.len;
        const char *name = entry->d_name.name;
        unsigned int hash = entry->d_name.hash;
-        BUG_ON(!list_empty(&entry->d_alias));
+        if (!inode) {
-        spin_lock(&dcache_lock);
+                entry->d_inode = NULL;
-        if (!inode)
+                return NULL;
-                goto do_negative;
+        }
        list_for_each_entry(alias, &inode->i_dentry, d_alias) {
                struct qstr *qstr = &alias->d_name;
@@ -851,19 +853,35 @@ struct dentry *d_instantiate_unique(struct dentry *entry, struct inode *inode)
                if (memcmp(qstr->name, name, len))
                        continue;
                dget_locked(alias);
-                spin_unlock(&dcache_lock);
-                BUG_ON(!d_unhashed(alias));
-                iput(inode);
                return alias;
        }
        list_add(&entry->d_alias, &inode->i_dentry);
-do_negative:
        entry->d_inode = inode;
        fsnotify_d_instantiate(entry, inode);
-        spin_unlock(&dcache_lock);
-        security_d_instantiate(entry, inode);
        return NULL;
 }
+struct dentry *d_instantiate_unique(struct dentry *entry, struct inode *inode)
+{
+        struct dentry *result;
+        BUG_ON(!list_empty(&entry->d_alias));
+        spin_lock(&dcache_lock);
+        result = __d_instantiate_unique(entry, inode);
+        spin_unlock(&dcache_lock);
+        if (!result) {
+                security_d_instantiate(entry, inode);
+                return NULL;
+        }
+        BUG_ON(!d_unhashed(result));
+        iput(inode);
+        return result;
+}
 EXPORT_SYMBOL(d_instantiate_unique);
 /**
@@ -1235,6 +1253,11 @@ static void __d_rehash(struct dentry * entry, struct hlist_head *list)
        hlist_add_head_rcu(&entry->d_hash, list);
 }
+static void _d_rehash(struct dentry * entry)
+{
+        __d_rehash(entry, d_hash(entry->d_parent, entry->d_name.hash));
+}
 /**
 * d_rehash     - add an entry back to the hash
 * @entry: dentry to add to the hash
@@ -1244,11 +1267,9 @@ static void __d_rehash(struct dentry * entry, struct hlist_head *list)
 
 void d_rehash(struct dentry * entry)
 {
-        struct hlist_head *list = d_hash(entry->d_parent, entry->d_name.hash);
        spin_lock(&dcache_lock);
        spin_lock(&entry->d_lock);
-        __d_rehash(entry, list);
+        _d_rehash(entry);
        spin_unlock(&entry->d_lock);
        spin_unlock(&dcache_lock);
 }
@@ -1386,6 +1407,120 @@ already_unhashed:
        spin_unlock(&dcache_lock);
 }
+/*
+ * Prepare an anonymous dentry for life in the superblock's dentry tree as a
+ * named dentry in place of the dentry to be replaced.
+ */
+static void __d_materialise_dentry(struct dentry *dentry, struct dentry *anon)
+{
+        struct dentry *dparent, *aparent;
+        switch_names(dentry, anon);
+        do_switch(dentry->d_name.len, anon->d_name.len);
+        do_switch(dentry->d_name.hash, anon->d_name.hash);
+        dparent = dentry->d_parent;
+        aparent = anon->d_parent;
+        dentry->d_parent = (aparent == anon) ? dentry : aparent;
+        list_del(&dentry->d_u.d_child);
+        if (!IS_ROOT(dentry))
+                list_add(&dentry->d_u.d_child, &dentry->d_parent->d_subdirs);
+        else
+                INIT_LIST_HEAD(&dentry->d_u.d_child);
+        anon->d_parent = (dparent == dentry) ? anon : dparent;
+        list_del(&anon->d_u.d_child);
+        if (!IS_ROOT(anon))
+                list_add(&anon->d_u.d_child, &anon->d_parent->d_subdirs);
+        else
+                INIT_LIST_HEAD(&anon->d_u.d_child);
+        anon->d_flags &= ~DCACHE_DISCONNECTED;
+}
+/**
+ * d_materialise_unique - introduce an inode into the tree
+ * @dentry: candidate dentry
+ * @inode: inode to bind to the dentry, to which aliases may be attached
+ *
+ * Introduces an dentry into the tree, substituting an extant disconnected
+ * root directory alias in its place if there is one
+ */
+struct dentry *d_materialise_unique(struct dentry *dentry, struct inode *inode)
+{
+        struct dentry *alias, *actual;
+        BUG_ON(!d_unhashed(dentry));
+        spin_lock(&dcache_lock);
+        if (!inode) {
+                actual = dentry;
+                dentry->d_inode = NULL;
+                goto found_lock;
+        }
+        /* See if a disconnected directory already exists as an anonymous root
+         * that we should splice into the tree instead */
+        if (S_ISDIR(inode->i_mode) && (alias = __d_find_alias(inode, 1))) {
+                spin_lock(&alias->d_lock);
+                /* Is this a mountpoint that we could splice into our tree? */
+                if (IS_ROOT(alias))
+                        goto connect_mountpoint;
+                if (alias->d_name.len == dentry->d_name.len &&
+                    alias->d_parent == dentry->d_parent &&
+                    memcmp(alias->d_name.name,
+                           dentry->d_name.name,
+                           dentry->d_name.len) == 0)
+                        goto replace_with_alias;
+                spin_unlock(&alias->d_lock);
+                /* Doh! Seem to be aliasing directories for some reason... */
+                dput(alias);
+        }
+        /* Add a unique reference */
+        actual = __d_instantiate_unique(dentry, inode);
+        if (!actual)
+                actual = dentry;
+        else if (unlikely(!d_unhashed(actual)))
+                goto shouldnt_be_hashed;
+found_lock:
+        spin_lock(&actual->d_lock);
+found:
+        _d_rehash(actual);
+        spin_unlock(&actual->d_lock);
+        spin_unlock(&dcache_lock);
+        if (actual == dentry) {
+                security_d_instantiate(dentry, inode);
+                return NULL;
+        }
+        iput(inode);
+        return actual;
+        /* Convert the anonymous/root alias into an ordinary dentry */
+connect_mountpoint:
+        __d_materialise_dentry(dentry, alias);
+        /* Replace the candidate dentry with the alias in the tree */
+replace_with_alias:
+        __d_drop(alias);
+        actual = alias;
+        goto found;
+shouldnt_be_hashed:
+        spin_unlock(&dcache_lock);
+        BUG();
+        goto shouldnt_be_hashed;
+}
 /**
 * d_path - return the path of a dentry
 * @dentry: dentry to report
@@ -1784,6 +1919,7 @@ EXPORT_SYMBOL(d_instantiate);
 EXPORT_SYMBOL(d_invalidate);
 EXPORT_SYMBOL(d_lookup);
 EXPORT_SYMBOL(d_move);
+EXPORT_SYMBOL_GPL(d_materialise_unique);
 EXPORT_SYMBOL(d_path);
 EXPORT_SYMBOL(d_prune_aliases);
 EXPORT_SYMBOL(d_rehash);
diff --git a/fs/hpfs/hpfs_fn.h b/fs/hpfs/hpfs_fn.h
index f687d54ed442..32ab51e42b96 100644
--- a/fs/hpfs/hpfs_fn.h
+++ b/fs/hpfs/hpfs_fn.h
@@ -12,7 +12,6 @@
 #include <linux/mutex.h>
 #include <linux/pagemap.h>
 #include <linux/buffer_head.h>
-#include <linux/hpfs_fs.h>
 #include <linux/slab.h>
 #include <linux/smp_lock.h>
diff --git a/fs/hpfs/super.c b/fs/hpfs/super.c
index f798480a363f..8fe51c343786 100644
--- a/fs/hpfs/super.c
+++ b/fs/hpfs/super.c
@@ -11,6 +11,7 @@
 #include <linux/parser.h>
 #include <linux/init.h>
 #include <linux/statfs.h>
+#include <linux/magic.h>
 /* Mark the filesystem dirty, so that chkdsk checks it when os/2 booted */
diff --git a/fs/jffs2/jffs2_fs_i.h b/fs/jffs2/jffs2_fs_i.h
index 2e0cc8e00b85..3a566077ac95 100644
--- a/fs/jffs2/jffs2_fs_i.h
+++ b/fs/jffs2/jffs2_fs_i.h
@@ -41,11 +41,7 @@ struct jffs2_inode_info {
        uint16_t flags;
        uint8_t usercompr;
-#if !defined (__ECOS)
-#if LINUX_VERSION_CODE > KERNEL_VERSION(2,5,2)
        struct inode vfs_inode;
-#endif
-#endif
 #ifdef CONFIG_JFFS2_FS_POSIX_ACL
        struct posix_acl *i_acl_access;
        struct posix_acl *i_acl_default;
diff --git a/fs/lockd/clntproc.c b/fs/lockd/clntproc.c
index 89ba0df14c22..50dbb67ae0c4 100644
--- a/fs/lockd/clntproc.c
+++ b/fs/lockd/clntproc.c
@@ -151,11 +151,13 @@ static void nlmclnt_release_lockargs(struct nlm_rqst *req)
 int
 nlmclnt_proc(struct inode *inode, int cmd, struct file_lock *fl)
 {
+        struct rpc_clnt         *client = NFS_CLIENT(inode);
+        struct sockaddr_in      addr;
        struct nlm_host         *host;
        struct nlm_rqst         *call;
        sigset_t                oldset;
        unsigned long           flags;
-        int                     status, proto, vers;
+        int                     status, vers;
        vers = (NFS_PROTO(inode)->version == 3) ? 4 : 1;
        if (NFS_PROTO(inode)->version > 3) {
@@ -163,10 +165,8 @@ nlmclnt_proc(struct inode *inode, int cmd, struct file_lock *fl)
                return -ENOLCK;
        }
-        /* Retrieve transport protocol from NFS client */
+        rpc_peeraddr(client, (struct sockaddr *) &addr, sizeof(addr));
-        proto = NFS_CLIENT(inode)->cl_xprt->prot;
+        host = nlmclnt_lookup_host(&addr, client->cl_xprt->prot, vers);
-        host = nlmclnt_lookup_host(NFS_ADDR(inode), proto, vers);
        if (host == NULL)
                return -ENOLCK;
diff --git a/fs/lockd/host.c b/fs/lockd/host.c
index 38b0e8a1aec0..703fb038c813 100644
--- a/fs/lockd/host.c
+++ b/fs/lockd/host.c
@@ -26,7 +26,6 @@
 #define NLM_HOST_REBIND         (60 * HZ)
 #define NLM_HOST_EXPIRE         ((nrhosts > NLM_HOST_MAX)? 300 * HZ : 120 * HZ)
 #define NLM_HOST_COLLECT        ((nrhosts > NLM_HOST_MAX)? 120 * HZ :  60 * HZ)
-#define NLM_HOST_ADDR(sv)       (&(sv)->s_nlmclnt->cl_xprt->addr)
 static struct nlm_host *        nlm_hosts[NLM_HOST_NRHASH];
 static unsigned long            next_gc;
@@ -167,7 +166,6 @@ struct rpc_clnt *
 nlm_bind_host(struct nlm_host *host)
 {
        struct rpc_clnt *clnt;
-        struct rpc_xprt *xprt;
        dprintk("lockd: nlm_bind_host(%08x)\n",
                        (unsigned)ntohl(host->h_addr.sin_addr.s_addr));
@@ -179,7 +177,6 @@ nlm_bind_host(struct nlm_host *host)
         * RPC rebind is required
         */
        if ((clnt = host->h_rpcclnt) != NULL) {
-                xprt = clnt->cl_xprt;
                if (time_after_eq(jiffies, host->h_nextrebind)) {
                        rpc_force_rebind(clnt);
                        host->h_nextrebind = jiffies + NLM_HOST_REBIND;
@@ -187,31 +184,37 @@ nlm_bind_host(struct nlm_host *host)
                                        host->h_nextrebind - jiffies);
                }
        } else {
-                xprt = xprt_create_proto(host->h_proto, &host->h_addr, NULL);
+                unsigned long increment = nlmsvc_timeout * HZ;
-                if (IS_ERR(xprt))
+                struct rpc_timeout timeparms = {
-                        goto forgetit;
+                        .to_initval     = increment,
+                        .to_increment   = increment,
-                xprt_set_timeout(&xprt->timeout, 5, nlmsvc_timeout);
+                        .to_maxval      = increment * 6UL,
-                xprt->resvport = 1;     /* NLM requires a reserved port */
+                        .to_retries     = 5U,
+                };
-                /* Existing NLM servers accept AUTH_UNIX only */
+                struct rpc_create_args args = {
-                clnt = rpc_new_client(xprt, host->h_name, &nlm_program,
+                        .protocol       = host->h_proto,
-                                        host->h_version, RPC_AUTH_UNIX);
+                        .address        = (struct sockaddr *)&host->h_addr,
-                if (IS_ERR(clnt))
+                        .addrsize       = sizeof(host->h_addr),
-                        goto forgetit;
+                        .timeout        = &timeparms,
-                clnt->cl_autobind = 1;  /* turn on pmap queries */
+                        .servername     = host->h_name,
-                clnt->cl_softrtry = 1; /* All queries are soft */
+                        .program        = &nlm_program,
+                        .version        = host->h_version,
-                host->h_rpcclnt = clnt;
+                        .authflavor     = RPC_AUTH_UNIX,
+                        .flags          = (RPC_CLNT_CREATE_HARDRTRY |
+                                           RPC_CLNT_CREATE_AUTOBIND),
+                };
+                clnt = rpc_create(&args);
+                if (!IS_ERR(clnt))
+                        host->h_rpcclnt = clnt;
+                else {
+                        printk("lockd: couldn't create RPC handle for %s\n", host->h_name);
+                        clnt = NULL;
+                }
        }
        mutex_unlock(&host->h_mutex);
        return clnt;
-forgetit:
-        printk("lockd: couldn't create RPC handle for %s\n", host->h_name);
-        mutex_unlock(&host->h_mutex);
-        return NULL;
 }
 /*
diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c
index 3fc683f46b3e..5954dcb497e4 100644
--- a/fs/lockd/mon.c
+++ b/fs/lockd/mon.c
@@ -109,30 +109,23 @@ nsm_unmonitor(struct nlm_host *host)
 static struct rpc_clnt *
 nsm_create(void)
 {
-        struct rpc_xprt         *xprt;
+        struct sockaddr_in      sin = {
-        struct rpc_clnt         *clnt;
+                .sin_family     = AF_INET,
-        struct sockaddr_in      sin;
+                .sin_addr.s_addr = htonl(INADDR_LOOPBACK),
+                .sin_port       = 0,
-        sin.sin_family = AF_INET;
+        };
-        sin.sin_addr.s_addr = htonl(INADDR_LOOPBACK);
+        struct rpc_create_args args = {
-        sin.sin_port = 0;
+                .protocol       = IPPROTO_UDP,
+                .address        = (struct sockaddr *)&sin,
-        xprt = xprt_create_proto(IPPROTO_UDP, &sin, NULL);
+                .addrsize       = sizeof(sin),
-        if (IS_ERR(xprt))
+                .servername     = "localhost",
-                return (struct rpc_clnt *)xprt;
+                .program        = &nsm_program,
-        xprt->resvport = 1;     /* NSM requires a reserved port */
+                .version        = SM_VERSION,
+                .authflavor     = RPC_AUTH_NULL,
-        clnt = rpc_create_client(xprt, "localhost",
+                .flags          = (RPC_CLNT_CREATE_ONESHOT),
-                                &nsm_program, SM_VERSION,
+        };
-                                RPC_AUTH_NULL);
-        if (IS_ERR(clnt))
+        return rpc_create(&args);
-                goto out_err;
-        clnt->cl_softrtry = 1;
-        clnt->cl_oneshot  = 1;
-        return clnt;
-out_err:
-        return clnt;
 }
 /*
diff --git a/fs/namei.c b/fs/namei.c
index 432d6bc6fab0..6b591c01b09f 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -2370,7 +2370,8 @@ static int vfs_rename_dir(struct inode *old_dir, struct dentry *old_dentry,
                dput(new_dentry);
        }
        if (!error)
-                d_move(old_dentry,new_dentry);
+                if (!(old_dir->i_sb->s_type->fs_flags & FS_RENAME_DOES_D_MOVE))
+                        d_move(old_dentry,new_dentry);
        return error;
 }
@@ -2393,8 +2394,7 @@ static int vfs_rename_other(struct inode *old_dir, struct dentry *old_dentry,
        else
                error = old_dir->i_op->rename(old_dir, old_dentry, new_dir, new_dentry);
        if (!error) {
-                /* The following d_move() should become unconditional */
+                if (!(old_dir->i_sb->s_type->fs_flags & FS_RENAME_DOES_D_MOVE))
-                if (!(old_dir->i_sb->s_type->fs_flags & FS_ODD_RENAME))
                        d_move(old_dentry, new_dentry);
        }
        if (target)
diff --git a/fs/nfs/Makefile b/fs/nfs/Makefile
index 0b572a0c1967..f4580b44eef4 100644
--- a/fs/nfs/Makefile
+++ b/fs/nfs/Makefile
@@ -4,9 +4,9 @@
 obj-$(CONFIG_NFS_FS) += nfs.o
-nfs-y                   := dir.o file.o inode.o super.o nfs2xdr.o pagelist.o \
+nfs-y                   := client.o dir.o file.o getroot.o inode.o super.o nfs2xdr.o \
-                           proc.o read.o symlink.o unlink.o write.o \
+                           pagelist.o proc.o read.o symlink.o unlink.o \
-                           namespace.o
+                           write.o namespace.o
 nfs-$(CONFIG_ROOT_NFS)  += nfsroot.o mount_clnt.o      
 nfs-$(CONFIG_NFS_V3)    += nfs3proc.o nfs3xdr.o
 nfs-$(CONFIG_NFS_V3_ACL)        += nfs3acl.o
diff --git a/fs/nfs/callback.c b/fs/nfs/callback.c
index fe0a6b8ac149..a3ee11364db0 100644
--- a/fs/nfs/callback.c
+++ b/fs/nfs/callback.c
@@ -19,6 +19,7 @@
 #include "nfs4_fs.h"
 #include "callback.h"
+#include "internal.h"
 #define NFSDBG_FACILITY NFSDBG_CALLBACK
@@ -36,6 +37,21 @@ static struct svc_program nfs4_callback_program;
 unsigned int nfs_callback_set_tcpport;
 unsigned short nfs_callback_tcpport;
+static const int nfs_set_port_min = 0;
+static const int nfs_set_port_max = 65535;
+static int param_set_port(const char *val, struct kernel_param *kp)
+{
+        char *endp;
+        int num = simple_strtol(val, &endp, 0);
+        if (endp == val || *endp || num < nfs_set_port_min || num > nfs_set_port_max)
+                return -EINVAL;
+        *((int *)kp->arg) = num;
+        return 0;
+}
+module_param_call(callback_tcpport, param_set_port, param_get_int,
+                 &nfs_callback_set_tcpport, 0644);
 /*
 * This is the callback kernel thread.
@@ -134,10 +150,8 @@ out_err:
 /*
 * Kill the server process if it is not already up.
 */
-int nfs_callback_down(void)
+void nfs_callback_down(void)
 {
-        int ret = 0;
        lock_kernel();
        mutex_lock(&nfs_callback_mutex);
        nfs_callback_info.users--;
@@ -149,20 +163,19 @@ int nfs_callback_down(void)
        } while (wait_for_completion_timeout(&nfs_callback_info.stopped, 5*HZ) == 0);
        mutex_unlock(&nfs_callback_mutex);
        unlock_kernel();
-        return ret;
 }
 static int nfs_callback_authenticate(struct svc_rqst *rqstp)
 {
-        struct in_addr *addr = &rqstp->rq_addr.sin_addr;
+        struct sockaddr_in *addr = &rqstp->rq_addr;
-        struct nfs4_client *clp;
+        struct nfs_client *clp;
        /* Don't talk to strangers */
-        clp = nfs4_find_client(addr);
+        clp = nfs_find_client(addr, 4);
        if (clp == NULL)
                return SVC_DROP;
-        dprintk("%s: %u.%u.%u.%u NFSv4 callback!\n", __FUNCTION__, NIPQUAD(addr));
+        dprintk("%s: %u.%u.%u.%u NFSv4 callback!\n", __FUNCTION__, NIPQUAD(addr->sin_addr));
-        nfs4_put_client(clp);
+        nfs_put_client(clp);
        switch (rqstp->rq_authop->flavour) {
                case RPC_AUTH_NULL:
                        if (rqstp->rq_proc != CB_NULL)
diff --git a/fs/nfs/callback.h b/fs/nfs/callback.h
index b252e7fe53a5..5676163d26e8 100644
--- a/fs/nfs/callback.h
+++ b/fs/nfs/callback.h
@@ -62,8 +62,13 @@ struct cb_recallargs {
 extern unsigned nfs4_callback_getattr(struct cb_getattrargs *args, struct cb_getattrres *res);
 extern unsigned nfs4_callback_recall(struct cb_recallargs *args, void *dummy);
+#ifdef CONFIG_NFS_V4
 extern int nfs_callback_up(void);
-extern int nfs_callback_down(void);
+extern void nfs_callback_down(void);
+#else
+#define nfs_callback_up()       (0)
+#define nfs_callback_down()     do {} while(0)
+#endif
 extern unsigned int nfs_callback_set_tcpport;
 extern unsigned short nfs_callback_tcpport;
diff --git a/fs/nfs/callback_proc.c b/fs/nfs/callback_proc.c
index 7719483ecdfc..97cf8f71451f 100644
--- a/fs/nfs/callback_proc.c
+++ b/fs/nfs/callback_proc.c
@@ -10,19 +10,20 @@
 #include "nfs4_fs.h"
 #include "callback.h"
 #include "delegation.h"
+#include "internal.h"
 #define NFSDBG_FACILITY NFSDBG_CALLBACK
 
 unsigned nfs4_callback_getattr(struct cb_getattrargs *args, struct cb_getattrres *res)
 {
-        struct nfs4_client *clp;
+        struct nfs_client *clp;
        struct nfs_delegation *delegation;
        struct nfs_inode *nfsi;
        struct inode *inode;
        
        res->bitmap[0] = res->bitmap[1] = 0;
        res->status = htonl(NFS4ERR_BADHANDLE);
-        clp = nfs4_find_client(&args->addr->sin_addr);
+        clp = nfs_find_client(args->addr, 4);
        if (clp == NULL)
                goto out;
        inode = nfs_delegation_find_inode(clp, &args->fh);
@@ -48,7 +49,7 @@ out_iput:
        up_read(&nfsi->rwsem);
        iput(inode);
 out_putclient:
-        nfs4_put_client(clp);
+        nfs_put_client(clp);
 out:
        dprintk("%s: exit with status = %d\n", __FUNCTION__, ntohl(res->status));
        return res->status;
@@ -56,12 +57,12 @@ out:
 unsigned nfs4_callback_recall(struct cb_recallargs *args, void *dummy)
 {
-        struct nfs4_client *clp;
+        struct nfs_client *clp;
        struct inode *inode;
        unsigned res;
        
        res = htonl(NFS4ERR_BADHANDLE);
-        clp = nfs4_find_client(&args->addr->sin_addr);
+        clp = nfs_find_client(args->addr, 4);
        if (clp == NULL)
                goto out;
        inode = nfs_delegation_find_inode(clp, &args->fh);
@@ -80,7 +81,7 @@ unsigned nfs4_callback_recall(struct cb_recallargs *args, void *dummy)
        }
        iput(inode);
 out_putclient:
-        nfs4_put_client(clp);
+        nfs_put_client(clp);
 out:
        dprintk("%s: exit with status = %d\n", __FUNCTION__, ntohl(res));
        return res;
diff --git a/fs/nfs/client.c b/fs/nfs/client.c
new file mode 100644
index 000000000000..ec1938d4b814
--- /dev/null
+++ b/fs/nfs/client.c
@@ -0,0 +1,1448 @@
+/* client.c: NFS client sharing and management code
+ *
+ * Copyright (C) 2006 Red Hat, Inc. All Rights Reserved.
+ * Written by David Howells (dhowells@redhat.com)
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version
+ * 2 of the License, or (at your option) any later version.
+ */
+#include <linux/config.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/time.h>
+#include <linux/kernel.h>
+#include <linux/mm.h>
+#include <linux/string.h>
+#include <linux/stat.h>
+#include <linux/errno.h>
+#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>
+#include <linux/lockd/bind.h>
+#include <linux/smp_lock.h>
+#include <linux/seq_file.h>
+#include <linux/mount.h>
+#include <linux/nfs_idmap.h>
+#include <linux/vfs.h>
+#include <linux/inet.h>
+#include <linux/nfs_xdr.h>
+#include <asm/system.h>
+#include "nfs4_fs.h"
+#include "callback.h"
+#include "delegation.h"
+#include "iostat.h"
+#include "internal.h"
+#define NFSDBG_FACILITY         NFSDBG_CLIENT
+static DEFINE_SPINLOCK(nfs_client_lock);
+static LIST_HEAD(nfs_client_list);
+static LIST_HEAD(nfs_volume_list);
+static DECLARE_WAIT_QUEUE_HEAD(nfs_client_active_wq);
+/*
+ * RPC cruft for NFS
+ */
+static struct rpc_version *nfs_version[5] = {
+        [2]                     = &nfs_version2,
+#ifdef CONFIG_NFS_V3
+        [3]                     = &nfs_version3,
+#endif
+#ifdef CONFIG_NFS_V4
+        [4]                     = &nfs_version4,
+#endif
+};
+struct rpc_program nfs_program = {
+        .name                   = "nfs",
+        .number                 = NFS_PROGRAM,
+        .nrvers                 = ARRAY_SIZE(nfs_version),
+        .version                = nfs_version,
+        .stats                  = &nfs_rpcstat,
+        .pipe_dir_name          = "/nfs",
+};
+struct rpc_stat nfs_rpcstat = {
+        .program                = &nfs_program
+};
+#ifdef CONFIG_NFS_V3_ACL
+static struct rpc_stat          nfsacl_rpcstat = { &nfsacl_program };
+static struct rpc_version *     nfsacl_version[] = {
+        [3]                     = &nfsacl_version3,
+};
+struct rpc_program              nfsacl_program = {
+        .name                   = "nfsacl",
+        .number                 = NFS_ACL_PROGRAM,
+        .nrvers                 = ARRAY_SIZE(nfsacl_version),
+        .version                = nfsacl_version,
+        .stats                  = &nfsacl_rpcstat,
+};
+#endif  /* CONFIG_NFS_V3_ACL */
+/*
+ * Allocate a shared client record
+ *
+ * Since these are allocated/deallocated very rarely, we don't
+ * bother putting them in a slab cache...
+ */
+static struct nfs_client *nfs_alloc_client(const char *hostname,
+                                           const struct sockaddr_in *addr,
+                                           int nfsversion)
+{
+        struct nfs_client *clp;
+        int error;
+        if ((clp = kzalloc(sizeof(*clp), GFP_KERNEL)) == NULL)
+                goto error_0;
+        error = rpciod_up();
+        if (error < 0) {
+                dprintk("%s: couldn't start rpciod! Error = %d\n",
+                                __FUNCTION__, error);
+                goto error_1;
+        }
+        __set_bit(NFS_CS_RPCIOD, &clp->cl_res_state);
+        if (nfsversion == 4) {
+                if (nfs_callback_up() < 0)
+                        goto error_2;
+                __set_bit(NFS_CS_CALLBACK, &clp->cl_res_state);
+        }
+        atomic_set(&clp->cl_count, 1);
+        clp->cl_cons_state = NFS_CS_INITING;
+        clp->cl_nfsversion = nfsversion;
+        memcpy(&clp->cl_addr, addr, sizeof(clp->cl_addr));
+        if (hostname) {
+                clp->cl_hostname = kstrdup(hostname, GFP_KERNEL);
+                if (!clp->cl_hostname)
+                        goto error_3;
+        }
+        INIT_LIST_HEAD(&clp->cl_superblocks);
+        clp->cl_rpcclient = ERR_PTR(-EINVAL);
+#ifdef CONFIG_NFS_V4
+        init_rwsem(&clp->cl_sem);
+        INIT_LIST_HEAD(&clp->cl_delegations);
+        INIT_LIST_HEAD(&clp->cl_state_owners);
+        INIT_LIST_HEAD(&clp->cl_unused);
+        spin_lock_init(&clp->cl_lock);
+        INIT_WORK(&clp->cl_renewd, nfs4_renew_state, clp);
+        rpc_init_wait_queue(&clp->cl_rpcwaitq, "NFS client");
+        clp->cl_boot_time = CURRENT_TIME;
+        clp->cl_state = 1 << NFS4CLNT_LEASE_EXPIRED;
+#endif
+        return clp;
+error_3:
+        if (__test_and_clear_bit(NFS_CS_CALLBACK, &clp->cl_res_state))
+                nfs_callback_down();
+error_2:
+        rpciod_down();
+        __clear_bit(NFS_CS_RPCIOD, &clp->cl_res_state);
+error_1:
+        kfree(clp);
+error_0:
+        return NULL;
+}
+static void nfs4_shutdown_client(struct nfs_client *clp)
+{
+#ifdef CONFIG_NFS_V4
+        if (__test_and_clear_bit(NFS_CS_RENEWD, &clp->cl_res_state))
+                nfs4_kill_renewd(clp);
+        while (!list_empty(&clp->cl_unused)) {
+                struct nfs4_state_owner *sp;
+                sp = list_entry(clp->cl_unused.next,
+                                struct nfs4_state_owner,
+                                so_list);
+                list_del(&sp->so_list);
+                kfree(sp);
+        }
+        BUG_ON(!list_empty(&clp->cl_state_owners));
+        if (__test_and_clear_bit(NFS_CS_IDMAP, &clp->cl_res_state))
+                nfs_idmap_delete(clp);
+#endif
+}
+/*
+ * Destroy a shared client record
+ */
+static void nfs_free_client(struct nfs_client *clp)
+{
+        dprintk("--> nfs_free_client(%d)\n", clp->cl_nfsversion);
+        nfs4_shutdown_client(clp);
+        /* -EIO all pending I/O */
+        if (!IS_ERR(clp->cl_rpcclient))
+                rpc_shutdown_client(clp->cl_rpcclient);
+        if (__test_and_clear_bit(NFS_CS_CALLBACK, &clp->cl_res_state))
+                nfs_callback_down();
+        if (__test_and_clear_bit(NFS_CS_RPCIOD, &clp->cl_res_state))
+                rpciod_down();
+        kfree(clp->cl_hostname);
+        kfree(clp);
+        dprintk("<-- nfs_free_client()\n");
+}
+/*
+ * Release a reference to a shared client record
+ */
+void nfs_put_client(struct nfs_client *clp)
+{
+        if (!clp)
+                return;
+        dprintk("--> nfs_put_client({%d})\n", atomic_read(&clp->cl_count));
+        if (atomic_dec_and_lock(&clp->cl_count, &nfs_client_lock)) {
+                list_del(&clp->cl_share_link);
+                spin_unlock(&nfs_client_lock);
+                BUG_ON(!list_empty(&clp->cl_superblocks));
+                nfs_free_client(clp);
+        }
+}
+/*
+ * Find a client by address
+ * - caller must hold nfs_client_lock
+ */
+static struct nfs_client *__nfs_find_client(const struct sockaddr_in *addr, int nfsversion)
+{
+        struct nfs_client *clp;
+        list_for_each_entry(clp, &nfs_client_list, cl_share_link) {
+                /* Different NFS versions cannot share the same nfs_client */
+                if (clp->cl_nfsversion != nfsversion)
+                        continue;
+                if (memcmp(&clp->cl_addr.sin_addr, &addr->sin_addr,
+                           sizeof(clp->cl_addr.sin_addr)) != 0)
+                        continue;
+                if (clp->cl_addr.sin_port == addr->sin_port)
+                        goto found;
+        }
+        return NULL;
+found:
+        atomic_inc(&clp->cl_count);
+        return clp;
+}
+/*
+ * Find a client by IP address and protocol version
+ * - returns NULL if no such client
+ */
+struct nfs_client *nfs_find_client(const struct sockaddr_in *addr, int nfsversion)
+{
+        struct nfs_client *clp;
+        spin_lock(&nfs_client_lock);
+        clp = __nfs_find_client(addr, nfsversion);
+        spin_unlock(&nfs_client_lock);
+        BUG_ON(clp && clp->cl_cons_state == 0);
+        return clp;
+}
+/*
+ * Look up a client by IP address and protocol version
+ * - creates a new record if one doesn't yet exist
+ */
+static struct nfs_client *nfs_get_client(const char *hostname,
+                                         const struct sockaddr_in *addr,
+                                         int nfsversion)
+{
+        struct nfs_client *clp, *new = NULL;
+        int error;
+        dprintk("--> nfs_get_client(%s,"NIPQUAD_FMT":%d,%d)\n",
+                hostname ?: "", NIPQUAD(addr->sin_addr),
+                addr->sin_port, nfsversion);
+        /* see if the client already exists */
+        do {
+                spin_lock(&nfs_client_lock);
+                clp = __nfs_find_client(addr, nfsversion);
+                if (clp)
+                        goto found_client;
+                if (new)
+                        goto install_client;
+                spin_unlock(&nfs_client_lock);
+                new = nfs_alloc_client(hostname, addr, nfsversion);
+        } while (new);
+        return ERR_PTR(-ENOMEM);
+        /* install a new client and return with it unready */
+install_client:
+        clp = new;
+        list_add(&clp->cl_share_link, &nfs_client_list);
+        spin_unlock(&nfs_client_lock);
+        dprintk("--> nfs_get_client() = %p [new]\n", clp);
+        return clp;
+        /* found an existing client
+         * - make sure it's ready before returning
+         */
+found_client:
+        spin_unlock(&nfs_client_lock);
+        if (new)
+                nfs_free_client(new);
+        if (clp->cl_cons_state == NFS_CS_INITING) {
+                DECLARE_WAITQUEUE(myself, current);
+                add_wait_queue(&nfs_client_active_wq, &myself);
+                for (;;) {
+                        set_current_state(TASK_INTERRUPTIBLE);
+                        if (signal_pending(current) ||
+                            clp->cl_cons_state > NFS_CS_READY)
+                                break;
+                        schedule();
+                }
+                remove_wait_queue(&nfs_client_active_wq, &myself);
+                if (signal_pending(current)) {
+                        nfs_put_client(clp);
+                        return ERR_PTR(-ERESTARTSYS);
+                }
+        }
+        if (clp->cl_cons_state < NFS_CS_READY) {
+                error = clp->cl_cons_state;
+                nfs_put_client(clp);
+                return ERR_PTR(error);
+        }
+        BUG_ON(clp->cl_cons_state != NFS_CS_READY);
+        dprintk("--> nfs_get_client() = %p [share]\n", clp);
+        return clp;
+}
+/*
+ * Mark a server as ready or failed
+ */
+static void nfs_mark_client_ready(struct nfs_client *clp, int state)
+{
+        clp->cl_cons_state = state;
+        wake_up_all(&nfs_client_active_wq);
+}
+/*
+ * Initialise the timeout values for a connection
+ */
+static void nfs_init_timeout_values(struct rpc_timeout *to, int proto,
+                                    unsigned int timeo, unsigned int retrans)
+{
+        to->to_initval = timeo * HZ / 10;
+        to->to_retries = retrans;
+        if (!to->to_retries)
+                to->to_retries = 2;
+        switch (proto) {
+        case IPPROTO_TCP:
+                if (!to->to_initval)
+                        to->to_initval = 60 * HZ;
+                if (to->to_initval > NFS_MAX_TCP_TIMEOUT)
+                        to->to_initval = NFS_MAX_TCP_TIMEOUT;
+                to->to_increment = to->to_initval;
+                to->to_maxval = to->to_initval + (to->to_increment * to->to_retries);
+                to->to_exponential = 0;
+                break;
+        case IPPROTO_UDP:
+        default:
+                if (!to->to_initval)
+                        to->to_initval = 11 * HZ / 10;
+                if (to->to_initval > NFS_MAX_UDP_TIMEOUT)
+                        to->to_initval = NFS_MAX_UDP_TIMEOUT;
+                to->to_maxval = NFS_MAX_UDP_TIMEOUT;
+                to->to_exponential = 1;
+                break;
+        }
+}
+/*
+ * Create an RPC client handle
+ */
+static int nfs_create_rpc_client(struct nfs_client *clp, int proto,
+                                                unsigned int timeo,
+                                                unsigned int retrans,
+                                                rpc_authflavor_t flavor)
+{
+        struct rpc_timeout      timeparms;
+        struct rpc_clnt         *clnt = NULL;
+        struct rpc_create_args args = {
+                .protocol       = proto,
+                .address        = (struct sockaddr *)&clp->cl_addr,
+                .addrsize       = sizeof(clp->cl_addr),
+                .timeout        = &timeparms,
+                .servername     = clp->cl_hostname,
+                .program        = &nfs_program,
+                .version        = clp->rpc_ops->version,
+                .authflavor     = flavor,
+        };
+        if (!IS_ERR(clp->cl_rpcclient))
+                return 0;
+        nfs_init_timeout_values(&timeparms, proto, timeo, retrans);
+        clp->retrans_timeo = timeparms.to_initval;
+        clp->retrans_count = timeparms.to_retries;
+        clnt = rpc_create(&args);
+        if (IS_ERR(clnt)) {
+                dprintk("%s: cannot create RPC client. Error = %ld\n",
+                                __FUNCTION__, PTR_ERR(clnt));
+                return PTR_ERR(clnt);
+        }
+        clp->cl_rpcclient = clnt;
+        return 0;
+}
+/*
+ * Version 2 or 3 client destruction
+ */
+static void nfs_destroy_server(struct nfs_server *server)
+{
+        if (!IS_ERR(server->client_acl))
+                rpc_shutdown_client(server->client_acl);
+        if (!(server->flags & NFS_MOUNT_NONLM))
+                lockd_down();   /* release rpc.lockd */
+}
+/*
+ * Version 2 or 3 lockd setup
+ */
+static int nfs_start_lockd(struct nfs_server *server)
+{
+        int error = 0;
+        if (server->nfs_client->cl_nfsversion > 3)
+                goto out;
+        if (server->flags & NFS_MOUNT_NONLM)
+                goto out;
+        error = lockd_up();
+        if (error < 0)
+                server->flags |= NFS_MOUNT_NONLM;
+        else
+                server->destroy = nfs_destroy_server;
+out:
+        return error;
+}
+/*
+ * Initialise an NFSv3 ACL client connection
+ */
+#ifdef CONFIG_NFS_V3_ACL
+static void nfs_init_server_aclclient(struct nfs_server *server)
+{
+        if (server->nfs_client->cl_nfsversion != 3)
+                goto out_noacl;
+        if (server->flags & NFS_MOUNT_NOACL)
+                goto out_noacl;
+        server->client_acl = rpc_bind_new_program(server->client, &nfsacl_program, 3);
+        if (IS_ERR(server->client_acl))
+                goto out_noacl;
+        /* No errors! Assume that Sun nfsacls are supported */
+        server->caps |= NFS_CAP_ACLS;
+        return;
+out_noacl:
+        server->caps &= ~NFS_CAP_ACLS;
+}
+#else
+static inline void nfs_init_server_aclclient(struct nfs_server *server)
+{
+        server->flags &= ~NFS_MOUNT_NOACL;
+        server->caps &= ~NFS_CAP_ACLS;
+}
+#endif
+/*
+ * Create a general RPC client
+ */
+static int nfs_init_server_rpcclient(struct nfs_server *server, rpc_authflavor_t pseudoflavour)
+{
+        struct nfs_client *clp = server->nfs_client;
+        server->client = rpc_clone_client(clp->cl_rpcclient);
+        if (IS_ERR(server->client)) {
+                dprintk("%s: couldn't create rpc_client!\n", __FUNCTION__);
+                return PTR_ERR(server->client);
+        }
+        if (pseudoflavour != clp->cl_rpcclient->cl_auth->au_flavor) {
+                struct rpc_auth *auth;
+                auth = rpcauth_create(pseudoflavour, server->client);
+                if (IS_ERR(auth)) {
+                        dprintk("%s: couldn't create credcache!\n", __FUNCTION__);
+                        return PTR_ERR(auth);
+                }
+        }
+        server->client->cl_softrtry = 0;
+        if (server->flags & NFS_MOUNT_SOFT)
+                server->client->cl_softrtry = 1;
+        server->client->cl_intr = 0;
+        if (server->flags & NFS4_MOUNT_INTR)
+                server->client->cl_intr = 1;
+        return 0;
+}
+/*
+ * Initialise an NFS2 or NFS3 client
+ */
+static int nfs_init_client(struct nfs_client *clp, const struct nfs_mount_data *data)
+{
+        int proto = (data->flags & NFS_MOUNT_TCP) ? IPPROTO_TCP : IPPROTO_UDP;
+        int error;
+        if (clp->cl_cons_state == NFS_CS_READY) {
+                /* the client is already initialised */
+                dprintk("<-- nfs_init_client() = 0 [already %p]\n", clp);
+                return 0;
+        }
+        /* Check NFS protocol revision and initialize RPC op vector */
+        clp->rpc_ops = &nfs_v2_clientops;
+#ifdef CONFIG_NFS_V3
+        if (clp->cl_nfsversion == 3)
+                clp->rpc_ops = &nfs_v3_clientops;
+#endif
+        /*
+         * Create a client RPC handle for doing FSSTAT with UNIX auth only
+         * - RFC 2623, sec 2.3.2
+         */
+        error = nfs_create_rpc_client(clp, proto, data->timeo, data->retrans,
+                        RPC_AUTH_UNIX);
+        if (error < 0)
+                goto error;
+        nfs_mark_client_ready(clp, NFS_CS_READY);
+        return 0;
+error:
+        nfs_mark_client_ready(clp, error);
+        dprintk("<-- nfs_init_client() = xerror %d\n", error);
+        return error;
+}
+/*
+ * Create a version 2 or 3 client
+ */
+static int nfs_init_server(struct nfs_server *server, const struct nfs_mount_data *data)
+{
+        struct nfs_client *clp;
+        int error, nfsvers = 2;
+        dprintk("--> nfs_init_server()\n");
+#ifdef CONFIG_NFS_V3
+        if (data->flags & NFS_MOUNT_VER3)
+                nfsvers = 3;
+#endif
+        /* Allocate or find a client reference we can use */
+        clp = nfs_get_client(data->hostname, &data->addr, nfsvers);
+        if (IS_ERR(clp)) {
+                dprintk("<-- nfs_init_server() = error %ld\n", PTR_ERR(clp));
+                return PTR_ERR(clp);
+        }
+        error = nfs_init_client(clp, data);
+        if (error < 0)
+                goto error;
+        server->nfs_client = clp;
+        /* Initialise the client representation from the mount data */
+        server->flags = data->flags & NFS_MOUNT_FLAGMASK;
+        if (data->rsize)
+                server->rsize = nfs_block_size(data->rsize, NULL);
+        if (data->wsize)
+                server->wsize = nfs_block_size(data->wsize, NULL);
+        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 */
+        error = nfs_start_lockd(server);
+        if (error < 0)
+                goto error;
+        error = nfs_init_server_rpcclient(server, data->pseudoflavor);
+        if (error < 0)
+                goto error;
+        server->namelen  = data->namlen;
+        /* Create a client RPC handle for the NFSv3 ACL management interface */
+        nfs_init_server_aclclient(server);
+        if (clp->cl_nfsversion == 3) {
+                if (server->namelen == 0 || server->namelen > NFS3_MAXNAMLEN)
+                        server->namelen = NFS3_MAXNAMLEN;
+                server->caps |= NFS_CAP_READDIRPLUS;
+        } else {
+                if (server->namelen == 0 || server->namelen > NFS2_MAXNAMLEN)
+                        server->namelen = NFS2_MAXNAMLEN;
+        }
+        dprintk("<-- nfs_init_server() = 0 [new %p]\n", clp);
+        return 0;
+error:
+        server->nfs_client = NULL;
+        nfs_put_client(clp);
+        dprintk("<-- nfs_init_server() = xerror %d\n", error);
+        return error;
+}
+/*
+ * Load up the server record from information gained in an fsinfo record
+ */
+static void nfs_server_set_fsinfo(struct nfs_server *server, struct nfs_fsinfo *fsinfo)
+{
+        unsigned long max_rpc_payload;
+        /* 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);
+        max_rpc_payload = nfs_block_size(rpc_max_payload(server->client), NULL);
+        if (server->rsize > max_rpc_payload)
+                server->rsize = max_rpc_payload;
+        if (server->rsize > NFS_MAX_FILE_IO_SIZE)
+                server->rsize = NFS_MAX_FILE_IO_SIZE;
+        server->rpages = (server->rsize + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
+        server->backing_dev_info.ra_pages = server->rpages * NFS_MAX_READAHEAD;
+        if (server->wsize > max_rpc_payload)
+                server->wsize = max_rpc_payload;
+        if (server->wsize > NFS_MAX_FILE_IO_SIZE)
+                server->wsize = NFS_MAX_FILE_IO_SIZE;
+        server->wpages = (server->wsize + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
+        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;
+        }
+        server->maxfilesize = fsinfo->maxfilesize;
+        /* We're airborne Set socket buffersize */
+        rpc_setbufsize(server->client, server->wsize + 100, server->rsize + 100);
+}
+/*
+ * Probe filesystem information, including the FSID on v2/v3
+ */
+static int nfs_probe_fsinfo(struct nfs_server *server, struct nfs_fh *mntfh, struct nfs_fattr *fattr)
+{
+        struct nfs_fsinfo fsinfo;
+        struct nfs_client *clp = server->nfs_client;
+        int error;
+        dprintk("--> nfs_probe_fsinfo()\n");
+        if (clp->rpc_ops->set_capabilities != NULL) {
+                error = clp->rpc_ops->set_capabilities(server, mntfh);
+                if (error < 0)
+                        goto out_error;
+        }
+        fsinfo.fattr = fattr;
+        nfs_fattr_init(fattr);
+        error = clp->rpc_ops->fsinfo(server, mntfh, &fsinfo);
+        if (error < 0)
+                goto out_error;
+        nfs_server_set_fsinfo(server, &fsinfo);
+        /* Get some general file system info */
+        if (server->namelen == 0) {
+                struct nfs_pathconf pathinfo;
+                pathinfo.fattr = fattr;
+                nfs_fattr_init(fattr);
+                if (clp->rpc_ops->pathconf(server, mntfh, &pathinfo) >= 0)
+                        server->namelen = pathinfo.max_namelen;
+        }
+        dprintk("<-- nfs_probe_fsinfo() = 0\n");
+        return 0;
+out_error:
+        dprintk("nfs_probe_fsinfo: error = %d\n", -error);
+        return error;
+}
+/*
+ * Copy useful information when duplicating a server record
+ */
+static void nfs_server_copy_userdata(struct nfs_server *target, struct nfs_server *source)
+{
+        target->flags = source->flags;
+        target->acregmin = source->acregmin;
+        target->acregmax = source->acregmax;
+        target->acdirmin = source->acdirmin;
+        target->acdirmax = source->acdirmax;
+        target->caps = source->caps;
+}
+/*
+ * Allocate and initialise a server record
+ */
+static struct nfs_server *nfs_alloc_server(void)
+{
+        struct nfs_server *server;
+        server = kzalloc(sizeof(struct nfs_server), GFP_KERNEL);
+        if (!server)
+                return NULL;
+        server->client = server->client_acl = ERR_PTR(-EINVAL);
+        /* Zero out the NFS state stuff */
+        INIT_LIST_HEAD(&server->client_link);
+        INIT_LIST_HEAD(&server->master_link);
+        server->io_stats = nfs_alloc_iostats();
+        if (!server->io_stats) {
+                kfree(server);
+                return NULL;
+        }
+        return server;
+}
+/*
+ * Free up a server record
+ */
+void nfs_free_server(struct nfs_server *server)
+{
+        dprintk("--> nfs_free_server()\n");
+        spin_lock(&nfs_client_lock);
+        list_del(&server->client_link);
+        list_del(&server->master_link);
+        spin_unlock(&nfs_client_lock);
+        if (server->destroy != NULL)
+                server->destroy(server);
+        if (!IS_ERR(server->client))
+                rpc_shutdown_client(server->client);
+        nfs_put_client(server->nfs_client);
+        nfs_free_iostats(server->io_stats);
+        kfree(server);
+        nfs_release_automount_timer();
+        dprintk("<-- nfs_free_server()\n");
+}
+/*
+ * Create a version 2 or 3 volume record
+ * - keyed on server and FSID
+ */
+struct nfs_server *nfs_create_server(const struct nfs_mount_data *data,
+                                     struct nfs_fh *mntfh)
+{
+        struct nfs_server *server;
+        struct nfs_fattr fattr;
+        int error;
+        server = nfs_alloc_server();
+        if (!server)
+                return ERR_PTR(-ENOMEM);
+        /* Get a client representation */
+        error = nfs_init_server(server, data);
+        if (error < 0)
+                goto error;
+        BUG_ON(!server->nfs_client);
+        BUG_ON(!server->nfs_client->rpc_ops);
+        BUG_ON(!server->nfs_client->rpc_ops->file_inode_ops);
+        /* Probe the root fh to retrieve its FSID */
+        error = nfs_probe_fsinfo(server, mntfh, &fattr);
+        if (error < 0)
+                goto error;
+        if (!(fattr.valid & NFS_ATTR_FATTR)) {
+                error = server->nfs_client->rpc_ops->getattr(server, mntfh, &fattr);
+                if (error < 0) {
+                        dprintk("nfs_create_server: getattr error = %d\n", -error);
+                        goto error;
+                }
+        }
+        memcpy(&server->fsid, &fattr.fsid, sizeof(server->fsid));
+        dprintk("Server FSID: %llx:%llx\n",
+                (unsigned long long) server->fsid.major,
+                (unsigned long long) server->fsid.minor);
+        BUG_ON(!server->nfs_client);
+        BUG_ON(!server->nfs_client->rpc_ops);
+        BUG_ON(!server->nfs_client->rpc_ops->file_inode_ops);
+        spin_lock(&nfs_client_lock);
+        list_add_tail(&server->client_link, &server->nfs_client->cl_superblocks);
+        list_add_tail(&server->master_link, &nfs_volume_list);
+        spin_unlock(&nfs_client_lock);
+        server->mount_time = jiffies;
+        return server;
+error:
+        nfs_free_server(server);
+        return ERR_PTR(error);
+}
+#ifdef CONFIG_NFS_V4
+/*
+ * Initialise an NFS4 client record
+ */
+static int nfs4_init_client(struct nfs_client *clp,
+                int proto, int timeo, int retrans,
+                rpc_authflavor_t authflavour)
+{
+        int error;
+        if (clp->cl_cons_state == NFS_CS_READY) {
+                /* the client is initialised already */
+                dprintk("<-- nfs4_init_client() = 0 [already %p]\n", clp);
+                return 0;
+        }
+        /* Check NFS protocol revision and initialize RPC op vector */
+        clp->rpc_ops = &nfs_v4_clientops;
+        error = nfs_create_rpc_client(clp, proto, timeo, retrans, authflavour);
+        if (error < 0)
+                goto error;
+        error = nfs_idmap_new(clp);
+        if (error < 0) {
+                dprintk("%s: failed to create idmapper. Error = %d\n",
+                        __FUNCTION__, error);
+                goto error;
+        }
+        __set_bit(NFS_CS_IDMAP, &clp->cl_res_state);
+        nfs_mark_client_ready(clp, NFS_CS_READY);
+        return 0;
+error:
+        nfs_mark_client_ready(clp, error);
+        dprintk("<-- nfs4_init_client() = xerror %d\n", error);
+        return error;
+}
+/*
+ * Set up an NFS4 client
+ */
+static int nfs4_set_client(struct nfs_server *server,
+                const char *hostname, const struct sockaddr_in *addr,
+                rpc_authflavor_t authflavour,
+                int proto, int timeo, int retrans)
+{
+        struct nfs_client *clp;
+        int error;
+        dprintk("--> nfs4_set_client()\n");
+        /* Allocate or find a client reference we can use */
+        clp = nfs_get_client(hostname, addr, 4);
+        if (IS_ERR(clp)) {
+                error = PTR_ERR(clp);
+                goto error;
+        }
+        error = nfs4_init_client(clp, proto, timeo, retrans, authflavour);
+        if (error < 0)
+                goto error_put;
+        server->nfs_client = clp;
+        dprintk("<-- nfs4_set_client() = 0 [new %p]\n", clp);
+        return 0;
+error_put:
+        nfs_put_client(clp);
+error:
+        dprintk("<-- nfs4_set_client() = xerror %d\n", error);
+        return error;
+}
+/*
+ * Create a version 4 volume record
+ */
+static int nfs4_init_server(struct nfs_server *server,
+                const struct nfs4_mount_data *data, rpc_authflavor_t authflavour)
+{
+        int error;
+        dprintk("--> nfs4_init_server()\n");
+        /* Initialise the client representation from the mount data */
+        server->flags = data->flags & NFS_MOUNT_FLAGMASK;
+        server->caps |= NFS_CAP_ATOMIC_OPEN;
+        if (data->rsize)
+                server->rsize = nfs_block_size(data->rsize, NULL);
+        if (data->wsize)
+                server->wsize = nfs_block_size(data->wsize, NULL);
+        server->acregmin = data->acregmin * HZ;
+        server->acregmax = data->acregmax * HZ;
+        server->acdirmin = data->acdirmin * HZ;
+        server->acdirmax = data->acdirmax * HZ;
+        error = nfs_init_server_rpcclient(server, authflavour);
+        /* Done */
+        dprintk("<-- nfs4_init_server() = %d\n", error);
+        return error;
+}
+/*
+ * Create a version 4 volume record
+ * - keyed on server and FSID
+ */
+struct nfs_server *nfs4_create_server(const struct nfs4_mount_data *data,
+                                      const char *hostname,
+                                      const struct sockaddr_in *addr,
+                                      const char *mntpath,
+                                      const char *ip_addr,
+                                      rpc_authflavor_t authflavour,
+                                      struct nfs_fh *mntfh)
+{
+        struct nfs_fattr fattr;
+        struct nfs_server *server;
+        int error;
+        dprintk("--> nfs4_create_server()\n");
+        server = nfs_alloc_server();
+        if (!server)
+                return ERR_PTR(-ENOMEM);
+        /* Get a client record */
+        error = nfs4_set_client(server, hostname, addr, authflavour,
+                        data->proto, data->timeo, data->retrans);
+        if (error < 0)
+                goto error;
+        /* set up the general RPC client */
+        error = nfs4_init_server(server, data, authflavour);
+        if (error < 0)
+                goto error;
+        BUG_ON(!server->nfs_client);
+        BUG_ON(!server->nfs_client->rpc_ops);
+        BUG_ON(!server->nfs_client->rpc_ops->file_inode_ops);
+        /* Probe the root fh to retrieve its FSID */
+        error = nfs4_path_walk(server, mntfh, mntpath);
+        if (error < 0)
+                goto error;
+        dprintk("Server FSID: %llx:%llx\n",
+                (unsigned long long) server->fsid.major,
+                (unsigned long long) server->fsid.minor);
+        dprintk("Mount FH: %d\n", mntfh->size);
+        error = nfs_probe_fsinfo(server, mntfh, &fattr);
+        if (error < 0)
+                goto error;
+        BUG_ON(!server->nfs_client);
+        BUG_ON(!server->nfs_client->rpc_ops);
+        BUG_ON(!server->nfs_client->rpc_ops->file_inode_ops);
+        spin_lock(&nfs_client_lock);
+        list_add_tail(&server->client_link, &server->nfs_client->cl_superblocks);
+        list_add_tail(&server->master_link, &nfs_volume_list);
+        spin_unlock(&nfs_client_lock);
+        server->mount_time = jiffies;
+        dprintk("<-- nfs4_create_server() = %p\n", server);
+        return server;
+error:
+        nfs_free_server(server);
+        dprintk("<-- nfs4_create_server() = error %d\n", error);
+        return ERR_PTR(error);
+}
+/*
+ * Create an NFS4 referral server record
+ */
+struct nfs_server *nfs4_create_referral_server(struct nfs_clone_mount *data,
+                                               struct nfs_fh *fh)
+{
+        struct nfs_client *parent_client;
+        struct nfs_server *server, *parent_server;
+        struct nfs_fattr fattr;
+        int error;
+        dprintk("--> nfs4_create_referral_server()\n");
+        server = nfs_alloc_server();
+        if (!server)
+                return ERR_PTR(-ENOMEM);
+        parent_server = NFS_SB(data->sb);
+        parent_client = parent_server->nfs_client;
+        /* Get a client representation.
+         * Note: NFSv4 always uses TCP, */
+        error = nfs4_set_client(server, data->hostname, data->addr,
+                        data->authflavor,
+                        parent_server->client->cl_xprt->prot,
+                        parent_client->retrans_timeo,
+                        parent_client->retrans_count);
+        if (error < 0)
+                goto error;
+        /* Initialise the client representation from the parent server */
+        nfs_server_copy_userdata(server, parent_server);
+        server->caps |= NFS_CAP_ATOMIC_OPEN;
+        error = nfs_init_server_rpcclient(server, data->authflavor);
+        if (error < 0)
+                goto error;
+        BUG_ON(!server->nfs_client);
+        BUG_ON(!server->nfs_client->rpc_ops);
+        BUG_ON(!server->nfs_client->rpc_ops->file_inode_ops);
+        /* probe the filesystem info for this server filesystem */
+        error = nfs_probe_fsinfo(server, fh, &fattr);
+        if (error < 0)
+                goto error;
+        dprintk("Referral FSID: %llx:%llx\n",
+                (unsigned long long) server->fsid.major,
+                (unsigned long long) server->fsid.minor);
+        spin_lock(&nfs_client_lock);
+        list_add_tail(&server->client_link, &server->nfs_client->cl_superblocks);
+        list_add_tail(&server->master_link, &nfs_volume_list);
+        spin_unlock(&nfs_client_lock);
+        server->mount_time = jiffies;
+        dprintk("<-- nfs_create_referral_server() = %p\n", server);
+        return server;
+error:
+        nfs_free_server(server);
+        dprintk("<-- nfs4_create_referral_server() = error %d\n", error);
+        return ERR_PTR(error);
+}
+#endif /* CONFIG_NFS_V4 */
+/*
+ * Clone an NFS2, NFS3 or NFS4 server record
+ */
+struct nfs_server *nfs_clone_server(struct nfs_server *source,
+                                    struct nfs_fh *fh,
+                                    struct nfs_fattr *fattr)
+{
+        struct nfs_server *server;
+        struct nfs_fattr fattr_fsinfo;
+        int error;
+        dprintk("--> nfs_clone_server(,%llx:%llx,)\n",
+                (unsigned long long) fattr->fsid.major,
+                (unsigned long long) fattr->fsid.minor);
+        server = nfs_alloc_server();
+        if (!server)
+                return ERR_PTR(-ENOMEM);
+        /* Copy data from the source */
+        server->nfs_client = source->nfs_client;
+        atomic_inc(&server->nfs_client->cl_count);
+        nfs_server_copy_userdata(server, source);
+        server->fsid = fattr->fsid;
+        error = nfs_init_server_rpcclient(server, source->client->cl_auth->au_flavor);
+        if (error < 0)
+                goto out_free_server;
+        if (!IS_ERR(source->client_acl))
+                nfs_init_server_aclclient(server);
+        /* probe the filesystem info for this server filesystem */
+        error = nfs_probe_fsinfo(server, fh, &fattr_fsinfo);
+        if (error < 0)
+                goto out_free_server;
+        dprintk("Cloned FSID: %llx:%llx\n",
+                (unsigned long long) server->fsid.major,
+                (unsigned long long) server->fsid.minor);
+        error = nfs_start_lockd(server);
+        if (error < 0)
+                goto out_free_server;
+        spin_lock(&nfs_client_lock);
+        list_add_tail(&server->client_link, &server->nfs_client->cl_superblocks);
+        list_add_tail(&server->master_link, &nfs_volume_list);
+        spin_unlock(&nfs_client_lock);
+        server->mount_time = jiffies;
+        dprintk("<-- nfs_clone_server() = %p\n", server);
+        return server;
+out_free_server:
+        nfs_free_server(server);
+        dprintk("<-- nfs_clone_server() = error %d\n", error);
+        return ERR_PTR(error);
+}
+#ifdef CONFIG_PROC_FS
+static struct proc_dir_entry *proc_fs_nfs;
+static int nfs_server_list_open(struct inode *inode, struct file *file);
+static void *nfs_server_list_start(struct seq_file *p, loff_t *pos);
+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);
+static int nfs_server_list_show(struct seq_file *m, void *v);
+static struct seq_operations nfs_server_list_ops = {
+        .start  = nfs_server_list_start,
+        .next   = nfs_server_list_next,
+        .stop   = nfs_server_list_stop,
+        .show   = nfs_server_list_show,
+};
+static struct file_operations nfs_server_list_fops = {
+        .open           = nfs_server_list_open,
+        .read           = seq_read,
+        .llseek         = seq_lseek,
+        .release        = seq_release,
+};
+static int nfs_volume_list_open(struct inode *inode, struct file *file);
+static void *nfs_volume_list_start(struct seq_file *p, loff_t *pos);
+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);
+static int nfs_volume_list_show(struct seq_file *m, void *v);
+static struct seq_operations nfs_volume_list_ops = {
+        .start  = nfs_volume_list_start,
+        .next   = nfs_volume_list_next,
+        .stop   = nfs_volume_list_stop,
+        .show   = nfs_volume_list_show,
+};
+static struct file_operations nfs_volume_list_fops = {
+        .open           = nfs_volume_list_open,
+        .read           = seq_read,
+        .llseek         = seq_lseek,
+        .release        = seq_release,
+};
+/*
+ * open "/proc/fs/nfsfs/servers" which provides a summary of servers with which
+ * we're dealing
+ */
+static int nfs_server_list_open(struct inode *inode, struct file *file)
+{
+        struct seq_file *m;
+        int ret;
+        ret = seq_open(file, &nfs_server_list_ops);
+        if (ret < 0)
+                return ret;
+        m = file->private_data;
+        m->private = PDE(inode)->data;
+        return 0;
+}
+/*
+ * set up the iterator to start reading from the server list and return the first item
+ */
+static void *nfs_server_list_start(struct seq_file *m, loff_t *_pos)
+{
+        struct list_head *_p;
+        loff_t pos = *_pos;
+        /* lock the list against modification */
+        spin_lock(&nfs_client_lock);
+        /* allow for the header line */
+        if (!pos)
+                return SEQ_START_TOKEN;
+        pos--;
+        /* find the n'th element in the list */
+        list_for_each(_p, &nfs_client_list)
+                if (!pos--)
+                        break;
+        return _p != &nfs_client_list ? _p : NULL;
+}
+/*
+ * move to next server
+ */
+static void *nfs_server_list_next(struct seq_file *p, void *v, loff_t *pos)
+{
+        struct list_head *_p;
+        (*pos)++;
+        _p = v;
+        _p = (v == SEQ_START_TOKEN) ? nfs_client_list.next : _p->next;
+        return _p != &nfs_client_list ? _p : NULL;
+}
+/*
+ * clean up after reading from the transports list
+ */
+static void nfs_server_list_stop(struct seq_file *p, void *v)
+{
+        spin_unlock(&nfs_client_lock);
+}
+/*
+ * display a header line followed by a load of call lines
+ */
+static int nfs_server_list_show(struct seq_file *m, void *v)
+{
+        struct nfs_client *clp;
+        /* display header on line 1 */
+        if (v == SEQ_START_TOKEN) {
+                seq_puts(m, "NV SERVER   PORT USE HOSTNAME\n");
+                return 0;
+        }
+        /* display one transport per line on subsequent lines */
+        clp = list_entry(v, struct nfs_client, cl_share_link);
+        seq_printf(m, "v%d %02x%02x%02x%02x %4hx %3d %s\n",
+                   clp->cl_nfsversion,
+                   NIPQUAD(clp->cl_addr.sin_addr),
+                   ntohs(clp->cl_addr.sin_port),
+                   atomic_read(&clp->cl_count),
+                   clp->cl_hostname);
+        return 0;
+}
+/*
+ * open "/proc/fs/nfsfs/volumes" which provides a summary of extant volumes
+ */
+static int nfs_volume_list_open(struct inode *inode, struct file *file)
+{
+        struct seq_file *m;
+        int ret;
+        ret = seq_open(file, &nfs_volume_list_ops);
+        if (ret < 0)
+                return ret;
+        m = file->private_data;
+        m->private = PDE(inode)->data;
+        return 0;
+}
+/*
+ * set up the iterator to start reading from the volume list and return the first item
+ */
+static void *nfs_volume_list_start(struct seq_file *m, loff_t *_pos)
+{
+        struct list_head *_p;
+        loff_t pos = *_pos;
+        /* lock the list against modification */
+        spin_lock(&nfs_client_lock);
+        /* allow for the header line */
+        if (!pos)
+                return SEQ_START_TOKEN;
+        pos--;
+        /* find the n'th element in the list */
+        list_for_each(_p, &nfs_volume_list)
+                if (!pos--)
+                        break;
+        return _p != &nfs_volume_list ? _p : NULL;
+}
+/*
+ * move to next volume
+ */
+static void *nfs_volume_list_next(struct seq_file *p, void *v, loff_t *pos)
+{
+        struct list_head *_p;
+        (*pos)++;
+        _p = v;
+        _p = (v == SEQ_START_TOKEN) ? nfs_volume_list.next : _p->next;
+        return _p != &nfs_volume_list ? _p : NULL;
+}
+/*
+ * clean up after reading from the transports list
+ */
+static void nfs_volume_list_stop(struct seq_file *p, void *v)
+{
+        spin_unlock(&nfs_client_lock);
+}
+/*
+ * display a header line followed by a load of call lines
+ */
+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];
+        /* display header on line 1 */
+        if (v == SEQ_START_TOKEN) {
+                seq_puts(m, "NV SERVER   PORT DEV     FSID\n");
+                return 0;
+        }
+        /* display one transport per line on subsequent lines */
+        server = list_entry(v, struct nfs_server, master_link);
+        clp = server->nfs_client;
+        snprintf(dev, 8, "%u:%u",
+                 MAJOR(server->s_dev), MINOR(server->s_dev));
+        snprintf(fsid, 17, "%llx:%llx",
+                 (unsigned long long) server->fsid.major,
+                 (unsigned long long) server->fsid.minor);
+        seq_printf(m, "v%d %02x%02x%02x%02x %4hx %-7s %-17s\n",
+                   clp->cl_nfsversion,
+                   NIPQUAD(clp->cl_addr.sin_addr),
+                   ntohs(clp->cl_addr.sin_port),
+                   dev,
+                   fsid);
+        return 0;
+}
+/*
+ * initialise the /proc/fs/nfsfs/ directory
+ */
+int __init nfs_fs_proc_init(void)
+{
+        struct proc_dir_entry *p;
+        proc_fs_nfs = proc_mkdir("nfsfs", proc_root_fs);
+        if (!proc_fs_nfs)
+                goto error_0;
+        proc_fs_nfs->owner = THIS_MODULE;
+        /* a file of servers with which we're dealing */
+        p = create_proc_entry("servers", S_IFREG|S_IRUGO, proc_fs_nfs);
+        if (!p)
+                goto error_1;
+        p->proc_fops = &nfs_server_list_fops;
+        p->owner = THIS_MODULE;
+        /* a file of volumes that we have mounted */
+        p = create_proc_entry("volumes", S_IFREG|S_IRUGO, proc_fs_nfs);
+        if (!p)
+                goto error_2;
+        p->proc_fops = &nfs_volume_list_fops;
+        p->owner = THIS_MODULE;
+        return 0;
+error_2:
+        remove_proc_entry("servers", proc_fs_nfs);
+error_1:
+        remove_proc_entry("nfsfs", proc_root_fs);
+error_0:
+        return -ENOMEM;
+}
+/*
+ * clean up the /proc/fs/nfsfs/ directory
+ */
+void nfs_fs_proc_exit(void)
+{
+        remove_proc_entry("volumes", proc_fs_nfs);
+        remove_proc_entry("servers", proc_fs_nfs);
+        remove_proc_entry("nfsfs", proc_root_fs);
+}
+#endif /* CONFIG_PROC_FS */
diff --git a/fs/nfs/delegation.c b/fs/nfs/delegation.c
index 9540a316c05e..57133678db16 100644
--- a/fs/nfs/delegation.c
+++ b/fs/nfs/delegation.c
@@ -18,6 +18,7 @@
 #include "nfs4_fs.h"
 #include "delegation.h"
+#include "internal.h"
 static struct nfs_delegation *nfs_alloc_delegation(void)
 {
@@ -52,7 +53,7 @@ static int nfs_delegation_claim_locks(struct nfs_open_context *ctx, struct nfs4_
                        case -NFS4ERR_EXPIRED:
                                /* kill_proc(fl->fl_pid, SIGLOST, 1); */
                        case -NFS4ERR_STALE_CLIENTID:
-                                nfs4_schedule_state_recovery(NFS_SERVER(inode)->nfs4_state);
+                                nfs4_schedule_state_recovery(NFS_SERVER(inode)->nfs_client);
                                goto out_err;
                }
        }
@@ -114,7 +115,7 @@ void nfs_inode_reclaim_delegation(struct inode *inode, struct rpc_cred *cred, st
 */
 int nfs_inode_set_delegation(struct inode *inode, struct rpc_cred *cred, struct nfs_openres *res)
 {
-        struct nfs4_client *clp = NFS_SERVER(inode)->nfs4_state;
+        struct nfs_client *clp = NFS_SERVER(inode)->nfs_client;
        struct nfs_inode *nfsi = NFS_I(inode);
        struct nfs_delegation *delegation;
        int status = 0;
@@ -145,7 +146,7 @@ int nfs_inode_set_delegation(struct inode *inode, struct rpc_cred *cred, struct
                                        sizeof(delegation->stateid)) != 0 ||
                                delegation->type != nfsi->delegation->type) {
                        printk("%s: server %u.%u.%u.%u, handed out a duplicate delegation!\n",
-                                        __FUNCTION__, NIPQUAD(clp->cl_addr));
+                                        __FUNCTION__, NIPQUAD(clp->cl_addr.sin_addr));
                        status = -EIO;
                }
        }
@@ -176,7 +177,7 @@ static void nfs_msync_inode(struct inode *inode)
 */
 int __nfs_inode_return_delegation(struct inode *inode)
 {
-        struct nfs4_client *clp = NFS_SERVER(inode)->nfs4_state;
+        struct nfs_client *clp = NFS_SERVER(inode)->nfs_client;
        struct nfs_inode *nfsi = NFS_I(inode);
        struct nfs_delegation *delegation;
        int res = 0;
@@ -208,7 +209,7 @@ int __nfs_inode_return_delegation(struct inode *inode)
 */
 void nfs_return_all_delegations(struct super_block *sb)
 {
-        struct nfs4_client *clp = NFS_SB(sb)->nfs4_state;
+        struct nfs_client *clp = NFS_SB(sb)->nfs_client;
        struct nfs_delegation *delegation;
        struct inode *inode;
@@ -232,7 +233,7 @@ restart:
 int nfs_do_expire_all_delegations(void *ptr)
 {
-        struct nfs4_client *clp = ptr;
+        struct nfs_client *clp = ptr;
        struct nfs_delegation *delegation;
        struct inode *inode;
@@ -254,11 +255,11 @@ restart:
        }
 out:
        spin_unlock(&clp->cl_lock);
-        nfs4_put_client(clp);
+        nfs_put_client(clp);
        module_put_and_exit(0);
 }
-void nfs_expire_all_delegations(struct nfs4_client *clp)
+void nfs_expire_all_delegations(struct nfs_client *clp)
 {
        struct task_struct *task;
@@ -266,17 +267,17 @@ void nfs_expire_all_delegations(struct nfs4_client *clp)
        atomic_inc(&clp->cl_count);
        task = kthread_run(nfs_do_expire_all_delegations, clp,
                        "%u.%u.%u.%u-delegreturn",
-                        NIPQUAD(clp->cl_addr));
+                        NIPQUAD(clp->cl_addr.sin_addr));
        if (!IS_ERR(task))
                return;
-        nfs4_put_client(clp);
+        nfs_put_client(clp);
        module_put(THIS_MODULE);
 }
 /*
 * Return all delegations following an NFS4ERR_CB_PATH_DOWN error.
 */
-void nfs_handle_cb_pathdown(struct nfs4_client *clp)
+void nfs_handle_cb_pathdown(struct nfs_client *clp)
 {
        struct nfs_delegation *delegation;
        struct inode *inode;
@@ -299,7 +300,7 @@ restart:
 struct recall_threadargs {
        struct inode *inode;
-        struct nfs4_client *clp;
+        struct nfs_client *clp;
        const nfs4_stateid *stateid;
        struct completion started;
@@ -310,7 +311,7 @@ static int recall_thread(void *data)
 {
        struct recall_threadargs *args = (struct recall_threadargs *)data;
        struct inode *inode = igrab(args->inode);
-        struct nfs4_client *clp = NFS_SERVER(inode)->nfs4_state;
+        struct nfs_client *clp = NFS_SERVER(inode)->nfs_client;
        struct nfs_inode *nfsi = NFS_I(inode);
        struct nfs_delegation *delegation;
@@ -371,7 +372,7 @@ out_module_put:
 /*
 * Retrieve the inode associated with a delegation
 */
-struct inode *nfs_delegation_find_inode(struct nfs4_client *clp, const struct nfs_fh *fhandle)
+struct inode *nfs_delegation_find_inode(struct nfs_client *clp, const struct nfs_fh *fhandle)
 {
        struct nfs_delegation *delegation;
        struct inode *res = NULL;
@@ -389,7 +390,7 @@ struct inode *nfs_delegation_find_inode(struct nfs4_client *clp, const struct nf
 /*
 * Mark all delegations as needing to be reclaimed
 */
-void nfs_delegation_mark_reclaim(struct nfs4_client *clp)
+void nfs_delegation_mark_reclaim(struct nfs_client *clp)
 {
        struct nfs_delegation *delegation;
        spin_lock(&clp->cl_lock);
@@ -401,7 +402,7 @@ void nfs_delegation_mark_reclaim(struct nfs4_client *clp)
 /*
 * Reap all unclaimed delegations after reboot recovery is done
 */
-void nfs_delegation_reap_unclaimed(struct nfs4_client *clp)
+void nfs_delegation_reap_unclaimed(struct nfs_client *clp)
 {
        struct nfs_delegation *delegation, *n;
        LIST_HEAD(head);
@@ -423,7 +424,7 @@ void nfs_delegation_reap_unclaimed(struct nfs4_client *clp)
 int nfs4_copy_delegation_stateid(nfs4_stateid *dst, struct inode *inode)
 {
-        struct nfs4_client *clp = NFS_SERVER(inode)->nfs4_state;
+        struct nfs_client *clp = NFS_SERVER(inode)->nfs_client;
        struct nfs_inode *nfsi = NFS_I(inode);
        struct nfs_delegation *delegation;
        int res = 0;
diff --git a/fs/nfs/delegation.h b/fs/nfs/delegation.h
index 3858694652fa..2cfd4b24c7fe 100644
--- a/fs/nfs/delegation.h
+++ b/fs/nfs/delegation.h
@@ -29,13 +29,13 @@ void nfs_inode_reclaim_delegation(struct inode *inode, struct rpc_cred *cred, st
 int __nfs_inode_return_delegation(struct inode *inode);
 int nfs_async_inode_return_delegation(struct inode *inode, const nfs4_stateid *stateid);
-struct inode *nfs_delegation_find_inode(struct nfs4_client *clp, const struct nfs_fh *fhandle);
+struct inode *nfs_delegation_find_inode(struct nfs_client *clp, const struct nfs_fh *fhandle);
 void nfs_return_all_delegations(struct super_block *sb);
-void nfs_expire_all_delegations(struct nfs4_client *clp);
+void nfs_expire_all_delegations(struct nfs_client *clp);
-void nfs_handle_cb_pathdown(struct nfs4_client *clp);
+void nfs_handle_cb_pathdown(struct nfs_client *clp);
-void nfs_delegation_mark_reclaim(struct nfs4_client *clp);
+void nfs_delegation_mark_reclaim(struct nfs_client *clp);
-void nfs_delegation_reap_unclaimed(struct nfs4_client *clp);
+void nfs_delegation_reap_unclaimed(struct nfs_client *clp);
 /* NFSv4 delegation-related procedures */
 int nfs4_proc_delegreturn(struct inode *inode, struct rpc_cred *cred, const nfs4_stateid *stateid);
diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index e7ffb4deb3e5..7432f1a43f3d 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -30,7 +30,9 @@
 #include <linux/nfs_mount.h>
 #include <linux/pagemap.h>
 #include <linux/smp_lock.h>
+#include <linux/pagevec.h>
 #include <linux/namei.h>
+#include <linux/mount.h>
 #include "nfs4_fs.h"
 #include "delegation.h"
@@ -870,14 +872,14 @@ int nfs_is_exclusive_create(struct inode *dir, struct nameidata *nd)
        return (nd->intent.open.flags & O_EXCL) != 0;
 }
-static inline int nfs_reval_fsid(struct inode *dir,
+static inline int nfs_reval_fsid(struct vfsmount *mnt, struct inode *dir,
-                struct nfs_fh *fh, struct nfs_fattr *fattr)
+                                 struct nfs_fh *fh, struct nfs_fattr *fattr)
 {
        struct nfs_server *server = NFS_SERVER(dir);
        if (!nfs_fsid_equal(&server->fsid, &fattr->fsid))
                /* Revalidate fsid on root dir */
-                return __nfs_revalidate_inode(server, dir->i_sb->s_root->d_inode);
+                return __nfs_revalidate_inode(server, mnt->mnt_root->d_inode);
        return 0;
 }
@@ -902,9 +904,15 @@ static struct dentry *nfs_lookup(struct inode *dir, struct dentry * dentry, stru
        lock_kernel();
-        /* If we're doing an exclusive create, optimize away the lookup */
+        /*
-        if (nfs_is_exclusive_create(dir, nd))
+         * If we're doing an exclusive create, optimize away the lookup
-                goto no_entry;
+         * but don't hash the dentry.
+         */
+        if (nfs_is_exclusive_create(dir, nd)) {
+                d_instantiate(dentry, NULL);
+                res = NULL;
+                goto out_unlock;
+        }
        error = NFS_PROTO(dir)->lookup(dir, &dentry->d_name, &fhandle, &fattr);
        if (error == -ENOENT)
@@ -913,7 +921,7 @@ static struct dentry *nfs_lookup(struct inode *dir, struct dentry * dentry, stru
                res = ERR_PTR(error);
                goto out_unlock;
        }
-        error = nfs_reval_fsid(dir, &fhandle, &fattr);
+        error = nfs_reval_fsid(nd->mnt, dir, &fhandle, &fattr);
        if (error < 0) {
                res = ERR_PTR(error);
                goto out_unlock;
@@ -922,8 +930,9 @@ static struct dentry *nfs_lookup(struct inode *dir, struct dentry * dentry, stru
        res = (struct dentry *)inode;
        if (IS_ERR(res))
                goto out_unlock;
 no_entry:
-        res = d_add_unique(dentry, inode);
+        res = d_materialise_unique(dentry, inode);
        if (res != NULL)
                dentry = res;
        nfs_renew_times(dentry);
@@ -1117,11 +1126,13 @@ static struct dentry *nfs_readdir_lookup(nfs_readdir_descriptor_t *desc)
                dput(dentry);
                return NULL;
        }
-        alias = d_add_unique(dentry, inode);
+        alias = d_materialise_unique(dentry, inode);
        if (alias != NULL) {
                dput(dentry);
                dentry = alias;
        }
        nfs_renew_times(dentry);
        nfs_set_verifier(dentry, nfs_save_change_attribute(dir));
        return dentry;
@@ -1143,23 +1154,22 @@ int nfs_instantiate(struct dentry *dentry, struct nfs_fh *fhandle,
                struct inode *dir = dentry->d_parent->d_inode;
                error = NFS_PROTO(dir)->lookup(dir, &dentry->d_name, fhandle, fattr);
                if (error)
-                        goto out_err;
+                        return error;
        }
        if (!(fattr->valid & NFS_ATTR_FATTR)) {
                struct nfs_server *server = NFS_SB(dentry->d_sb);
-                error = server->rpc_ops->getattr(server, fhandle, fattr);
+                error = server->nfs_client->rpc_ops->getattr(server, fhandle, fattr);
                if (error < 0)
-                        goto out_err;
+                        return error;
        }
        inode = nfs_fhget(dentry->d_sb, fhandle, fattr);
        error = PTR_ERR(inode);
        if (IS_ERR(inode))
-                goto out_err;
+                return error;
        d_instantiate(dentry, inode);
+        if (d_unhashed(dentry))
+                d_rehash(dentry);
        return 0;
-out_err:
-        d_drop(dentry);
-        return error;
 }
 /*
@@ -1440,48 +1450,82 @@ static int nfs_unlink(struct inode *dir, struct dentry *dentry)
        return error;
 }
-static int
+/*
-nfs_symlink(struct inode *dir, struct dentry *dentry, const char *symname)
+ * To create a symbolic link, most file systems instantiate a new inode,
+ * add a page to it containing the path, then write it out to the disk
+ * using prepare_write/commit_write.
+ *
+ * Unfortunately the NFS client can't create the in-core inode first
+ * because it needs a file handle to create an in-core inode (see
+ * fs/nfs/inode.c:nfs_fhget).  We only have a file handle *after* the
+ * symlink request has completed on the server.
+ *
+ * So instead we allocate a raw page, copy the symname into it, then do
+ * the SYMLINK request with the page as the buffer.  If it succeeds, we
+ * now have a new file handle and can instantiate an in-core NFS inode
+ * and move the raw page into its mapping.
+ */
+static int nfs_symlink(struct inode *dir, struct dentry *dentry, const char *symname)
 {
+        struct pagevec lru_pvec;
+        struct page *page;
+        char *kaddr;
        struct iattr attr;
-        struct nfs_fattr sym_attr;
+        unsigned int pathlen = strlen(symname);
-        struct nfs_fh sym_fh;
-        struct qstr qsymname;
        int error;
        dfprintk(VFS, "NFS: symlink(%s/%ld, %s, %s)\n", dir->i_sb->s_id,
                dir->i_ino, dentry->d_name.name, symname);
-#ifdef NFS_PARANOIA
+        if (pathlen > PAGE_SIZE)
-if (dentry->d_inode)
+                return -ENAMETOOLONG;
-printk("nfs_proc_symlink: %s/%s not negative!\n",
-dentry->d_parent->d_name.name, dentry->d_name.name);
-#endif
-        /*
-         * Fill in the sattr for the call.
-         * Note: SunOS 4.1.2 crashes if the mode isn't initialized!
-         */
-        attr.ia_valid = ATTR_MODE;
-        attr.ia_mode = S_IFLNK | S_IRWXUGO;
-        qsymname.name = symname;
+        attr.ia_mode = S_IFLNK | S_IRWXUGO;
-        qsymname.len  = strlen(symname);
+        attr.ia_valid = ATTR_MODE;
        lock_kernel();
+        page = alloc_page(GFP_KERNEL);
+        if (!page) {
+                unlock_kernel();
+                return -ENOMEM;
+        }
+        kaddr = kmap_atomic(page, KM_USER0);
+        memcpy(kaddr, symname, pathlen);
+        if (pathlen < PAGE_SIZE)
+                memset(kaddr + pathlen, 0, PAGE_SIZE - pathlen);
+        kunmap_atomic(kaddr, KM_USER0);
        nfs_begin_data_update(dir);
-        error = NFS_PROTO(dir)->symlink(dir, &dentry->d_name, &qsymname,
+        error = NFS_PROTO(dir)->symlink(dir, dentry, page, pathlen, &attr);
-                                          &attr, &sym_fh, &sym_attr);
        nfs_end_data_update(dir);
-        if (!error) {
+        if (error != 0) {
-                error = nfs_instantiate(dentry, &sym_fh, &sym_attr);
+                dfprintk(VFS, "NFS: symlink(%s/%ld, %s, %s) error %d\n",
-        } else {
+                        dir->i_sb->s_id, dir->i_ino,
-                if (error == -EEXIST)
+                        dentry->d_name.name, symname, error);
-                        printk("nfs_proc_symlink: %s/%s already exists??\n",
-                               dentry->d_parent->d_name.name, dentry->d_name.name);
                d_drop(dentry);
+                __free_page(page);
+                unlock_kernel();
+                return error;
        }
+        /*
+         * No big deal if we can't add this page to the page cache here.
+         * READLINK will get the missing page from the server if needed.
+         */
+        pagevec_init(&lru_pvec, 0);
+        if (!add_to_page_cache(page, dentry->d_inode->i_mapping, 0,
+                                                        GFP_KERNEL)) {
+                if (!pagevec_add(&lru_pvec, page))
+                        __pagevec_lru_add(&lru_pvec);
+                SetPageUptodate(page);
+                unlock_page(page);
+        } else
+                __free_page(page);
        unlock_kernel();
-        return error;
+        return 0;
 }
 static int 
@@ -1625,8 +1669,7 @@ out:
        if (rehash)
                d_rehash(rehash);
        if (!error) {
-                if (!S_ISDIR(old_inode->i_mode))
+                d_move(old_dentry, new_dentry);
-                        d_move(old_dentry, new_dentry);
                nfs_renew_times(new_dentry);
                nfs_set_verifier(new_dentry, nfs_save_change_attribute(new_dir));
        }
@@ -1638,35 +1681,211 @@ out:
        return error;
 }
+static DEFINE_SPINLOCK(nfs_access_lru_lock);
+static LIST_HEAD(nfs_access_lru_list);
+static atomic_long_t nfs_access_nr_entries;
+static void nfs_access_free_entry(struct nfs_access_entry *entry)
+{
+        put_rpccred(entry->cred);
+        kfree(entry);
+        smp_mb__before_atomic_dec();
+        atomic_long_dec(&nfs_access_nr_entries);
+        smp_mb__after_atomic_dec();
+}
+int nfs_access_cache_shrinker(int nr_to_scan, gfp_t gfp_mask)
+{
+        LIST_HEAD(head);
+        struct nfs_inode *nfsi;
+        struct nfs_access_entry *cache;
+        spin_lock(&nfs_access_lru_lock);
+restart:
+        list_for_each_entry(nfsi, &nfs_access_lru_list, access_cache_inode_lru) {
+                struct inode *inode;
+                if (nr_to_scan-- == 0)
+                        break;
+                inode = igrab(&nfsi->vfs_inode);
+                if (inode == NULL)
+                        continue;
+                spin_lock(&inode->i_lock);
+                if (list_empty(&nfsi->access_cache_entry_lru))
+                        goto remove_lru_entry;
+                cache = list_entry(nfsi->access_cache_entry_lru.next,
+                                struct nfs_access_entry, lru);
+                list_move(&cache->lru, &head);
+                rb_erase(&cache->rb_node, &nfsi->access_cache);
+                if (!list_empty(&nfsi->access_cache_entry_lru))
+                        list_move_tail(&nfsi->access_cache_inode_lru,
+                                        &nfs_access_lru_list);
+                else {
+remove_lru_entry:
+                        list_del_init(&nfsi->access_cache_inode_lru);
+                        clear_bit(NFS_INO_ACL_LRU_SET, &nfsi->flags);
+                }
+                spin_unlock(&inode->i_lock);
+                iput(inode);
+                goto restart;
+        }
+        spin_unlock(&nfs_access_lru_lock);
+        while (!list_empty(&head)) {
+                cache = list_entry(head.next, struct nfs_access_entry, lru);
+                list_del(&cache->lru);
+                nfs_access_free_entry(cache);
+        }
+        return (atomic_long_read(&nfs_access_nr_entries) / 100) * sysctl_vfs_cache_pressure;
+}
+static void __nfs_access_zap_cache(struct inode *inode)
+{
+        struct nfs_inode *nfsi = NFS_I(inode);
+        struct rb_root *root_node = &nfsi->access_cache;
+        struct rb_node *n, *dispose = NULL;
+        struct nfs_access_entry *entry;
+        /* Unhook entries from the cache */
+        while ((n = rb_first(root_node)) != NULL) {
+                entry = rb_entry(n, struct nfs_access_entry, rb_node);
+                rb_erase(n, root_node);
+                list_del(&entry->lru);
+                n->rb_left = dispose;
+                dispose = n;
+        }
+        nfsi->cache_validity &= ~NFS_INO_INVALID_ACCESS;
+        spin_unlock(&inode->i_lock);
+        /* Now kill them all! */
+        while (dispose != NULL) {
+                n = dispose;
+                dispose = n->rb_left;
+                nfs_access_free_entry(rb_entry(n, struct nfs_access_entry, rb_node));
+        }
+}
+void nfs_access_zap_cache(struct inode *inode)
+{
+        /* Remove from global LRU init */
+        if (test_and_clear_bit(NFS_INO_ACL_LRU_SET, &NFS_FLAGS(inode))) {
+                spin_lock(&nfs_access_lru_lock);
+                list_del_init(&NFS_I(inode)->access_cache_inode_lru);
+                spin_unlock(&nfs_access_lru_lock);
+        }
+        spin_lock(&inode->i_lock);
+        /* This will release the spinlock */
+        __nfs_access_zap_cache(inode);
+}
+static struct nfs_access_entry *nfs_access_search_rbtree(struct inode *inode, struct rpc_cred *cred)
+{
+        struct rb_node *n = NFS_I(inode)->access_cache.rb_node;
+        struct nfs_access_entry *entry;
+        while (n != NULL) {
+                entry = rb_entry(n, struct nfs_access_entry, rb_node);
+                if (cred < entry->cred)
+                        n = n->rb_left;
+                else if (cred > entry->cred)
+                        n = n->rb_right;
+                else
+                        return entry;
+        }
+        return NULL;
+}
 int nfs_access_get_cached(struct inode *inode, struct rpc_cred *cred, struct nfs_access_entry *res)
 {
        struct nfs_inode *nfsi = NFS_I(inode);
-        struct nfs_access_entry *cache = &nfsi->cache_access;
+        struct nfs_access_entry *cache;
+        int err = -ENOENT;
-        if (cache->cred != cred
+        spin_lock(&inode->i_lock);
-                        || time_after(jiffies, cache->jiffies + NFS_ATTRTIMEO(inode))
+        if (nfsi->cache_validity & NFS_INO_INVALID_ACCESS)
-                        || (nfsi->cache_validity & NFS_INO_INVALID_ACCESS))
+                goto out_zap;
-                return -ENOENT;
+        cache = nfs_access_search_rbtree(inode, cred);
-        memcpy(res, cache, sizeof(*res));
+        if (cache == NULL)
-        return 0;
+                goto out;
+        if (time_after(jiffies, cache->jiffies + NFS_ATTRTIMEO(inode)))
+                goto out_stale;
+        res->jiffies = cache->jiffies;
+        res->cred = cache->cred;
+        res->mask = cache->mask;
+        list_move_tail(&cache->lru, &nfsi->access_cache_entry_lru);
+        err = 0;
+out:
+        spin_unlock(&inode->i_lock);
+        return err;
+out_stale:
+        rb_erase(&cache->rb_node, &nfsi->access_cache);
+        list_del(&cache->lru);
+        spin_unlock(&inode->i_lock);
+        nfs_access_free_entry(cache);
+        return -ENOENT;
+out_zap:
+        /* This will release the spinlock */
+        __nfs_access_zap_cache(inode);
+        return -ENOENT;
 }
-void nfs_access_add_cache(struct inode *inode, struct nfs_access_entry *set)
+static void nfs_access_add_rbtree(struct inode *inode, struct nfs_access_entry *set)
 {
        struct nfs_inode *nfsi = NFS_I(inode);
-        struct nfs_access_entry *cache = &nfsi->cache_access;
+        struct rb_root *root_node = &nfsi->access_cache;
+        struct rb_node **p = &root_node->rb_node;
+        struct rb_node *parent = NULL;
+        struct nfs_access_entry *entry;
-        if (cache->cred != set->cred) {
-                if (cache->cred)
-                        put_rpccred(cache->cred);
-                cache->cred = get_rpccred(set->cred);
-        }
-        /* FIXME: replace current access_cache BKL reliance with inode->i_lock */
        spin_lock(&inode->i_lock);
-        nfsi->cache_validity &= ~NFS_INO_INVALID_ACCESS;
+        while (*p != NULL) {
+                parent = *p;
+                entry = rb_entry(parent, struct nfs_access_entry, rb_node);
+                if (set->cred < entry->cred)
+                        p = &parent->rb_left;
+                else if (set->cred > entry->cred)
+                        p = &parent->rb_right;
+                else
+                        goto found;
+        }
+        rb_link_node(&set->rb_node, parent, p);
+        rb_insert_color(&set->rb_node, root_node);
+        list_add_tail(&set->lru, &nfsi->access_cache_entry_lru);
        spin_unlock(&inode->i_lock);
+        return;
+found:
+        rb_replace_node(parent, &set->rb_node, root_node);
+        list_add_tail(&set->lru, &nfsi->access_cache_entry_lru);
+        list_del(&entry->lru);
+        spin_unlock(&inode->i_lock);
+        nfs_access_free_entry(entry);
+}
+void nfs_access_add_cache(struct inode *inode, struct nfs_access_entry *set)
+{
+        struct nfs_access_entry *cache = kmalloc(sizeof(*cache), GFP_KERNEL);
+        if (cache == NULL)
+                return;
+        RB_CLEAR_NODE(&cache->rb_node);
        cache->jiffies = set->jiffies;
+        cache->cred = get_rpccred(set->cred);
        cache->mask = set->mask;
+        nfs_access_add_rbtree(inode, cache);
+        /* Update accounting */
+        smp_mb__before_atomic_inc();
+        atomic_long_inc(&nfs_access_nr_entries);
+        smp_mb__after_atomic_inc();
+        /* Add inode to global LRU list */
+        if (!test_and_set_bit(NFS_INO_ACL_LRU_SET, &NFS_FLAGS(inode))) {
+                spin_lock(&nfs_access_lru_lock);
+                list_add_tail(&NFS_I(inode)->access_cache_inode_lru, &nfs_access_lru_list);
+                spin_unlock(&nfs_access_lru_lock);
+        }
 }
 static int nfs_do_access(struct inode *inode, struct rpc_cred *cred, int mask)
diff --git a/fs/nfs/file.c b/fs/nfs/file.c
index 48e892880d5b..be997d649127 100644
--- a/fs/nfs/file.c
+++ b/fs/nfs/file.c
@@ -111,7 +111,7 @@ nfs_file_open(struct inode *inode, struct file *filp)
        nfs_inc_stats(inode, NFSIOS_VFSOPEN);
        lock_kernel();
-        res = NFS_SERVER(inode)->rpc_ops->file_open(inode, filp);
+        res = NFS_PROTO(inode)->file_open(inode, filp);
        unlock_kernel();
        return res;
 }
@@ -157,7 +157,7 @@ force_reval:
 static loff_t nfs_file_llseek(struct file *filp, loff_t offset, int origin)
 {
        /* origin == SEEK_END => we must revalidate the cached file length */
-        if (origin == 2) {
+        if (origin == SEEK_END) {
                struct inode *inode = filp->f_mapping->host;
                int retval = nfs_revalidate_file_size(inode, filp);
                if (retval < 0)
diff --git a/fs/nfs/getroot.c b/fs/nfs/getroot.c
new file mode 100644
index 000000000000..76b08ae9ed82
--- /dev/null
+++ b/fs/nfs/getroot.c
@@ -0,0 +1,311 @@
+/* getroot.c: get the root dentry for an NFS mount
+ *
+ * Copyright (C) 2006 Red Hat, Inc. All Rights Reserved.
+ * Written by David Howells (dhowells@redhat.com)
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version
+ * 2 of the License, or (at your option) any later version.
+ */
+#include <linux/config.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/time.h>
+#include <linux/kernel.h>
+#include <linux/mm.h>
+#include <linux/string.h>
+#include <linux/stat.h>
+#include <linux/errno.h>
+#include <linux/unistd.h>
+#include <linux/sunrpc/clnt.h>
+#include <linux/sunrpc/stats.h>
+#include <linux/nfs_fs.h>
+#include <linux/nfs_mount.h>
+#include <linux/nfs4_mount.h>
+#include <linux/lockd/bind.h>
+#include <linux/smp_lock.h>
+#include <linux/seq_file.h>
+#include <linux/mount.h>
+#include <linux/nfs_idmap.h>
+#include <linux/vfs.h>
+#include <linux/namei.h>
+#include <linux/namespace.h>
+#include <linux/security.h>
+#include <asm/system.h>
+#include <asm/uaccess.h>
+#include "nfs4_fs.h"
+#include "delegation.h"
+#include "internal.h"
+#define NFSDBG_FACILITY         NFSDBG_CLIENT
+#define NFS_PARANOIA 1
+/*
+ * get an NFS2/NFS3 root dentry from the root filehandle
+ */
+struct dentry *nfs_get_root(struct super_block *sb, struct nfs_fh *mntfh)
+{
+        struct nfs_server *server = NFS_SB(sb);
+        struct nfs_fsinfo fsinfo;
+        struct nfs_fattr fattr;
+        struct dentry *mntroot;
+        struct inode *inode;
+        int error;
+        /* create a dummy root dentry with dummy inode for this superblock */
+        if (!sb->s_root) {
+                struct nfs_fh dummyfh;
+                struct dentry *root;
+                struct inode *iroot;
+                memset(&dummyfh, 0, sizeof(dummyfh));
+                memset(&fattr, 0, sizeof(fattr));
+                nfs_fattr_init(&fattr);
+                fattr.valid = NFS_ATTR_FATTR;
+                fattr.type = NFDIR;
+                fattr.mode = S_IFDIR | S_IRUSR | S_IWUSR;
+                fattr.nlink = 2;
+                iroot = nfs_fhget(sb, &dummyfh, &fattr);
+                if (IS_ERR(iroot))
+                        return ERR_PTR(PTR_ERR(iroot));
+                root = d_alloc_root(iroot);
+                if (!root) {
+                        iput(iroot);
+                        return ERR_PTR(-ENOMEM);
+                }
+                sb->s_root = root;
+        }
+        /* get the actual root for this mount */
+        fsinfo.fattr = &fattr;
+        error = server->nfs_client->rpc_ops->getroot(server, mntfh, &fsinfo);
+        if (error < 0) {
+                dprintk("nfs_get_root: getattr error = %d\n", -error);
+                return ERR_PTR(error);
+        }
+        inode = nfs_fhget(sb, mntfh, fsinfo.fattr);
+        if (IS_ERR(inode)) {
+                dprintk("nfs_get_root: get root inode failed\n");
+                return ERR_PTR(PTR_ERR(inode));
+        }
+        /* root dentries normally start off anonymous and get spliced in later
+         * 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
+         */
+        mntroot = d_alloc_anon(inode);
+        if (!mntroot) {
+                iput(inode);
+                dprintk("nfs_get_root: get root dentry failed\n");
+                return ERR_PTR(-ENOMEM);
+        }
+        security_d_instantiate(mntroot, inode);
+        if (!mntroot->d_op)
+                mntroot->d_op = server->nfs_client->rpc_ops->dentry_ops;
+        return mntroot;
+}
+#ifdef CONFIG_NFS_V4
+/*
+ * Do a simple pathwalk from the root FH of the server to the nominated target
+ * of the mountpoint
+ * - give error on symlinks
+ * - give error on ".." occurring in the path
+ * - follow traversals
+ */
+int nfs4_path_walk(struct nfs_server *server,
+                   struct nfs_fh *mntfh,
+                   const char *path)
+{
+        struct nfs_fsinfo fsinfo;
+        struct nfs_fattr fattr;
+        struct nfs_fh lastfh;
+        struct qstr name;
+        int ret;
+        //int referral_count = 0;
+        dprintk("--> nfs4_path_walk(,,%s)\n", path);
+        fsinfo.fattr = &fattr;
+        nfs_fattr_init(&fattr);
+        if (*path++ != '/') {
+                dprintk("nfs4_get_root: Path does not begin with a slash\n");
+                return -EINVAL;
+        }
+        /* Start by getting the root filehandle from the server */
+        ret = server->nfs_client->rpc_ops->getroot(server, mntfh, &fsinfo);
+        if (ret < 0) {
+                dprintk("nfs4_get_root: getroot error = %d\n", -ret);
+                return ret;
+        }
+        if (fattr.type != NFDIR) {
+                printk(KERN_ERR "nfs4_get_root:"
+                       " getroot encountered non-directory\n");
+                return -ENOTDIR;
+        }
+        if (fattr.valid & NFS_ATTR_FATTR_V4_REFERRAL) {
+                printk(KERN_ERR "nfs4_get_root:"
+                       " getroot obtained referral\n");
+                return -EREMOTE;
+        }
+next_component:
+        dprintk("Next: %s\n", path);
+        /* extract the next bit of the path */
+        if (!*path)
+                goto path_walk_complete;
+        name.name = path;
+        while (*path && *path != '/')
+                path++;
+        name.len = path - (const char *) name.name;
+eat_dot_dir:
+        while (*path == '/')
+                path++;
+        if (path[0] == '.' && (path[1] == '/' || !path[1])) {
+                path += 2;
+                goto eat_dot_dir;
+        }
+        if (path[0] == '.' && path[1] == '.' && (path[2] == '/' || !path[2])
+            ) {
+                printk(KERN_ERR "nfs4_get_root:"
+                       " Mount path contains reference to \"..\"\n");
+                return -EINVAL;
+        }
+        /* lookup the next FH in the sequence */
+        memcpy(&lastfh, mntfh, sizeof(lastfh));
+        dprintk("LookupFH: %*.*s [%s]\n", name.len, name.len, name.name, path);
+        ret = server->nfs_client->rpc_ops->lookupfh(server, &lastfh, &name,
+                                                    mntfh, &fattr);
+        if (ret < 0) {
+                dprintk("nfs4_get_root: getroot error = %d\n", -ret);
+                return ret;
+        }
+        if (fattr.type != NFDIR) {
+                printk(KERN_ERR "nfs4_get_root:"
+                       " lookupfh encountered non-directory\n");
+                return -ENOTDIR;
+        }
+        if (fattr.valid & NFS_ATTR_FATTR_V4_REFERRAL) {
+                printk(KERN_ERR "nfs4_get_root:"
+                       " lookupfh obtained referral\n");
+                return -EREMOTE;
+        }
+        goto next_component;
+path_walk_complete:
+        memcpy(&server->fsid, &fattr.fsid, sizeof(server->fsid));
+        dprintk("<-- nfs4_path_walk() = 0\n");
+        return 0;
+}
+/*
+ * get an NFS4 root dentry from the root filehandle
+ */
+struct dentry *nfs4_get_root(struct super_block *sb, struct nfs_fh *mntfh)
+{
+        struct nfs_server *server = NFS_SB(sb);
+        struct nfs_fattr fattr;
+        struct dentry *mntroot;
+        struct inode *inode;
+        int error;
+        dprintk("--> nfs4_get_root()\n");
+        /* create a dummy root dentry with dummy inode for this superblock */
+        if (!sb->s_root) {
+                struct nfs_fh dummyfh;
+                struct dentry *root;
+                struct inode *iroot;
+                memset(&dummyfh, 0, sizeof(dummyfh));
+                memset(&fattr, 0, sizeof(fattr));
+                nfs_fattr_init(&fattr);
+                fattr.valid = NFS_ATTR_FATTR;
+                fattr.type = NFDIR;
+                fattr.mode = S_IFDIR | S_IRUSR | S_IWUSR;
+                fattr.nlink = 2;
+                iroot = nfs_fhget(sb, &dummyfh, &fattr);
+                if (IS_ERR(iroot))
+                        return ERR_PTR(PTR_ERR(iroot));
+                root = d_alloc_root(iroot);
+                if (!root) {
+                        iput(iroot);
+                        return ERR_PTR(-ENOMEM);
+                }
+                sb->s_root = root;
+        }
+        /* get the info about the server and filesystem */
+        error = nfs4_server_capabilities(server, mntfh);
+        if (error < 0) {
+                dprintk("nfs_get_root: getcaps error = %d\n",
+                        -error);
+                return ERR_PTR(error);
+        }
+        /* get the actual root for this mount */
+        error = server->nfs_client->rpc_ops->getattr(server, mntfh, &fattr);
+        if (error < 0) {
+                dprintk("nfs_get_root: getattr error = %d\n", -error);
+                return ERR_PTR(error);
+        }
+        inode = nfs_fhget(sb, mntfh, &fattr);
+        if (IS_ERR(inode)) {
+                dprintk("nfs_get_root: get root inode failed\n");
+                return ERR_PTR(PTR_ERR(inode));
+        }
+        /* root dentries normally start off anonymous and get spliced in later
+         * 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
+         */
+        mntroot = d_alloc_anon(inode);
+        if (!mntroot) {
+                iput(inode);
+                dprintk("nfs_get_root: get root dentry failed\n");
+                return ERR_PTR(-ENOMEM);
+        }
+        security_d_instantiate(mntroot, inode);
+        if (!mntroot->d_op)
+                mntroot->d_op = server->nfs_client->rpc_ops->dentry_ops;
+        dprintk("<-- nfs4_get_root()\n");
+        return mntroot;
+}
+#endif /* CONFIG_NFS_V4 */
diff --git a/fs/nfs/idmap.c b/fs/nfs/idmap.c
index 07a5dd57646e..82ad7110a1c0 100644
--- a/fs/nfs/idmap.c
+++ b/fs/nfs/idmap.c
@@ -57,6 +57,20 @@
 /* Default cache timeout is 10 minutes */
 unsigned int nfs_idmap_cache_timeout = 600 * HZ;
+static int param_set_idmap_timeout(const char *val, struct kernel_param *kp)
+{
+        char *endp;
+        int num = simple_strtol(val, &endp, 0);
+        int jif = num * HZ;
+        if (endp == val || *endp || num < 0 || jif < num)
+                return -EINVAL;
+        *((int *)kp->arg) = jif;
+        return 0;
+}
+module_param_call(idmap_cache_timeout, param_set_idmap_timeout, param_get_int,
+                 &nfs_idmap_cache_timeout, 0644);
 struct idmap_hashent {
        unsigned long ih_expires;
        __u32 ih_id;
@@ -70,7 +84,6 @@ struct idmap_hashtable {
 };
 struct idmap {
-        char                  idmap_path[48];
        struct dentry        *idmap_dentry;
        wait_queue_head_t     idmap_wq;
        struct idmap_msg      idmap_im;
@@ -94,24 +107,23 @@ static struct rpc_pipe_ops idmap_upcall_ops = {
        .destroy_msg    = idmap_pipe_destroy_msg,
 };
-void
+int
-nfs_idmap_new(struct nfs4_client *clp)
+nfs_idmap_new(struct nfs_client *clp)
 {
        struct idmap *idmap;
+        int error;
-        if (clp->cl_idmap != NULL)
+        BUG_ON(clp->cl_idmap != NULL);
-                return;
-        if ((idmap = kzalloc(sizeof(*idmap), GFP_KERNEL)) == NULL)
-                return;
-        snprintf(idmap->idmap_path, sizeof(idmap->idmap_path),
+        if ((idmap = kzalloc(sizeof(*idmap), GFP_KERNEL)) == NULL)
-            "%s/idmap", clp->cl_rpcclient->cl_pathname);
+                return -ENOMEM;
-        idmap->idmap_dentry = rpc_mkpipe(idmap->idmap_path,
+        idmap->idmap_dentry = rpc_mkpipe(clp->cl_rpcclient->cl_dentry, "idmap",
            idmap, &idmap_upcall_ops, 0);
        if (IS_ERR(idmap->idmap_dentry)) {
+                error = PTR_ERR(idmap->idmap_dentry);
                kfree(idmap);
-                return;
+                return error;
        }
        mutex_init(&idmap->idmap_lock);
@@ -121,10 +133,11 @@ nfs_idmap_new(struct nfs4_client *clp)
        idmap->idmap_group_hash.h_type = IDMAP_TYPE_GROUP;
        clp->cl_idmap = idmap;
+        return 0;
 }
 void
-nfs_idmap_delete(struct nfs4_client *clp)
+nfs_idmap_delete(struct nfs_client *clp)
 {
        struct idmap *idmap = clp->cl_idmap;
@@ -477,27 +490,27 @@ static unsigned int fnvhash32(const void *buf, size_t buflen)
        return (hash);
 }
-int nfs_map_name_to_uid(struct nfs4_client *clp, const char *name, size_t namelen, __u32 *uid)
+int nfs_map_name_to_uid(struct nfs_client *clp, const char *name, size_t namelen, __u32 *uid)
 {
        struct idmap *idmap = clp->cl_idmap;
        return nfs_idmap_id(idmap, &idmap->idmap_user_hash, name, namelen, uid);
 }
-int nfs_map_group_to_gid(struct nfs4_client *clp, const char *name, size_t namelen, __u32 *uid)
+int nfs_map_group_to_gid(struct nfs_client *clp, const char *name, size_t namelen, __u32 *uid)
 {
        struct idmap *idmap = clp->cl_idmap;
        return nfs_idmap_id(idmap, &idmap->idmap_group_hash, name, namelen, uid);
 }
-int nfs_map_uid_to_name(struct nfs4_client *clp, __u32 uid, char *buf)
+int nfs_map_uid_to_name(struct nfs_client *clp, __u32 uid, char *buf)
 {
        struct idmap *idmap = clp->cl_idmap;
        return nfs_idmap_name(idmap, &idmap->idmap_user_hash, uid, buf);
 }
-int nfs_map_gid_to_group(struct nfs4_client *clp, __u32 uid, char *buf)
+int nfs_map_gid_to_group(struct nfs_client *clp, __u32 uid, char *buf)
 {
        struct idmap *idmap = clp->cl_idmap;
diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c
index d349fb2245da..e8c143d182c4 100644
--- a/fs/nfs/inode.c
+++ b/fs/nfs/inode.c
@@ -76,19 +76,14 @@ int nfs_write_inode(struct inode *inode, int sync)
 void nfs_clear_inode(struct inode *inode)
 {
-        struct nfs_inode *nfsi = NFS_I(inode);
-        struct rpc_cred *cred;
        /*
         * The following should never happen...
         */
        BUG_ON(nfs_have_writebacks(inode));
-        BUG_ON (!list_empty(&nfsi->open_files));
+        BUG_ON(!list_empty(&NFS_I(inode)->open_files));
+        BUG_ON(atomic_read(&NFS_I(inode)->data_updates) != 0);
        nfs_zap_acl_cache(inode);
-        cred = nfsi->cache_access.cred;
+        nfs_access_zap_cache(inode);
-        if (cred)
-                put_rpccred(cred);
-        BUG_ON(atomic_read(&nfsi->data_updates) != 0);
 }
 /**
@@ -242,13 +237,13 @@ 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_SB(sb)->rpc_ops->file_inode_ops;
+                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)
@@ -290,7 +285,7 @@ nfs_fhget(struct super_block *sb, struct nfs_fh *fh, struct nfs_fattr *fattr)
                nfsi->attrtimeo = NFS_MINATTRTIMEO(inode);
                nfsi->attrtimeo_timestamp = jiffies;
                memset(nfsi->cookieverf, 0, sizeof(nfsi->cookieverf));
-                nfsi->cache_access.cred = NULL;
+                nfsi->access_cache = RB_ROOT;
                unlock_new_inode(inode);
        } else
@@ -722,13 +717,11 @@ void nfs_end_data_update(struct inode *inode)
 {
        struct nfs_inode *nfsi = NFS_I(inode);
-        if (!nfs_have_delegation(inode, FMODE_READ)) {
+        /* Directories: invalidate page cache */
-                /* Directories and symlinks: invalidate page cache */
+        if (S_ISDIR(inode->i_mode)) {
-                if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode)) {
+                spin_lock(&inode->i_lock);
-                        spin_lock(&inode->i_lock);
+                nfsi->cache_validity |= NFS_INO_INVALID_DATA;
-                        nfsi->cache_validity |= NFS_INO_INVALID_DATA;
+                spin_unlock(&inode->i_lock);
-                        spin_unlock(&inode->i_lock);
-                }
        }
        nfsi->cache_change_attribute = jiffies;
        atomic_dec(&nfsi->data_updates);
@@ -847,6 +840,12 @@ int nfs_refresh_inode(struct inode *inode, struct nfs_fattr *fattr)
 *
 * After an operation that has changed the inode metadata, mark the
 * attribute cache as being invalid, then try to update it.
+ *
+ * NB: if the server didn't return any post op attributes, this
+ * function will force the retrieval of attributes before the next
+ * NFS request.  Thus it should be used only for operations that
+ * are expected to change one or more attributes, to avoid
+ * unnecessary NFS requests and trips through nfs_update_inode().
 */
 int nfs_post_op_update_inode(struct inode *inode, struct nfs_fattr *fattr)
 {
@@ -1025,7 +1024,7 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr)
 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)->hostname, inode->i_sb->s_id,
+                NFS_SERVER(inode)->nfs_client->cl_hostname, inode->i_sb->s_id,
                (long long)nfsi->fileid, (long long)fattr->fileid);
        goto out_err;
 }
@@ -1109,6 +1108,8 @@ static void init_once(void * foo, kmem_cache_t * cachep, unsigned long flags)
                INIT_LIST_HEAD(&nfsi->dirty);
                INIT_LIST_HEAD(&nfsi->commit);
                INIT_LIST_HEAD(&nfsi->open_files);
+                INIT_LIST_HEAD(&nfsi->access_cache_entry_lru);
+                INIT_LIST_HEAD(&nfsi->access_cache_inode_lru);
                INIT_RADIX_TREE(&nfsi->nfs_page_tree, GFP_ATOMIC);
                atomic_set(&nfsi->data_updates, 0);
                nfsi->ndirty = 0;
@@ -1144,6 +1145,10 @@ static int __init init_nfs_fs(void)
 {
        int err;
+        err = nfs_fs_proc_init();
+        if (err)
+                goto out5;
        err = nfs_init_nfspagecache();
        if (err)
                goto out4;
@@ -1184,6 +1189,8 @@ out2:
 out3:
        nfs_destroy_nfspagecache();
 out4:
+        nfs_fs_proc_exit();
+out5:
        return err;
 }
@@ -1198,6 +1205,7 @@ static void __exit exit_nfs_fs(void)
        rpc_proc_unregister("nfs");
 #endif
        unregister_nfs_fs();
+        nfs_fs_proc_exit();
 }
 /* Not quite true; I just maintain it */
diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h
index e4f4e5def0fc..bea0b016bd70 100644
--- a/fs/nfs/internal.h
+++ b/fs/nfs/internal.h
@@ -4,6 +4,18 @@
 #include <linux/mount.h>
+struct nfs_string;
+struct nfs_mount_data;
+struct nfs4_mount_data;
+/* 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)
 struct nfs_clone_mount {
        const struct super_block *sb;
        const struct dentry *dentry;
@@ -15,7 +27,40 @@ struct nfs_clone_mount {
        rpc_authflavor_t authflavor;
 };
-/* namespace-nfs4.c */
+/* client.c */
+extern struct rpc_program nfs_program;
+extern void nfs_put_client(struct nfs_client *);
+extern struct nfs_client *nfs_find_client(const struct sockaddr_in *, int);
+extern struct nfs_server *nfs_create_server(const struct nfs_mount_data *,
+                                            struct nfs_fh *);
+extern struct nfs_server *nfs4_create_server(const struct nfs4_mount_data *,
+                                             const char *,
+                                             const struct sockaddr_in *,
+                                             const char *,
+                                             const char *,
+                                             rpc_authflavor_t,
+                                             struct nfs_fh *);
+extern struct nfs_server *nfs4_create_referral_server(struct nfs_clone_mount *,
+                                                      struct nfs_fh *);
+extern void nfs_free_server(struct nfs_server *server);
+extern struct nfs_server *nfs_clone_server(struct nfs_server *,
+                                           struct nfs_fh *,
+                                           struct nfs_fattr *);
+#ifdef CONFIG_PROC_FS
+extern int __init nfs_fs_proc_init(void);
+extern void nfs_fs_proc_exit(void);
+#else
+static inline int nfs_fs_proc_init(void)
+{
+        return 0;
+}
+static inline void nfs_fs_proc_exit(void)
+{
+}
+#endif
+/* nfs4namespace.c */
 #ifdef CONFIG_NFS_V4
 extern struct vfsmount *nfs_do_refmount(const struct vfsmount *mnt_parent, struct dentry *dentry);
 #else
@@ -46,6 +91,7 @@ extern void nfs_destroy_directcache(void);
 #endif
 /* nfs2xdr.c */
+extern int nfs_stat_to_errno(int);
 extern struct rpc_procinfo nfs_procedures[];
 extern u32 * nfs_decode_dirent(u32 *, struct nfs_entry *, int);
@@ -54,8 +100,9 @@ extern struct rpc_procinfo nfs3_procedures[];
 extern u32 *nfs3_decode_dirent(u32 *, struct nfs_entry *, int);
 /* nfs4xdr.c */
-extern int nfs_stat_to_errno(int);
+#ifdef CONFIG_NFS_V4
 extern u32 *nfs4_decode_dirent(u32 *p, struct nfs_entry *entry, int plus);
+#endif
 /* nfs4proc.c */
 #ifdef CONFIG_NFS_V4
@@ -66,6 +113,9 @@ extern int nfs4_proc_fs_locations(struct inode *dir, struct dentry *dentry,
                                  struct page *page);
 #endif
+/* dir.c */
+extern int nfs_access_cache_shrinker(int nr_to_scan, gfp_t gfp_mask);
 /* inode.c */
 extern struct inode *nfs_alloc_inode(struct super_block *sb);
 extern void nfs_destroy_inode(struct inode *);
@@ -76,10 +126,10 @@ extern void nfs4_clear_inode(struct inode *);
 #endif
 /* super.c */
-extern struct file_system_type nfs_referral_nfs4_fs_type;
+extern struct file_system_type nfs_xdev_fs_type;
-extern struct file_system_type clone_nfs_fs_type;
 #ifdef CONFIG_NFS_V4
-extern struct file_system_type clone_nfs4_fs_type;
+extern struct file_system_type nfs4_xdev_fs_type;
+extern struct file_system_type nfs4_referral_fs_type;
 #endif
 extern struct rpc_stat nfs_rpcstat;
@@ -88,30 +138,30 @@ extern int __init register_nfs_fs(void);
 extern void __exit unregister_nfs_fs(void);
 /* namespace.c */
-extern char *nfs_path(const char *base, const struct dentry *dentry,
+extern char *nfs_path(const char *base,
+                      const struct dentry *droot,
+                      const struct dentry *dentry,
                      char *buffer, ssize_t buflen);
-/*
+/* getroot.c */
- * Determine the mount path as a string
+extern struct dentry *nfs_get_root(struct super_block *, struct nfs_fh *);
- */
-static inline char *
-nfs4_path(const struct dentry *dentry, char *buffer, ssize_t buflen)
-{
 #ifdef CONFIG_NFS_V4
-        return nfs_path(NFS_SB(dentry->d_sb)->mnt_path, dentry, buffer, buflen);
+extern struct dentry *nfs4_get_root(struct super_block *, struct nfs_fh *);
-#else
-        return NULL;
+extern int nfs4_path_walk(struct nfs_server *server,
+                          struct nfs_fh *mntfh,
+                          const char *path);
 #endif
-}
 /*
 * Determine the device name as a string
 */
 static inline char *nfs_devname(const struct vfsmount *mnt_parent,
-                         const struct dentry *dentry,
+                                const struct dentry *dentry,
-                         char *buffer, ssize_t buflen)
+                                char *buffer, ssize_t buflen)
 {
-        return nfs_path(mnt_parent->mnt_devname, dentry, buffer, buflen);
+        return nfs_path(mnt_parent->mnt_devname, mnt_parent->mnt_root,
+                        dentry, buffer, buflen);
 }
 /*
@@ -167,20 +217,3 @@ void nfs_super_set_maxbytes(struct super_block *sb, __u64 maxfilesize)
        if (sb->s_maxbytes > MAX_LFS_FILESIZE || sb->s_maxbytes <= 0)
                sb->s_maxbytes = MAX_LFS_FILESIZE;
 }
-/*
- * Check if the string represents a "valid" IPv4 address
- */
-static inline int valid_ipaddr4(const char *buf)
-{
-        int rc, count, in[4];
-        rc = sscanf(buf, "%d.%d.%d.%d", &in[0], &in[1], &in[2], &in[3]);
-        if (rc != 4)
-                return -EINVAL;
-        for (count = 0; count < 4; count++) {
-                if (in[count] > 255)
-                        return -EINVAL;
-        }
-        return 0;
-}
diff --git a/fs/nfs/mount_clnt.c b/fs/nfs/mount_clnt.c
index 445abb4d4214..d507b021207f 100644
--- a/fs/nfs/mount_clnt.c
+++ b/fs/nfs/mount_clnt.c
@@ -14,7 +14,6 @@
 #include <linux/net.h>
 #include <linux/in.h>
 #include <linux/sunrpc/clnt.h>
-#include <linux/sunrpc/xprt.h>
 #include <linux/sunrpc/sched.h>
 #include <linux/nfs_fs.h>
@@ -77,22 +76,19 @@ static struct rpc_clnt *
 mnt_create(char *hostname, struct sockaddr_in *srvaddr, int version,
                int protocol)
 {
-        struct rpc_xprt *xprt;
+        struct rpc_create_args args = {
-        struct rpc_clnt *clnt;
+                .protocol       = protocol,
+                .address        = (struct sockaddr *)srvaddr,
-        xprt = xprt_create_proto(protocol, srvaddr, NULL);
+                .addrsize       = sizeof(*srvaddr),
-        if (IS_ERR(xprt))
+                .servername     = hostname,
-                return (struct rpc_clnt *)xprt;
+                .program        = &mnt_program,
+                .version        = version,
-        clnt = rpc_create_client(xprt, hostname,
+                .authflavor     = RPC_AUTH_UNIX,
-                                &mnt_program, version,
+                .flags          = (RPC_CLNT_CREATE_ONESHOT |
-                                RPC_AUTH_UNIX);
+                                   RPC_CLNT_CREATE_INTR),
-        if (!IS_ERR(clnt)) {
+        };
-                clnt->cl_softrtry = 1;
-                clnt->cl_oneshot  = 1;
+        return rpc_create(&args);
-                clnt->cl_intr = 1;
-        }
-        return clnt;
 }
 /*
diff --git a/fs/nfs/namespace.c b/fs/nfs/namespace.c
index 86b3169c8cac..77b00684894d 100644
--- a/fs/nfs/namespace.c
+++ b/fs/nfs/namespace.c
@@ -2,6 +2,7 @@
 * linux/fs/nfs/namespace.c
 *
 * Copyright (C) 2005 Trond Myklebust <Trond.Myklebust@netapp.com>
+ * - Modified by David Howells <dhowells@redhat.com>
 *
 * NFS namespace
 */
@@ -28,6 +29,7 @@ int nfs_mountpoint_expiry_timeout = 500 * HZ;
 /*
 * nfs_path - reconstruct the path given an arbitrary dentry
 * @base - arbitrary string to prepend to the path
+ * @droot - pointer to root dentry for mountpoint
 * @dentry - pointer to dentry
 * @buffer - result buffer
 * @buflen - length of buffer
@@ -38,7 +40,9 @@ int nfs_mountpoint_expiry_timeout = 500 * HZ;
 * This is mainly for use in figuring out the path on the
 * server side when automounting on top of an existing partition.
 */
-char *nfs_path(const char *base, const struct dentry *dentry,
+char *nfs_path(const char *base,
+               const struct dentry *droot,
+               const struct dentry *dentry,
               char *buffer, ssize_t buflen)
 {
        char *end = buffer+buflen;
@@ -47,7 +51,7 @@ char *nfs_path(const char *base, const struct dentry *dentry,
        *--end = '\0';
        buflen--;
        spin_lock(&dcache_lock);
-        while (!IS_ROOT(dentry)) {
+        while (!IS_ROOT(dentry) && dentry != droot) {
                namelen = dentry->d_name.len;
                buflen -= namelen + 1;
                if (buflen < 0)
@@ -96,15 +100,18 @@ static void * nfs_follow_mountpoint(struct dentry *dentry, struct nameidata *nd)
        struct nfs_fattr fattr;
        int err;
+        dprintk("--> nfs_follow_mountpoint()\n");
        BUG_ON(IS_ROOT(dentry));
        dprintk("%s: enter\n", __FUNCTION__);
        dput(nd->dentry);
        nd->dentry = dget(dentry);
-        if (d_mountpoint(nd->dentry))
-                goto out_follow;
        /* Look it up again */
        parent = dget_parent(nd->dentry);
-        err = server->rpc_ops->lookup(parent->d_inode, &nd->dentry->d_name, &fh, &fattr);
+        err = server->nfs_client->rpc_ops->lookup(parent->d_inode,
+                                                  &nd->dentry->d_name,
+                                                  &fh, &fattr);
        dput(parent);
        if (err != 0)
                goto out_err;
@@ -132,6 +139,8 @@ static void * nfs_follow_mountpoint(struct dentry *dentry, struct nameidata *nd)
        schedule_delayed_work(&nfs_automount_task, nfs_mountpoint_expiry_timeout);
 out:
        dprintk("%s: done, returned %d\n", __FUNCTION__, err);
+        dprintk("<-- nfs_follow_mountpoint() = %d\n", err);
        return ERR_PTR(err);
 out_err:
        path_release(nd);
@@ -172,22 +181,23 @@ void nfs_release_automount_timer(void)
 /*
 * Clone a mountpoint of the appropriate type
 */
-static struct vfsmount *nfs_do_clone_mount(struct nfs_server *server, char *devname,
+static struct vfsmount *nfs_do_clone_mount(struct nfs_server *server,
+                                           const char *devname,
                                           struct nfs_clone_mount *mountdata)
 {
 #ifdef CONFIG_NFS_V4
        struct vfsmount *mnt = NULL;
-        switch (server->rpc_ops->version) {
+        switch (server->nfs_client->cl_nfsversion) {
                case 2:
                case 3:
-                        mnt = vfs_kern_mount(&clone_nfs_fs_type, 0, devname, mountdata);
+                        mnt = vfs_kern_mount(&nfs_xdev_fs_type, 0, devname, mountdata);
                        break;
                case 4:
-                        mnt = vfs_kern_mount(&clone_nfs4_fs_type, 0, devname, mountdata);
+                        mnt = vfs_kern_mount(&nfs4_xdev_fs_type, 0, devname, mountdata);
        }
        return mnt;
 #else
-        return vfs_kern_mount(&clone_nfs_fs_type, 0, devname, mountdata);
+        return vfs_kern_mount(&nfs_xdev_fs_type, 0, devname, mountdata);
 #endif
 }
@@ -213,6 +223,8 @@ struct vfsmount *nfs_do_submount(const struct vfsmount *mnt_parent,
        char *page = (char *) __get_free_page(GFP_USER);
        char *devname;
+        dprintk("--> nfs_do_submount()\n");
        dprintk("%s: submounting on %s/%s\n", __FUNCTION__,
                        dentry->d_parent->d_name.name,
                        dentry->d_name.name);
@@ -227,5 +239,7 @@ free_page:
        free_page((unsigned long)page);
 out:
        dprintk("%s: done\n", __FUNCTION__);
+        dprintk("<-- nfs_do_submount() = %p\n", mnt);
        return mnt;
 }
diff --git a/fs/nfs/nfs2xdr.c b/fs/nfs/nfs2xdr.c
index 67391eef6b93..b49501fc0a79 100644
--- a/fs/nfs/nfs2xdr.c
+++ b/fs/nfs/nfs2xdr.c
@@ -51,7 +51,7 @@
 #define NFS_createargs_sz       (NFS_diropargs_sz+NFS_sattr_sz)
 #define NFS_renameargs_sz       (NFS_diropargs_sz+NFS_diropargs_sz)
 #define NFS_linkargs_sz         (NFS_fhandle_sz+NFS_diropargs_sz)
-#define NFS_symlinkargs_sz      (NFS_diropargs_sz+NFS_path_sz+NFS_sattr_sz)
+#define NFS_symlinkargs_sz      (NFS_diropargs_sz+1+NFS_sattr_sz)
 #define NFS_readdirargs_sz      (NFS_fhandle_sz+2)
 #define NFS_attrstat_sz         (1+NFS_fattr_sz)
@@ -351,11 +351,26 @@ nfs_xdr_linkargs(struct rpc_rqst *req, u32 *p, struct nfs_linkargs *args)
 static int
 nfs_xdr_symlinkargs(struct rpc_rqst *req, u32 *p, struct nfs_symlinkargs *args)
 {
+        struct xdr_buf *sndbuf = &req->rq_snd_buf;
+        size_t pad;
        p = xdr_encode_fhandle(p, args->fromfh);
        p = xdr_encode_array(p, args->fromname, args->fromlen);
-        p = xdr_encode_array(p, args->topath, args->tolen);
+        *p++ = htonl(args->pathlen);
+        sndbuf->len = xdr_adjust_iovec(sndbuf->head, p);
+        xdr_encode_pages(sndbuf, args->pages, 0, args->pathlen);
+        /*
+         * xdr_encode_pages may have added a few bytes to ensure the
+         * pathname ends on a 4-byte boundary.  Start encoding the
+         * attributes after the pad bytes.
+         */
+        pad = sndbuf->tail->iov_len;
+        if (pad > 0)
+                p++;
        p = xdr_encode_sattr(p, args->sattr);
-        req->rq_slen = xdr_adjust_iovec(req->rq_svec, p);
+        sndbuf->len += xdr_adjust_iovec(sndbuf->tail, p) - pad;
        return 0;
 }
diff --git a/fs/nfs/nfs3proc.c b/fs/nfs/nfs3proc.c
index 7143b1f82cea..f8688eaa0001 100644
--- a/fs/nfs/nfs3proc.c
+++ b/fs/nfs/nfs3proc.c
@@ -81,7 +81,7 @@ do_proc_get_root(struct rpc_clnt *client, struct nfs_fh *fhandle,
 }
 /*
- * Bare-bones access to getattr: this is for nfs_read_super.
+ * Bare-bones access to getattr: this is for nfs_get_root/nfs_get_sb
 */
 static int
 nfs3_proc_get_root(struct nfs_server *server, struct nfs_fh *fhandle,
@@ -90,8 +90,8 @@ nfs3_proc_get_root(struct nfs_server *server, struct nfs_fh *fhandle,
        int     status;
        status = do_proc_get_root(server->client, fhandle, info);
-        if (status && server->client_sys != server->client)
+        if (status && server->nfs_client->cl_rpcclient != server->client)
-                status = do_proc_get_root(server->client_sys, fhandle, info);
+                status = do_proc_get_root(server->nfs_client->cl_rpcclient, fhandle, info);
        return status;
 }
@@ -544,23 +544,23 @@ nfs3_proc_link(struct inode *inode, struct inode *dir, struct qstr *name)
 }
 static int
-nfs3_proc_symlink(struct inode *dir, struct qstr *name, struct qstr *path,
+nfs3_proc_symlink(struct inode *dir, struct dentry *dentry, struct page *page,
-                  struct iattr *sattr, struct nfs_fh *fhandle,
+                  unsigned int len, struct iattr *sattr)
-                  struct nfs_fattr *fattr)
 {
-        struct nfs_fattr        dir_attr;
+        struct nfs_fh fhandle;
+        struct nfs_fattr fattr, dir_attr;
        struct nfs3_symlinkargs arg = {
                .fromfh         = NFS_FH(dir),
-                .fromname       = name->name,
+                .fromname       = dentry->d_name.name,
-                .fromlen        = name->len,
+                .fromlen        = dentry->d_name.len,
-                .topath         = path->name,
+                .pages          = &page,
-                .tolen          = path->len,
+                .pathlen        = len,
                .sattr          = sattr
        };
        struct nfs3_diropres    res = {
                .dir_attr       = &dir_attr,
-                .fh             = fhandle,
+                .fh             = &fhandle,
-                .fattr          = fattr
+                .fattr          = &fattr
        };
        struct rpc_message msg = {
                .rpc_proc       = &nfs3_procedures[NFS3PROC_SYMLINK],
@@ -569,13 +569,19 @@ nfs3_proc_symlink(struct inode *dir, struct qstr *name, struct qstr *path,
        };
        int                     status;
-        if (path->len > NFS3_MAXPATHLEN)
+        if (len > NFS3_MAXPATHLEN)
                return -ENAMETOOLONG;
-        dprintk("NFS call  symlink %s -> %s\n", name->name, path->name);
+        dprintk("NFS call  symlink %s\n", dentry->d_name.name);
        nfs_fattr_init(&dir_attr);
-        nfs_fattr_init(fattr);
+        nfs_fattr_init(&fattr);
        status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0);
        nfs_post_op_update_inode(dir, &dir_attr);
+        if (status != 0)
+                goto out;
+        status = nfs_instantiate(dentry, &fhandle, &fattr);
+out:
        dprintk("NFS reply symlink: %d\n", status);
        return status;
 }
@@ -785,7 +791,7 @@ nfs3_proc_fsinfo(struct nfs_server *server, struct nfs_fh *fhandle,
        dprintk("NFS call  fsinfo\n");
        nfs_fattr_init(info->fattr);
-        status = rpc_call_sync(server->client_sys, &msg, 0);
+        status = rpc_call_sync(server->nfs_client->cl_rpcclient, &msg, 0);
        dprintk("NFS reply fsinfo: %d\n", status);
        return status;
 }
@@ -886,7 +892,7 @@ nfs3_proc_lock(struct file *filp, int cmd, struct file_lock *fl)
        return nlmclnt_proc(filp->f_dentry->d_inode, cmd, fl);
 }
-struct nfs_rpc_ops      nfs_v3_clientops = {
+const struct nfs_rpc_ops nfs_v3_clientops = {
        .version        = 3,                    /* protocol version */
        .dentry_ops     = &nfs_dentry_operations,
        .dir_inode_ops  = &nfs3_dir_inode_operations,
diff --git a/fs/nfs/nfs3xdr.c b/fs/nfs/nfs3xdr.c
index 0250269e9753..16556fa4effb 100644
--- a/fs/nfs/nfs3xdr.c
+++ b/fs/nfs/nfs3xdr.c
@@ -56,7 +56,7 @@
 #define NFS3_writeargs_sz       (NFS3_fh_sz+5)
 #define NFS3_createargs_sz      (NFS3_diropargs_sz+NFS3_sattr_sz)
 #define NFS3_mkdirargs_sz       (NFS3_diropargs_sz+NFS3_sattr_sz)
-#define NFS3_symlinkargs_sz     (NFS3_diropargs_sz+NFS3_path_sz+NFS3_sattr_sz)
+#define NFS3_symlinkargs_sz     (NFS3_diropargs_sz+1+NFS3_sattr_sz)
 #define NFS3_mknodargs_sz       (NFS3_diropargs_sz+2+NFS3_sattr_sz)
 #define NFS3_renameargs_sz      (NFS3_diropargs_sz+NFS3_diropargs_sz)
 #define NFS3_linkargs_sz                (NFS3_fh_sz+NFS3_diropargs_sz)
@@ -398,8 +398,11 @@ nfs3_xdr_symlinkargs(struct rpc_rqst *req, u32 *p, struct nfs3_symlinkargs *args
        p = xdr_encode_fhandle(p, args->fromfh);
        p = xdr_encode_array(p, args->fromname, args->fromlen);
        p = xdr_encode_sattr(p, args->sattr);
-        p = xdr_encode_array(p, args->topath, args->tolen);
+        *p++ = htonl(args->pathlen);
        req->rq_slen = xdr_adjust_iovec(req->rq_svec, p);
+        /* Copy the page */
+        xdr_encode_pages(&req->rq_snd_buf, args->pages, 0, args->pathlen);
        return 0;
 }
diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h
index 9a102860df37..61095fe4b5ca 100644
--- a/fs/nfs/nfs4_fs.h
+++ b/fs/nfs/nfs4_fs.h
@@ -43,55 +43,6 @@ enum nfs4_client_state {
 };
 /*
- * The nfs4_client identifies our client state to the server.
- */
-struct nfs4_client {
-        struct list_head        cl_servers;     /* Global list of servers */
-        struct in_addr          cl_addr;        /* Server identifier */
-        u64                     cl_clientid;    /* constant */
-        nfs4_verifier           cl_confirm;
-        unsigned long           cl_state;
-        u32                     cl_lockowner_id;
-        /*
-         * The following rwsem ensures exclusive access to the server
-         * while we recover the state following a lease expiration.
-         */
-        struct rw_semaphore     cl_sem;
-        struct list_head        cl_delegations;
-        struct list_head        cl_state_owners;
-        struct list_head        cl_unused;
-        int                     cl_nunused;
-        spinlock_t              cl_lock;
-        atomic_t                cl_count;
-        struct rpc_clnt *       cl_rpcclient;
-        struct list_head        cl_superblocks; /* List of nfs_server structs */
-        unsigned long           cl_lease_time;
-        unsigned long           cl_last_renewal;
-        struct work_struct      cl_renewd;
-        struct work_struct      cl_recoverd;
-        struct rpc_wait_queue   cl_rpcwaitq;
-        /* used for the setclientid verifier */
-        struct timespec         cl_boot_time;
-        /* idmapper */
-        struct idmap *          cl_idmap;
-        /* Our own IP address, as a null-terminated string.
-         * This is used to generate the clientid, and the callback address.
-         */
-        char                    cl_ipaddr[16];
-        unsigned char           cl_id_uniquifier;
-};
-/*
 * struct rpc_sequence ensures that RPC calls are sent in the exact
 * order that they appear on the list.
 */
@@ -127,7 +78,7 @@ static inline void nfs_confirm_seqid(struct nfs_seqid_counter *seqid, int status
 struct nfs4_state_owner {
        spinlock_t           so_lock;
        struct list_head     so_list;    /* per-clientid list of state_owners */
-        struct nfs4_client   *so_client;
+        struct nfs_client    *so_client;
        u32                  so_id;      /* 32-bit identifier, unique */
        atomic_t             so_count;
@@ -210,10 +161,10 @@ extern ssize_t nfs4_listxattr(struct dentry *, char *, size_t);
 /* nfs4proc.c */
 extern int nfs4_map_errors(int err);
-extern int nfs4_proc_setclientid(struct nfs4_client *, u32, unsigned short, struct rpc_cred *);
+extern int nfs4_proc_setclientid(struct nfs_client *, u32, unsigned short, struct rpc_cred *);
-extern int nfs4_proc_setclientid_confirm(struct nfs4_client *, struct rpc_cred *);
+extern int nfs4_proc_setclientid_confirm(struct nfs_client *, struct rpc_cred *);
-extern int nfs4_proc_async_renew(struct nfs4_client *, struct rpc_cred *);
+extern int nfs4_proc_async_renew(struct nfs_client *, struct rpc_cred *);
-extern int nfs4_proc_renew(struct nfs4_client *, struct rpc_cred *);
+extern int nfs4_proc_renew(struct nfs_client *, struct rpc_cred *);
 extern int nfs4_do_close(struct inode *inode, struct nfs4_state *state);
 extern struct dentry *nfs4_atomic_open(struct inode *, struct dentry *, struct nameidata *);
 extern int nfs4_open_revalidate(struct inode *, struct dentry *, int, struct nameidata *);
@@ -231,19 +182,14 @@ extern const u32 nfs4_fsinfo_bitmap[2];
 extern const u32 nfs4_fs_locations_bitmap[2];
 /* nfs4renewd.c */
-extern void nfs4_schedule_state_renewal(struct nfs4_client *);
+extern void nfs4_schedule_state_renewal(struct nfs_client *);
 extern void nfs4_renewd_prepare_shutdown(struct nfs_server *);
-extern void nfs4_kill_renewd(struct nfs4_client *);
+extern void nfs4_kill_renewd(struct nfs_client *);
 extern void nfs4_renew_state(void *);
 /* nfs4state.c */
-extern void init_nfsv4_state(struct nfs_server *);
+struct rpc_cred *nfs4_get_renew_cred(struct nfs_client *clp);
-extern void destroy_nfsv4_state(struct nfs_server *);
+extern u32 nfs4_alloc_lockowner_id(struct nfs_client *);
-extern struct nfs4_client *nfs4_get_client(struct in_addr *);
-extern void nfs4_put_client(struct nfs4_client *clp);
-extern struct nfs4_client *nfs4_find_client(struct in_addr *);
-struct rpc_cred *nfs4_get_renew_cred(struct nfs4_client *clp);
-extern u32 nfs4_alloc_lockowner_id(struct nfs4_client *);
 extern struct nfs4_state_owner * nfs4_get_state_owner(struct nfs_server *, struct rpc_cred *);
 extern void nfs4_put_state_owner(struct nfs4_state_owner *);
@@ -252,7 +198,7 @@ extern struct nfs4_state * nfs4_get_open_state(struct inode *, struct nfs4_state
 extern void nfs4_put_open_state(struct nfs4_state *);
 extern void nfs4_close_state(struct nfs4_state *, mode_t);
 extern void nfs4_state_set_mode_locked(struct nfs4_state *, mode_t);
-extern void nfs4_schedule_state_recovery(struct nfs4_client *);
+extern void nfs4_schedule_state_recovery(struct nfs_client *);
 extern void nfs4_put_lock_state(struct nfs4_lock_state *lsp);
 extern int nfs4_set_lock_state(struct nfs4_state *state, struct file_lock *fl);
 extern void nfs4_copy_stateid(nfs4_stateid *, struct nfs4_state *, fl_owner_t);
@@ -276,10 +222,6 @@ extern struct svc_version nfs4_callback_version1;
 #else
-#define init_nfsv4_state(server)  do { } while (0)
-#define destroy_nfsv4_state(server)       do { } while (0)
-#define nfs4_put_state_owner(inode, owner) do { } while (0)
-#define nfs4_put_open_state(state) do { } while (0)
 #define nfs4_close_state(a, b) do { } while (0)
 #endif /* CONFIG_NFS_V4 */
diff --git a/fs/nfs/nfs4namespace.c b/fs/nfs/nfs4namespace.c
index ea38d27b74e6..24e47f3bbd17 100644
--- a/fs/nfs/nfs4namespace.c
+++ b/fs/nfs/nfs4namespace.c
@@ -2,6 +2,7 @@
 * linux/fs/nfs/nfs4namespace.c
 *
 * Copyright (C) 2005 Trond Myklebust <Trond.Myklebust@netapp.com>
+ * - Modified by David Howells <dhowells@redhat.com>
 *
 * NFSv4 namespace
 */
@@ -23,7 +24,7 @@
 /*
 * Check if fs_root is valid
 */
-static inline char *nfs4_pathname_string(struct nfs4_pathname *pathname,
+static inline char *nfs4_pathname_string(const struct nfs4_pathname *pathname,
                                         char *buffer, ssize_t buflen)
 {
        char *end = buffer + buflen;
@@ -34,7 +35,7 @@ static inline char *nfs4_pathname_string(struct nfs4_pathname *pathname,
        n = pathname->ncomponents;
        while (--n >= 0) {
-                struct nfs4_string *component = &pathname->components[n];
+                const struct nfs4_string *component = &pathname->components[n];
                buflen -= component->len + 1;
                if (buflen < 0)
                        goto Elong;
@@ -47,6 +48,68 @@ Elong:
        return ERR_PTR(-ENAMETOOLONG);
 }
+/*
+ * Determine the mount path as a string
+ */
+static char *nfs4_path(const struct vfsmount *mnt_parent,
+                       const struct dentry *dentry,
+                       char *buffer, ssize_t buflen)
+{
+        const char *srvpath;
+        srvpath = strchr(mnt_parent->mnt_devname, ':');
+        if (srvpath)
+                srvpath++;
+        else
+                srvpath = mnt_parent->mnt_devname;
+        return nfs_path(srvpath, mnt_parent->mnt_root, dentry, buffer, buflen);
+}
+/*
+ * Check that fs_locations::fs_root [RFC3530 6.3] is a prefix for what we
+ * believe to be the server path to this dentry
+ */
+static int nfs4_validate_fspath(const struct vfsmount *mnt_parent,
+                                const struct dentry *dentry,
+                                const struct nfs4_fs_locations *locations,
+                                char *page, char *page2)
+{
+        const char *path, *fs_path;
+        path = nfs4_path(mnt_parent, dentry, page, PAGE_SIZE);
+        if (IS_ERR(path))
+                return PTR_ERR(path);
+        fs_path = nfs4_pathname_string(&locations->fs_path, page2, PAGE_SIZE);
+        if (IS_ERR(fs_path))
+                return PTR_ERR(fs_path);
+        if (strncmp(path, fs_path, strlen(fs_path)) != 0) {
+                dprintk("%s: path %s does not begin with fsroot %s\n",
+                        __FUNCTION__, path, fs_path);
+                return -ENOENT;
+        }
+        return 0;
+}
+/*
+ * Check if the string represents a "valid" IPv4 address
+ */
+static inline int valid_ipaddr4(const char *buf)
+{
+        int rc, count, in[4];
+        rc = sscanf(buf, "%d.%d.%d.%d", &in[0], &in[1], &in[2], &in[3]);
+        if (rc != 4)
+                return -EINVAL;
+        for (count = 0; count < 4; count++) {
+                if (in[count] > 255)
+                        return -EINVAL;
+        }
+        return 0;
+}
 /**
 * nfs_follow_referral - set up mountpoint when hitting a referral on moved error
@@ -60,7 +123,7 @@ Elong:
 */
 static struct vfsmount *nfs_follow_referral(const struct vfsmount *mnt_parent,
                                            const struct dentry *dentry,
-                                            struct nfs4_fs_locations *locations)
+                                            const struct nfs4_fs_locations *locations)
 {
        struct vfsmount *mnt = ERR_PTR(-ENOENT);
        struct nfs_clone_mount mountdata = {
@@ -68,10 +131,9 @@ static struct vfsmount *nfs_follow_referral(const struct vfsmount *mnt_parent,
                .dentry = dentry,
                .authflavor = NFS_SB(mnt_parent->mnt_sb)->client->cl_auth->au_flavor,
        };
-        char *page, *page2;
+        char *page = NULL, *page2 = NULL;
-        char *path, *fs_path;
        char *devname;
-        int loc, s;
+        int loc, s, error;
        if (locations == NULL || locations->nlocations <= 0)
                goto out;
@@ -79,36 +141,30 @@ static struct vfsmount *nfs_follow_referral(const struct vfsmount *mnt_parent,
        dprintk("%s: referral at %s/%s\n", __FUNCTION__,
                dentry->d_parent->d_name.name, dentry->d_name.name);
-        /* Ensure fs path is a prefix of current dentry path */
        page = (char *) __get_free_page(GFP_USER);
-        if (page == NULL)
+        if (!page)
                goto out;
        page2 = (char *) __get_free_page(GFP_USER);
-        if (page2 == NULL)
+        if (!page2)
                goto out;
-        path = nfs4_path(dentry, page, PAGE_SIZE);
+        /* Ensure fs path is a prefix of current dentry path */
-        if (IS_ERR(path))
+        error = nfs4_validate_fspath(mnt_parent, dentry, locations, page, page2);
-                goto out_free;
+        if (error < 0) {
+                mnt = ERR_PTR(error);
-        fs_path = nfs4_pathname_string(&locations->fs_path, page2, PAGE_SIZE);
+                goto out;
-        if (IS_ERR(fs_path))
-                goto out_free;
-        if (strncmp(path, fs_path, strlen(fs_path)) != 0) {
-                dprintk("%s: path %s does not begin with fsroot %s\n", __FUNCTION__, path, fs_path);
-                goto out_free;
        }
        devname = nfs_devname(mnt_parent, dentry, page, PAGE_SIZE);
        if (IS_ERR(devname)) {
                mnt = (struct vfsmount *)devname;
-                goto out_free;
+                goto out;
        }
        loc = 0;
        while (loc < locations->nlocations && IS_ERR(mnt)) {
-                struct nfs4_fs_location *location = &locations->locations[loc];
+                const struct nfs4_fs_location *location = &locations->locations[loc];
                char *mnt_path;
                if (location == NULL || location->nservers <= 0 ||
@@ -140,7 +196,7 @@ static struct vfsmount *nfs_follow_referral(const struct vfsmount *mnt_parent,
                        addr.sin_port = htons(NFS_PORT);
                        mountdata.addr = &addr;
-                        mnt = vfs_kern_mount(&nfs_referral_nfs4_fs_type, 0, devname, &mountdata);
+                        mnt = vfs_kern_mount(&nfs4_referral_fs_type, 0, devname, &mountdata);
                        if (!IS_ERR(mnt)) {
                                break;
                        }
@@ -149,10 +205,9 @@ static struct vfsmount *nfs_follow_referral(const struct vfsmount *mnt_parent,
                loc++;
        }
-out_free:
-        free_page((unsigned long)page);
-        free_page((unsigned long)page2);
 out:
+        free_page((unsigned long) page);
+        free_page((unsigned long) page2);
        dprintk("%s: done\n", __FUNCTION__);
        return mnt;
 }
@@ -165,7 +220,7 @@ out:
 */
 struct vfsmount *nfs_do_refmount(const struct vfsmount *mnt_parent, struct dentry *dentry)
 {
-        struct vfsmount *mnt = ERR_PTR(-ENOENT);
+        struct vfsmount *mnt = ERR_PTR(-ENOMEM);
        struct dentry *parent;
        struct nfs4_fs_locations *fs_locations = NULL;
        struct page *page;
@@ -183,11 +238,16 @@ struct vfsmount *nfs_do_refmount(const struct vfsmount *mnt_parent, struct dentr
                goto out_free;
        /* Get locations */
+        mnt = ERR_PTR(-ENOENT);
        parent = dget_parent(dentry);
-        dprintk("%s: getting locations for %s/%s\n", __FUNCTION__, parent->d_name.name, dentry->d_name.name);
+        dprintk("%s: getting locations for %s/%s\n",
+                __FUNCTION__, parent->d_name.name, dentry->d_name.name);
        err = nfs4_proc_fs_locations(parent->d_inode, dentry, fs_locations, page);
        dput(parent);
-        if (err != 0 || fs_locations->nlocations <= 0 ||
+        if (err != 0 ||
+            fs_locations->nlocations <= 0 ||
            fs_locations->fs_path.ncomponents <= 0)
                goto out_free;
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index b14145b7b87f..47c7e6e3910d 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -55,7 +55,7 @@
 #define NFSDBG_FACILITY         NFSDBG_PROC
-#define NFS4_POLL_RETRY_MIN     (1*HZ)
+#define NFS4_POLL_RETRY_MIN     (HZ/10)
 #define NFS4_POLL_RETRY_MAX     (15*HZ)
 struct nfs4_opendata;
@@ -64,7 +64,7 @@ static int nfs4_do_fsinfo(struct nfs_server *, struct nfs_fh *, struct nfs_fsinf
 static int nfs4_async_handle_error(struct rpc_task *, const struct nfs_server *);
 static int _nfs4_proc_access(struct inode *inode, struct nfs_access_entry *entry);
 static int nfs4_handle_exception(const struct nfs_server *server, int errorcode, struct nfs4_exception *exception);
-static int nfs4_wait_clnt_recover(struct rpc_clnt *clnt, struct nfs4_client *clp);
+static int nfs4_wait_clnt_recover(struct rpc_clnt *clnt, struct nfs_client *clp);
 /* Prevent leaks of NFSv4 errors into userland */
 int nfs4_map_errors(int err)
@@ -195,7 +195,7 @@ static void nfs4_setup_readdir(u64 cookie, u32 *verifier, struct dentry *dentry,
 static void renew_lease(const struct nfs_server *server, unsigned long timestamp)
 {
-        struct nfs4_client *clp = server->nfs4_state;
+        struct nfs_client *clp = server->nfs_client;
        spin_lock(&clp->cl_lock);
        if (time_before(clp->cl_last_renewal,timestamp))
                clp->cl_last_renewal = timestamp;
@@ -252,7 +252,7 @@ static struct nfs4_opendata *nfs4_opendata_alloc(struct dentry *dentry,
        atomic_inc(&sp->so_count);
        p->o_arg.fh = NFS_FH(dir);
        p->o_arg.open_flags = flags,
-        p->o_arg.clientid = server->nfs4_state->cl_clientid;
+        p->o_arg.clientid = server->nfs_client->cl_clientid;
        p->o_arg.id = sp->so_id;
        p->o_arg.name = &dentry->d_name;
        p->o_arg.server = server;
@@ -550,7 +550,7 @@ int nfs4_open_delegation_recall(struct dentry *dentry, struct nfs4_state *state)
                        case -NFS4ERR_STALE_STATEID:
                        case -NFS4ERR_EXPIRED:
                                /* Don't recall a delegation if it was lost */
-                                nfs4_schedule_state_recovery(server->nfs4_state);
+                                nfs4_schedule_state_recovery(server->nfs_client);
                                return err;
                }
                err = nfs4_handle_exception(server, err, &exception);
@@ -758,7 +758,7 @@ static int _nfs4_proc_open(struct nfs4_opendata *data)
        }
        nfs_confirm_seqid(&data->owner->so_seqid, 0);
        if (!(o_res->f_attr->valid & NFS_ATTR_FATTR))
-                return server->rpc_ops->getattr(server, &o_res->fh, o_res->f_attr);
+                return server->nfs_client->rpc_ops->getattr(server, &o_res->fh, o_res->f_attr);
        return 0;
 }
@@ -792,11 +792,18 @@ out:
 int nfs4_recover_expired_lease(struct nfs_server *server)
 {
-        struct nfs4_client *clp = server->nfs4_state;
+        struct nfs_client *clp = server->nfs_client;
+        int ret;
-        if (test_and_clear_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state))
+        for (;;) {
+                ret = nfs4_wait_clnt_recover(server->client, clp);
+                if (ret != 0)
+                        return ret;
+                if (!test_and_clear_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state))
+                        break;
                nfs4_schedule_state_recovery(clp);
-        return nfs4_wait_clnt_recover(server->client, clp);
+        }
+        return 0;
 }
 /*
@@ -867,7 +874,7 @@ static int _nfs4_open_delegated(struct inode *inode, int flags, struct rpc_cred
 {
        struct nfs_delegation *delegation;
        struct nfs_server *server = NFS_SERVER(inode);
-        struct nfs4_client *clp = server->nfs4_state;
+        struct nfs_client *clp = server->nfs_client;
        struct nfs_inode *nfsi = NFS_I(inode);
        struct nfs4_state_owner *sp = NULL;
        struct nfs4_state *state = NULL;
@@ -953,7 +960,7 @@ static int _nfs4_do_open(struct inode *dir, struct dentry *dentry, int flags, st
        struct nfs4_state_owner  *sp;
        struct nfs4_state     *state = NULL;
        struct nfs_server       *server = NFS_SERVER(dir);
-        struct nfs4_client *clp = server->nfs4_state;
+        struct nfs_client *clp = server->nfs_client;
        struct nfs4_opendata *opendata;
        int                     status;
@@ -1133,7 +1140,7 @@ static void nfs4_close_done(struct rpc_task *task, void *data)
                        break;
                case -NFS4ERR_STALE_STATEID:
                case -NFS4ERR_EXPIRED:
-                        nfs4_schedule_state_recovery(server->nfs4_state);
+                        nfs4_schedule_state_recovery(server->nfs_client);
                        break;
                default:
                        if (nfs4_async_handle_error(task, server) == -EAGAIN) {
@@ -1268,7 +1275,7 @@ nfs4_atomic_open(struct inode *dir, struct dentry *dentry, struct nameidata *nd)
                BUG_ON(nd->intent.open.flags & O_CREAT);
        }
-        cred = rpcauth_lookupcred(NFS_SERVER(dir)->client->cl_auth, 0);
+        cred = rpcauth_lookupcred(NFS_CLIENT(dir)->cl_auth, 0);
        if (IS_ERR(cred))
                return (struct dentry *)cred;
        state = nfs4_do_open(dir, dentry, nd->intent.open.flags, &attr, cred);
@@ -1291,7 +1298,7 @@ nfs4_open_revalidate(struct inode *dir, struct dentry *dentry, int openflags, st
        struct rpc_cred *cred;
        struct nfs4_state *state;
-        cred = rpcauth_lookupcred(NFS_SERVER(dir)->client->cl_auth, 0);
+        cred = rpcauth_lookupcred(NFS_CLIENT(dir)->cl_auth, 0);
        if (IS_ERR(cred))
                return PTR_ERR(cred);
        state = nfs4_open_delegated(dentry->d_inode, openflags, cred);
@@ -1393,70 +1400,19 @@ static int nfs4_lookup_root(struct nfs_server *server, struct nfs_fh *fhandle,
        return err;
 }
+/*
+ * get the file handle for the "/" directory on the server
+ */
 static int nfs4_proc_get_root(struct nfs_server *server, struct nfs_fh *fhandle,
-                struct nfs_fsinfo *info)
+                              struct nfs_fsinfo *info)
 {
-        struct nfs_fattr *      fattr = info->fattr;
-        unsigned char *         p;
-        struct qstr             q;
-        struct nfs4_lookup_arg args = {
-                .dir_fh = fhandle,
-                .name = &q,
-                .bitmask = nfs4_fattr_bitmap,
-        };
-        struct nfs4_lookup_res res = {
-                .server = server,
-                .fattr = fattr,
-                .fh = fhandle,
-        };
-        struct rpc_message msg = {
-                .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_LOOKUP],
-                .rpc_argp = &args,
-                .rpc_resp = &res,
-        };
        int status;
-        /*
-         * Now we do a separate LOOKUP for each component of the mount path.
-         * The LOOKUPs are done separately so that we can conveniently
-         * catch an ERR_WRONGSEC if it occurs along the way...
-         */
        status = nfs4_lookup_root(server, fhandle, info);
-        if (status)
-                goto out;
-        p = server->mnt_path;
-        for (;;) {
-                struct nfs4_exception exception = { };
-                while (*p == '/')
-                        p++;
-                if (!*p)
-                        break;
-                q.name = p;
-                while (*p && (*p != '/'))
-                        p++;
-                q.len = p - q.name;
-                do {
-                        nfs_fattr_init(fattr);
-                        status = nfs4_handle_exception(server,
-                                        rpc_call_sync(server->client, &msg, 0),
-                                        &exception);
-                } while (exception.retry);
-                if (status == 0)
-                        continue;
-                if (status == -ENOENT) {
-                        printk(KERN_NOTICE "NFS: mount path %s does not exist!\n", server->mnt_path);
-                        printk(KERN_NOTICE "NFS: suggestion: try mounting '/' instead.\n");
-                }
-                break;
-        }
        if (status == 0)
                status = nfs4_server_capabilities(server, fhandle);
        if (status == 0)
                status = nfs4_do_fsinfo(server, fhandle, info);
-out:
        return nfs4_map_errors(status);
 }
@@ -1565,7 +1521,7 @@ nfs4_proc_setattr(struct dentry *dentry, struct nfs_fattr *fattr,
        nfs_fattr_init(fattr);
        
-        cred = rpcauth_lookupcred(NFS_SERVER(inode)->client->cl_auth, 0);
+        cred = rpcauth_lookupcred(NFS_CLIENT(inode)->cl_auth, 0);
        if (IS_ERR(cred))
                return PTR_ERR(cred);
@@ -1583,6 +1539,52 @@ nfs4_proc_setattr(struct dentry *dentry, struct nfs_fattr *fattr,
        return status;
 }
+static int _nfs4_proc_lookupfh(struct nfs_server *server, struct nfs_fh *dirfh,
+                struct qstr *name, struct nfs_fh *fhandle,
+                struct nfs_fattr *fattr)
+{
+        int                    status;
+        struct nfs4_lookup_arg args = {
+                .bitmask = server->attr_bitmask,
+                .dir_fh = dirfh,
+                .name = name,
+        };
+        struct nfs4_lookup_res res = {
+                .server = server,
+                .fattr = fattr,
+                .fh = fhandle,
+        };
+        struct rpc_message msg = {
+                .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_LOOKUP],
+                .rpc_argp = &args,
+                .rpc_resp = &res,
+        };
+        nfs_fattr_init(fattr);
+        dprintk("NFS call  lookupfh %s\n", name->name);
+        status = rpc_call_sync(server->client, &msg, 0);
+        dprintk("NFS reply lookupfh: %d\n", status);
+        if (status == -NFS4ERR_MOVED)
+                status = -EREMOTE;
+        return status;
+}
+static int nfs4_proc_lookupfh(struct nfs_server *server, struct nfs_fh *dirfh,
+                              struct qstr *name, struct nfs_fh *fhandle,
+                              struct nfs_fattr *fattr)
+{
+        struct nfs4_exception exception = { };
+        int err;
+        do {
+                err = nfs4_handle_exception(server,
+                                _nfs4_proc_lookupfh(server, dirfh, name,
+                                                    fhandle, fattr),
+                                &exception);
+        } while (exception.retry);
+        return err;
+}
 static int _nfs4_proc_lookup(struct inode *dir, struct qstr *name,
                struct nfs_fh *fhandle, struct nfs_fattr *fattr)
 {
@@ -1881,7 +1883,7 @@ nfs4_proc_create(struct inode *dir, struct dentry *dentry, struct iattr *sattr,
        struct rpc_cred *cred;
        int status = 0;
-        cred = rpcauth_lookupcred(NFS_SERVER(dir)->client->cl_auth, 0);
+        cred = rpcauth_lookupcred(NFS_CLIENT(dir)->cl_auth, 0);
        if (IS_ERR(cred)) {
                status = PTR_ERR(cred);
                goto out;
@@ -2089,24 +2091,24 @@ static int nfs4_proc_link(struct inode *inode, struct inode *dir, struct qstr *n
        return err;
 }
-static int _nfs4_proc_symlink(struct inode *dir, struct qstr *name,
+static int _nfs4_proc_symlink(struct inode *dir, struct dentry *dentry,
-                struct qstr *path, struct iattr *sattr, struct nfs_fh *fhandle,
+                struct page *page, unsigned int len, struct iattr *sattr)
-                struct nfs_fattr *fattr)
 {
        struct nfs_server *server = NFS_SERVER(dir);
-        struct nfs_fattr dir_fattr;
+        struct nfs_fh fhandle;
+        struct nfs_fattr fattr, dir_fattr;
        struct nfs4_create_arg arg = {
                .dir_fh = NFS_FH(dir),
                .server = server,
-                .name = name,
+                .name = &dentry->d_name,
                .attrs = sattr,
                .ftype = NF4LNK,
                .bitmask = server->attr_bitmask,
        };
        struct nfs4_create_res res = {
                .server = server,
-                .fh = fhandle,
+                .fh = &fhandle,
-                .fattr = fattr,
+                .fattr = &fattr,
                .dir_fattr = &dir_fattr,
        };
        struct rpc_message msg = {
@@ -2116,29 +2118,32 @@ static int _nfs4_proc_symlink(struct inode *dir, struct qstr *name,
        };
        int                     status;
-        if (path->len > NFS4_MAXPATHLEN)
+        if (len > NFS4_MAXPATHLEN)
                return -ENAMETOOLONG;
-        arg.u.symlink = path;
-        nfs_fattr_init(fattr);
+        arg.u.symlink.pages = &page;
+        arg.u.symlink.len = len;
+        nfs_fattr_init(&fattr);
        nfs_fattr_init(&dir_fattr);
        
        status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0);
-        if (!status)
+        if (!status) {
                update_changeattr(dir, &res.dir_cinfo);
-        nfs_post_op_update_inode(dir, res.dir_fattr);
+                nfs_post_op_update_inode(dir, res.dir_fattr);
+                status = nfs_instantiate(dentry, &fhandle, &fattr);
+        }
        return status;
 }
-static int nfs4_proc_symlink(struct inode *dir, struct qstr *name,
+static int nfs4_proc_symlink(struct inode *dir, struct dentry *dentry,
-                struct qstr *path, struct iattr *sattr, struct nfs_fh *fhandle,
+                struct page *page, unsigned int len, struct iattr *sattr)
-                struct nfs_fattr *fattr)
 {
        struct nfs4_exception exception = { };
        int err;
        do {
                err = nfs4_handle_exception(NFS_SERVER(dir),
-                                _nfs4_proc_symlink(dir, name, path, sattr,
+                                _nfs4_proc_symlink(dir, dentry, page,
-                                        fhandle, fattr),
+                                                        len, sattr),
                                &exception);
        } while (exception.retry);
        return err;
@@ -2521,7 +2526,7 @@ static void nfs4_proc_commit_setup(struct nfs_write_data *data, int how)
 */
 static void nfs4_renew_done(struct rpc_task *task, void *data)
 {
-        struct nfs4_client *clp = (struct nfs4_client *)task->tk_msg.rpc_argp;
+        struct nfs_client *clp = (struct nfs_client *)task->tk_msg.rpc_argp;
        unsigned long timestamp = (unsigned long)data;
        if (task->tk_status < 0) {
@@ -2543,7 +2548,7 @@ static const struct rpc_call_ops nfs4_renew_ops = {
        .rpc_call_done = nfs4_renew_done,
 };
-int nfs4_proc_async_renew(struct nfs4_client *clp, struct rpc_cred *cred)
+int nfs4_proc_async_renew(struct nfs_client *clp, struct rpc_cred *cred)
 {
        struct rpc_message msg = {
                .rpc_proc       = &nfs4_procedures[NFSPROC4_CLNT_RENEW],
@@ -2555,7 +2560,7 @@ int nfs4_proc_async_renew(struct nfs4_client *clp, struct rpc_cred *cred)
                        &nfs4_renew_ops, (void *)jiffies);
 }
-int nfs4_proc_renew(struct nfs4_client *clp, struct rpc_cred *cred)
+int nfs4_proc_renew(struct nfs_client *clp, struct rpc_cred *cred)
 {
        struct rpc_message msg = {
                .rpc_proc       = &nfs4_procedures[NFSPROC4_CLNT_RENEW],
@@ -2770,7 +2775,7 @@ static int __nfs4_proc_set_acl(struct inode *inode, const void *buf, size_t bufl
                return -EOPNOTSUPP;
        nfs_inode_return_delegation(inode);
        buf_to_pages(buf, buflen, arg.acl_pages, &arg.acl_pgbase);
-        ret = rpc_call_sync(NFS_SERVER(inode)->client, &msg, 0);
+        ret = rpc_call_sync(NFS_CLIENT(inode), &msg, 0);
        if (ret == 0)
                nfs4_write_cached_acl(inode, buf, buflen);
        return ret;
@@ -2791,7 +2796,7 @@ static int nfs4_proc_set_acl(struct inode *inode, const void *buf, size_t buflen
 static int
 nfs4_async_handle_error(struct rpc_task *task, const struct nfs_server *server)
 {
-        struct nfs4_client *clp = server->nfs4_state;
+        struct nfs_client *clp = server->nfs_client;
        if (!clp || task->tk_status >= 0)
                return 0;
@@ -2828,7 +2833,7 @@ static int nfs4_wait_bit_interruptible(void *word)
        return 0;
 }
-static int nfs4_wait_clnt_recover(struct rpc_clnt *clnt, struct nfs4_client *clp)
+static int nfs4_wait_clnt_recover(struct rpc_clnt *clnt, struct nfs_client *clp)
 {
        sigset_t oldset;
        int res;
@@ -2871,7 +2876,7 @@ static int nfs4_delay(struct rpc_clnt *clnt, long *timeout)
 */
 int nfs4_handle_exception(const struct nfs_server *server, int errorcode, struct nfs4_exception *exception)
 {
-        struct nfs4_client *clp = server->nfs4_state;
+        struct nfs_client *clp = server->nfs_client;
        int ret = errorcode;
        exception->retry = 0;
@@ -2886,6 +2891,7 @@ int nfs4_handle_exception(const struct nfs_server *server, int errorcode, struct
                        if (ret == 0)
                                exception->retry = 1;
                        break;
+                case -NFS4ERR_FILE_OPEN:
                case -NFS4ERR_GRACE:
                case -NFS4ERR_DELAY:
                        ret = nfs4_delay(server->client, &exception->timeout);
@@ -2898,7 +2904,7 @@ int nfs4_handle_exception(const struct nfs_server *server, int errorcode, struct
        return nfs4_map_errors(ret);
 }
-int nfs4_proc_setclientid(struct nfs4_client *clp, u32 program, unsigned short port, struct rpc_cred *cred)
+int nfs4_proc_setclientid(struct nfs_client *clp, u32 program, unsigned short port, struct rpc_cred *cred)
 {
        nfs4_verifier sc_verifier;
        struct nfs4_setclientid setclientid = {
@@ -2922,7 +2928,7 @@ int nfs4_proc_setclientid(struct nfs4_client *clp, u32 program, unsigned short p
        for(;;) {
                setclientid.sc_name_len = scnprintf(setclientid.sc_name,
                                sizeof(setclientid.sc_name), "%s/%u.%u.%u.%u %s %u",
-                                clp->cl_ipaddr, NIPQUAD(clp->cl_addr.s_addr),
+                                clp->cl_ipaddr, NIPQUAD(clp->cl_addr.sin_addr),
                                cred->cr_ops->cr_name,
                                clp->cl_id_uniquifier);
                setclientid.sc_netid_len = scnprintf(setclientid.sc_netid,
@@ -2945,7 +2951,7 @@ int nfs4_proc_setclientid(struct nfs4_client *clp, u32 program, unsigned short p
        return status;
 }
-static int _nfs4_proc_setclientid_confirm(struct nfs4_client *clp, struct rpc_cred *cred)
+static int _nfs4_proc_setclientid_confirm(struct nfs_client *clp, struct rpc_cred *cred)
 {
        struct nfs_fsinfo fsinfo;
        struct rpc_message msg = {
@@ -2969,7 +2975,7 @@ static int _nfs4_proc_setclientid_confirm(struct nfs4_client *clp, struct rpc_cr
        return status;
 }
-int nfs4_proc_setclientid_confirm(struct nfs4_client *clp, struct rpc_cred *cred)
+int nfs4_proc_setclientid_confirm(struct nfs_client *clp, struct rpc_cred *cred)
 {
        long timeout;
        int err;
@@ -3077,7 +3083,7 @@ int nfs4_proc_delegreturn(struct inode *inode, struct rpc_cred *cred, const nfs4
                switch (err) {
                        case -NFS4ERR_STALE_STATEID:
                        case -NFS4ERR_EXPIRED:
-                                nfs4_schedule_state_recovery(server->nfs4_state);
+                                nfs4_schedule_state_recovery(server->nfs_client);
                        case 0:
                                return 0;
                }
@@ -3106,7 +3112,7 @@ static int _nfs4_proc_getlk(struct nfs4_state *state, int cmd, struct file_lock
 {
        struct inode *inode = state->inode;
        struct nfs_server *server = NFS_SERVER(inode);
-        struct nfs4_client *clp = server->nfs4_state;
+        struct nfs_client *clp = server->nfs_client;
        struct nfs_lockt_args arg = {
                .fh = NFS_FH(inode),
                .fl = request,
@@ -3231,7 +3237,7 @@ static void nfs4_locku_done(struct rpc_task *task, void *data)
                        break;
                case -NFS4ERR_STALE_STATEID:
                case -NFS4ERR_EXPIRED:
-                        nfs4_schedule_state_recovery(calldata->server->nfs4_state);
+                        nfs4_schedule_state_recovery(calldata->server->nfs_client);
                        break;
                default:
                        if (nfs4_async_handle_error(task, calldata->server) == -EAGAIN) {
@@ -3343,7 +3349,7 @@ static struct nfs4_lockdata *nfs4_alloc_lockdata(struct file_lock *fl,
        if (p->arg.lock_seqid == NULL)
                goto out_free;
        p->arg.lock_stateid = &lsp->ls_stateid;
-        p->arg.lock_owner.clientid = server->nfs4_state->cl_clientid;
+        p->arg.lock_owner.clientid = server->nfs_client->cl_clientid;
        p->arg.lock_owner.id = lsp->ls_id;
        p->lsp = lsp;
        atomic_inc(&lsp->ls_count);
@@ -3513,7 +3519,7 @@ static int nfs4_lock_expired(struct nfs4_state *state, struct file_lock *request
 static int _nfs4_proc_setlk(struct nfs4_state *state, int cmd, struct file_lock *request)
 {
-        struct nfs4_client *clp = state->owner->so_client;
+        struct nfs_client *clp = state->owner->so_client;
        unsigned char fl_flags = request->fl_flags;
        int status;
@@ -3715,7 +3721,7 @@ static struct inode_operations nfs4_file_inode_operations = {
        .listxattr      = nfs4_listxattr,
 };
-struct nfs_rpc_ops      nfs_v4_clientops = {
+const struct nfs_rpc_ops nfs_v4_clientops = {
        .version        = 4,                    /* protocol version */
        .dentry_ops     = &nfs4_dentry_operations,
        .dir_inode_ops  = &nfs4_dir_inode_operations,
@@ -3723,6 +3729,7 @@ struct nfs_rpc_ops	nfs_v4_clientops = {
        .getroot        = nfs4_proc_get_root,
        .getattr        = nfs4_proc_getattr,
        .setattr        = nfs4_proc_setattr,
+        .lookupfh       = nfs4_proc_lookupfh,
        .lookup         = nfs4_proc_lookup,
        .access         = nfs4_proc_access,
        .readlink       = nfs4_proc_readlink,
@@ -3743,6 +3750,7 @@ struct nfs_rpc_ops	nfs_v4_clientops = {
        .statfs         = nfs4_proc_statfs,
        .fsinfo         = nfs4_proc_fsinfo,
        .pathconf       = nfs4_proc_pathconf,
+        .set_capabilities = nfs4_server_capabilities,
        .decode_dirent  = nfs4_decode_dirent,
        .read_setup     = nfs4_proc_read_setup,
        .read_done      = nfs4_read_done,
diff --git a/fs/nfs/nfs4renewd.c b/fs/nfs/nfs4renewd.c
index 5d764d8e6d8a..7b6df1852e75 100644
--- a/fs/nfs/nfs4renewd.c
+++ b/fs/nfs/nfs4renewd.c
@@ -61,7 +61,7 @@
 void
 nfs4_renew_state(void *data)
 {
-        struct nfs4_client *clp = (struct nfs4_client *)data;
+        struct nfs_client *clp = (struct nfs_client *)data;
        struct rpc_cred *cred;
        long lease, timeout;
        unsigned long last, now;
@@ -108,7 +108,7 @@ out:
 /* Must be called with clp->cl_sem locked for writes */
 void
-nfs4_schedule_state_renewal(struct nfs4_client *clp)
+nfs4_schedule_state_renewal(struct nfs_client *clp)
 {
        long timeout;
@@ -121,32 +121,20 @@ nfs4_schedule_state_renewal(struct nfs4_client *clp)
                        __FUNCTION__, (timeout + HZ - 1) / HZ);
        cancel_delayed_work(&clp->cl_renewd);
        schedule_delayed_work(&clp->cl_renewd, timeout);
+        set_bit(NFS_CS_RENEWD, &clp->cl_res_state);
        spin_unlock(&clp->cl_lock);
 }
 void
 nfs4_renewd_prepare_shutdown(struct nfs_server *server)
 {
-        struct nfs4_client *clp = server->nfs4_state;
-        if (!clp)
-                return;
        flush_scheduled_work();
-        down_write(&clp->cl_sem);
-        if (!list_empty(&server->nfs4_siblings))
-                list_del_init(&server->nfs4_siblings);
-        up_write(&clp->cl_sem);
 }
-/* Must be called with clp->cl_sem locked for writes */
 void
-nfs4_kill_renewd(struct nfs4_client *clp)
+nfs4_kill_renewd(struct nfs_client *clp)
 {
        down_read(&clp->cl_sem);
-        if (!list_empty(&clp->cl_superblocks)) {
-                up_read(&clp->cl_sem);
-                return;
-        }
        cancel_delayed_work(&clp->cl_renewd);
        up_read(&clp->cl_sem);
        flush_scheduled_work();
diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c
index 090a36b07a22..5fffbdfa971f 100644
--- a/fs/nfs/nfs4state.c
+++ b/fs/nfs/nfs4state.c
@@ -50,149 +50,15 @@
 #include "nfs4_fs.h"
 #include "callback.h"
 #include "delegation.h"
+#include "internal.h"
 #define OPENOWNER_POOL_SIZE     8
 const nfs4_stateid zero_stateid;
-static DEFINE_SPINLOCK(state_spinlock);
 static LIST_HEAD(nfs4_clientid_list);
-void
+static int nfs4_init_client(struct nfs_client *clp, struct rpc_cred *cred)
-init_nfsv4_state(struct nfs_server *server)
-{
-        server->nfs4_state = NULL;
-        INIT_LIST_HEAD(&server->nfs4_siblings);
-}
-void
-destroy_nfsv4_state(struct nfs_server *server)
-{
-        kfree(server->mnt_path);
-        server->mnt_path = NULL;
-        if (server->nfs4_state) {
-                nfs4_put_client(server->nfs4_state);
-                server->nfs4_state = NULL;
-        }
-}
-/*
- * nfs4_get_client(): returns an empty client structure
- * nfs4_put_client(): drops reference to client structure
- *
- * Since these are allocated/deallocated very rarely, we don't
- * bother putting them in a slab cache...
- */
-static struct nfs4_client *
-nfs4_alloc_client(struct in_addr *addr)
-{
-        struct nfs4_client *clp;
-        if (nfs_callback_up() < 0)
-                return NULL;
-        if ((clp = kzalloc(sizeof(*clp), GFP_KERNEL)) == NULL) {
-                nfs_callback_down();
-                return NULL;
-        }
-        memcpy(&clp->cl_addr, addr, sizeof(clp->cl_addr));
-        init_rwsem(&clp->cl_sem);
-        INIT_LIST_HEAD(&clp->cl_delegations);
-        INIT_LIST_HEAD(&clp->cl_state_owners);
-        INIT_LIST_HEAD(&clp->cl_unused);
-        spin_lock_init(&clp->cl_lock);
-        atomic_set(&clp->cl_count, 1);
-        INIT_WORK(&clp->cl_renewd, nfs4_renew_state, clp);
-        INIT_LIST_HEAD(&clp->cl_superblocks);
-        rpc_init_wait_queue(&clp->cl_rpcwaitq, "NFS4 client");
-        clp->cl_rpcclient = ERR_PTR(-EINVAL);
-        clp->cl_boot_time = CURRENT_TIME;
-        clp->cl_state = 1 << NFS4CLNT_LEASE_EXPIRED;
-        return clp;
-}
-static void
-nfs4_free_client(struct nfs4_client *clp)
-{
-        struct nfs4_state_owner *sp;
-        while (!list_empty(&clp->cl_unused)) {
-                sp = list_entry(clp->cl_unused.next,
-                                struct nfs4_state_owner,
-                                so_list);
-                list_del(&sp->so_list);
-                kfree(sp);
-        }
-        BUG_ON(!list_empty(&clp->cl_state_owners));
-        nfs_idmap_delete(clp);
-        if (!IS_ERR(clp->cl_rpcclient))
-                rpc_shutdown_client(clp->cl_rpcclient);
-        kfree(clp);
-        nfs_callback_down();
-}
-static struct nfs4_client *__nfs4_find_client(struct in_addr *addr)
-{
-        struct nfs4_client *clp;
-        list_for_each_entry(clp, &nfs4_clientid_list, cl_servers) {
-                if (memcmp(&clp->cl_addr, addr, sizeof(clp->cl_addr)) == 0) {
-                        atomic_inc(&clp->cl_count);
-                        return clp;
-                }
-        }
-        return NULL;
-}
-struct nfs4_client *nfs4_find_client(struct in_addr *addr)
-{
-        struct nfs4_client *clp;
-        spin_lock(&state_spinlock);
-        clp = __nfs4_find_client(addr);
-        spin_unlock(&state_spinlock);
-        return clp;
-}
-struct nfs4_client *
-nfs4_get_client(struct in_addr *addr)
-{
-        struct nfs4_client *clp, *new = NULL;
-        spin_lock(&state_spinlock);
-        for (;;) {
-                clp = __nfs4_find_client(addr);
-                if (clp != NULL)
-                        break;
-                clp = new;
-                if (clp != NULL) {
-                        list_add(&clp->cl_servers, &nfs4_clientid_list);
-                        new = NULL;
-                        break;
-                }
-                spin_unlock(&state_spinlock);
-                new = nfs4_alloc_client(addr);
-                spin_lock(&state_spinlock);
-                if (new == NULL)
-                        break;
-        }
-        spin_unlock(&state_spinlock);
-        if (new)
-                nfs4_free_client(new);
-        return clp;
-}
-void
-nfs4_put_client(struct nfs4_client *clp)
-{
-        if (!atomic_dec_and_lock(&clp->cl_count, &state_spinlock))
-                return;
-        list_del(&clp->cl_servers);
-        spin_unlock(&state_spinlock);
-        BUG_ON(!list_empty(&clp->cl_superblocks));
-        rpc_wake_up(&clp->cl_rpcwaitq);
-        nfs4_kill_renewd(clp);
-        nfs4_free_client(clp);
-}
-static int nfs4_init_client(struct nfs4_client *clp, struct rpc_cred *cred)
 {
        int status = nfs4_proc_setclientid(clp, NFS4_CALLBACK,
                        nfs_callback_tcpport, cred);
@@ -204,13 +70,13 @@ static int nfs4_init_client(struct nfs4_client *clp, struct rpc_cred *cred)
 }
 u32
-nfs4_alloc_lockowner_id(struct nfs4_client *clp)
+nfs4_alloc_lockowner_id(struct nfs_client *clp)
 {
        return clp->cl_lockowner_id ++;
 }
 static struct nfs4_state_owner *
-nfs4_client_grab_unused(struct nfs4_client *clp, struct rpc_cred *cred)
+nfs4_client_grab_unused(struct nfs_client *clp, struct rpc_cred *cred)
 {
        struct nfs4_state_owner *sp = NULL;
@@ -224,7 +90,7 @@ nfs4_client_grab_unused(struct nfs4_client *clp, struct rpc_cred *cred)
        return sp;
 }
-struct rpc_cred *nfs4_get_renew_cred(struct nfs4_client *clp)
+struct rpc_cred *nfs4_get_renew_cred(struct nfs_client *clp)
 {
        struct nfs4_state_owner *sp;
        struct rpc_cred *cred = NULL;
@@ -238,7 +104,7 @@ struct rpc_cred *nfs4_get_renew_cred(struct nfs4_client *clp)
        return cred;
 }
-struct rpc_cred *nfs4_get_setclientid_cred(struct nfs4_client *clp)
+struct rpc_cred *nfs4_get_setclientid_cred(struct nfs_client *clp)
 {
        struct nfs4_state_owner *sp;
@@ -251,7 +117,7 @@ struct rpc_cred *nfs4_get_setclientid_cred(struct nfs4_client *clp)
 }
 static struct nfs4_state_owner *
-nfs4_find_state_owner(struct nfs4_client *clp, struct rpc_cred *cred)
+nfs4_find_state_owner(struct nfs_client *clp, struct rpc_cred *cred)
 {
        struct nfs4_state_owner *sp, *res = NULL;
@@ -294,7 +160,7 @@ nfs4_alloc_state_owner(void)
 void
 nfs4_drop_state_owner(struct nfs4_state_owner *sp)
 {
-        struct nfs4_client *clp = sp->so_client;
+        struct nfs_client *clp = sp->so_client;
        spin_lock(&clp->cl_lock);
        list_del_init(&sp->so_list);
        spin_unlock(&clp->cl_lock);
@@ -306,7 +172,7 @@ nfs4_drop_state_owner(struct nfs4_state_owner *sp)
 */
 struct nfs4_state_owner *nfs4_get_state_owner(struct nfs_server *server, struct rpc_cred *cred)
 {
-        struct nfs4_client *clp = server->nfs4_state;
+        struct nfs_client *clp = server->nfs_client;
        struct nfs4_state_owner *sp, *new;
        get_rpccred(cred);
@@ -337,7 +203,7 @@ struct nfs4_state_owner *nfs4_get_state_owner(struct nfs_server *server, struct
 */
 void nfs4_put_state_owner(struct nfs4_state_owner *sp)
 {
-        struct nfs4_client *clp = sp->so_client;
+        struct nfs_client *clp = sp->so_client;
        struct rpc_cred *cred = sp->so_cred;
        if (!atomic_dec_and_lock(&sp->so_count, &clp->cl_lock))
@@ -540,7 +406,7 @@ __nfs4_find_lock_state(struct nfs4_state *state, fl_owner_t fl_owner)
 static struct nfs4_lock_state *nfs4_alloc_lock_state(struct nfs4_state *state, fl_owner_t fl_owner)
 {
        struct nfs4_lock_state *lsp;
-        struct nfs4_client *clp = state->owner->so_client;
+        struct nfs_client *clp = state->owner->so_client;
        lsp = kzalloc(sizeof(*lsp), GFP_KERNEL);
        if (lsp == NULL)
@@ -752,7 +618,7 @@ out:
 static int reclaimer(void *);
-static inline void nfs4_clear_recover_bit(struct nfs4_client *clp)
+static inline void nfs4_clear_recover_bit(struct nfs_client *clp)
 {
        smp_mb__before_clear_bit();
        clear_bit(NFS4CLNT_STATE_RECOVER, &clp->cl_state);
@@ -764,25 +630,25 @@ static inline void nfs4_clear_recover_bit(struct nfs4_client *clp)
 /*
 * State recovery routine
 */
-static void nfs4_recover_state(struct nfs4_client *clp)
+static void nfs4_recover_state(struct nfs_client *clp)
 {
        struct task_struct *task;
        __module_get(THIS_MODULE);
        atomic_inc(&clp->cl_count);
        task = kthread_run(reclaimer, clp, "%u.%u.%u.%u-reclaim",
-                        NIPQUAD(clp->cl_addr));
+                        NIPQUAD(clp->cl_addr.sin_addr));
        if (!IS_ERR(task))
                return;
        nfs4_clear_recover_bit(clp);
-        nfs4_put_client(clp);
+        nfs_put_client(clp);
        module_put(THIS_MODULE);
 }
 /*
 * Schedule a state recovery attempt
 */
-void nfs4_schedule_state_recovery(struct nfs4_client *clp)
+void nfs4_schedule_state_recovery(struct nfs_client *clp)
 {
        if (!clp)
                return;
@@ -879,7 +745,7 @@ out_err:
        return status;
 }
-static void nfs4_state_mark_reclaim(struct nfs4_client *clp)
+static void nfs4_state_mark_reclaim(struct nfs_client *clp)
 {
        struct nfs4_state_owner *sp;
        struct nfs4_state *state;
@@ -903,7 +769,7 @@ static void nfs4_state_mark_reclaim(struct nfs4_client *clp)
 static int reclaimer(void *ptr)
 {
-        struct nfs4_client *clp = ptr;
+        struct nfs_client *clp = ptr;
        struct nfs4_state_owner *sp;
        struct nfs4_state_recovery_ops *ops;
        struct rpc_cred *cred;
@@ -970,12 +836,12 @@ out:
        if (status == -NFS4ERR_CB_PATH_DOWN)
                nfs_handle_cb_pathdown(clp);
        nfs4_clear_recover_bit(clp);
-        nfs4_put_client(clp);
+        nfs_put_client(clp);
        module_put_and_exit(0);
        return 0;
 out_error:
        printk(KERN_WARNING "Error: state recovery failed on NFSv4 server %u.%u.%u.%u with error %d\n",
-                                NIPQUAD(clp->cl_addr.s_addr), -status);
+                                NIPQUAD(clp->cl_addr.sin_addr), -status);
        set_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state);
        goto out;
 }
diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c
index 730ec8fb31c6..3dd413f52da1 100644
--- a/fs/nfs/nfs4xdr.c
+++ b/fs/nfs/nfs4xdr.c
@@ -58,7 +58,7 @@
 /* Mapping from NFS error code to "errno" error code. */
 #define errno_NFSERR_IO         EIO
-static int nfs_stat_to_errno(int);
+static int nfs4_stat_to_errno(int);
 /* NFSv4 COMPOUND tags are only wanted for debugging purposes */
 #ifdef DEBUG
@@ -128,7 +128,7 @@ static int nfs_stat_to_errno(int);
 #define decode_link_maxsz       (op_decode_hdr_maxsz + 5)
 #define encode_symlink_maxsz    (op_encode_hdr_maxsz + \
                                1 + nfs4_name_maxsz + \
-                                nfs4_path_maxsz + \
+                                1 + \
                                nfs4_fattr_maxsz)
 #define decode_symlink_maxsz    (op_decode_hdr_maxsz + 8)
 #define encode_create_maxsz     (op_encode_hdr_maxsz + \
@@ -529,7 +529,7 @@ static int encode_attrs(struct xdr_stream *xdr, const struct iattr *iap, const s
        if (iap->ia_valid & ATTR_MODE)
                len += 4;
        if (iap->ia_valid & ATTR_UID) {
-                owner_namelen = nfs_map_uid_to_name(server->nfs4_state, iap->ia_uid, owner_name);
+                owner_namelen = nfs_map_uid_to_name(server->nfs_client, iap->ia_uid, owner_name);
                if (owner_namelen < 0) {
                        printk(KERN_WARNING "nfs: couldn't resolve uid %d to string\n",
                               iap->ia_uid);
@@ -541,7 +541,7 @@ static int encode_attrs(struct xdr_stream *xdr, const struct iattr *iap, const s
                len += 4 + (XDR_QUADLEN(owner_namelen) << 2);
        }
        if (iap->ia_valid & ATTR_GID) {
-                owner_grouplen = nfs_map_gid_to_group(server->nfs4_state, iap->ia_gid, owner_group);
+                owner_grouplen = nfs_map_gid_to_group(server->nfs_client, iap->ia_gid, owner_group);
                if (owner_grouplen < 0) {
                        printk(KERN_WARNING "nfs4: couldn't resolve gid %d to string\n",
                               iap->ia_gid);
@@ -673,9 +673,9 @@ static int encode_create(struct xdr_stream *xdr, const struct nfs4_create_arg *c
        switch (create->ftype) {
        case NF4LNK:
-                RESERVE_SPACE(4 + create->u.symlink->len);
+                RESERVE_SPACE(4);
-                WRITE32(create->u.symlink->len);
+                WRITE32(create->u.symlink.len);
-                WRITEMEM(create->u.symlink->name, create->u.symlink->len);
+                xdr_write_pages(xdr, create->u.symlink.pages, 0, create->u.symlink.len);
                break;
        case NF4BLK: case NF4CHR:
@@ -1160,7 +1160,7 @@ static int encode_rename(struct xdr_stream *xdr, const struct qstr *oldname, con
        return 0;
 }
-static int encode_renew(struct xdr_stream *xdr, const struct nfs4_client *client_stateid)
+static int encode_renew(struct xdr_stream *xdr, const struct nfs_client *client_stateid)
 {
        uint32_t *p;
@@ -1246,7 +1246,7 @@ static int encode_setclientid(struct xdr_stream *xdr, const struct nfs4_setclien
        return 0;
 }
-static int encode_setclientid_confirm(struct xdr_stream *xdr, const struct nfs4_client *client_state)
+static int encode_setclientid_confirm(struct xdr_stream *xdr, const struct nfs_client *client_state)
 {
        uint32_t *p;
@@ -1945,7 +1945,7 @@ static int nfs4_xdr_enc_server_caps(struct rpc_rqst *req, uint32_t *p, const str
 /*
 * a RENEW request
 */
-static int nfs4_xdr_enc_renew(struct rpc_rqst *req, uint32_t *p, struct nfs4_client *clp)
+static int nfs4_xdr_enc_renew(struct rpc_rqst *req, uint32_t *p, struct nfs_client *clp)
 {
        struct xdr_stream xdr;
        struct compound_hdr hdr = {
@@ -1975,7 +1975,7 @@ static int nfs4_xdr_enc_setclientid(struct rpc_rqst *req, uint32_t *p, struct nf
 /*
 * a SETCLIENTID_CONFIRM request
 */
-static int nfs4_xdr_enc_setclientid_confirm(struct rpc_rqst *req, uint32_t *p, struct nfs4_client *clp)
+static int nfs4_xdr_enc_setclientid_confirm(struct rpc_rqst *req, uint32_t *p, struct nfs_client *clp)
 {
        struct xdr_stream xdr;
        struct compound_hdr hdr = {
@@ -2127,12 +2127,12 @@ static int decode_op_hdr(struct xdr_stream *xdr, enum nfs_opnum4 expected)
        }
        READ32(nfserr);
        if (nfserr != NFS_OK)
-                return -nfs_stat_to_errno(nfserr);
+                return -nfs4_stat_to_errno(nfserr);
        return 0;
 }
 /* Dummy routine */
-static int decode_ace(struct xdr_stream *xdr, void *ace, struct nfs4_client *clp)
+static int decode_ace(struct xdr_stream *xdr, void *ace, struct nfs_client *clp)
 {
        uint32_t *p;
        unsigned int strlen;
@@ -2636,7 +2636,7 @@ static int decode_attr_nlink(struct xdr_stream *xdr, uint32_t *bitmap, uint32_t
        return 0;
 }
-static int decode_attr_owner(struct xdr_stream *xdr, uint32_t *bitmap, struct nfs4_client *clp, int32_t *uid)
+static int decode_attr_owner(struct xdr_stream *xdr, uint32_t *bitmap, struct nfs_client *clp, int32_t *uid)
 {
        uint32_t len, *p;
@@ -2660,7 +2660,7 @@ static int decode_attr_owner(struct xdr_stream *xdr, uint32_t *bitmap, struct nf
        return 0;
 }
-static int decode_attr_group(struct xdr_stream *xdr, uint32_t *bitmap, struct nfs4_client *clp, int32_t *gid)
+static int decode_attr_group(struct xdr_stream *xdr, uint32_t *bitmap, struct nfs_client *clp, int32_t *gid)
 {
        uint32_t len, *p;
@@ -3051,9 +3051,9 @@ static int decode_getfattr(struct xdr_stream *xdr, struct nfs_fattr *fattr, cons
        fattr->mode |= fmode;
        if ((status = decode_attr_nlink(xdr, bitmap, &fattr->nlink)) != 0)
                goto xdr_error;
-        if ((status = decode_attr_owner(xdr, bitmap, server->nfs4_state, &fattr->uid)) != 0)
+        if ((status = decode_attr_owner(xdr, bitmap, server->nfs_client, &fattr->uid)) != 0)
                goto xdr_error;
-        if ((status = decode_attr_group(xdr, bitmap, server->nfs4_state, &fattr->gid)) != 0)
+        if ((status = decode_attr_group(xdr, bitmap, server->nfs_client, &fattr->gid)) != 0)
                goto xdr_error;
        if ((status = decode_attr_rdev(xdr, bitmap, &fattr->rdev)) != 0)
                goto xdr_error;
@@ -3254,7 +3254,7 @@ static int decode_delegation(struct xdr_stream *xdr, struct nfs_openres *res)
                        if (decode_space_limit(xdr, &res->maxsize) < 0)
                                return -EIO;
        }
-        return decode_ace(xdr, NULL, res->server->nfs4_state);
+        return decode_ace(xdr, NULL, res->server->nfs_client);
 }
 static int decode_open(struct xdr_stream *xdr, struct nfs_openres *res)
@@ -3565,7 +3565,7 @@ static int decode_setattr(struct xdr_stream *xdr, struct nfs_setattrres *res)
        return 0;
 }
-static int decode_setclientid(struct xdr_stream *xdr, struct nfs4_client *clp)
+static int decode_setclientid(struct xdr_stream *xdr, struct nfs_client *clp)
 {
        uint32_t *p;
        uint32_t opnum;
@@ -3598,7 +3598,7 @@ static int decode_setclientid(struct xdr_stream *xdr, struct nfs4_client *clp)
                READ_BUF(len);
                return -NFSERR_CLID_INUSE;
        } else
-                return -nfs_stat_to_errno(nfserr);
+                return -nfs4_stat_to_errno(nfserr);
        return 0;
 }
@@ -4256,7 +4256,7 @@ static int nfs4_xdr_dec_fsinfo(struct rpc_rqst *req, uint32_t *p, struct nfs_fsi
        if (!status)
                status = decode_fsinfo(&xdr, fsinfo);
        if (!status)
-                status = -nfs_stat_to_errno(hdr.status);
+                status = -nfs4_stat_to_errno(hdr.status);
        return status;
 }
@@ -4335,7 +4335,7 @@ static int nfs4_xdr_dec_renew(struct rpc_rqst *rqstp, uint32_t *p, void *dummy)
 * a SETCLIENTID request
 */
 static int nfs4_xdr_dec_setclientid(struct rpc_rqst *req, uint32_t *p,
-                struct nfs4_client *clp)
+                struct nfs_client *clp)
 {
        struct xdr_stream xdr;
        struct compound_hdr hdr;
@@ -4346,7 +4346,7 @@ static int nfs4_xdr_dec_setclientid(struct rpc_rqst *req, uint32_t *p,
        if (!status)
                status = decode_setclientid(&xdr, clp);
        if (!status)
-                status = -nfs_stat_to_errno(hdr.status);
+                status = -nfs4_stat_to_errno(hdr.status);
        return status;
 }
@@ -4368,7 +4368,7 @@ static int nfs4_xdr_dec_setclientid_confirm(struct rpc_rqst *req, uint32_t *p, s
        if (!status)
                status = decode_fsinfo(&xdr, fsinfo);
        if (!status)
-                status = -nfs_stat_to_errno(hdr.status);
+                status = -nfs4_stat_to_errno(hdr.status);
        return status;
 }
@@ -4521,7 +4521,7 @@ static struct {
 * This one is used jointly by NFSv2 and NFSv3.
 */
 static int
-nfs_stat_to_errno(int stat)
+nfs4_stat_to_errno(int stat)
 {
        int i;
        for (i = 0; nfs_errtbl[i].stat != -1; i++) {
diff --git a/fs/nfs/proc.c b/fs/nfs/proc.c
index b3899ea3229e..630e50647bbb 100644
--- a/fs/nfs/proc.c
+++ b/fs/nfs/proc.c
@@ -66,14 +66,14 @@ nfs_proc_get_root(struct nfs_server *server, struct nfs_fh *fhandle,
        dprintk("%s: call getattr\n", __FUNCTION__);
        nfs_fattr_init(fattr);
-        status = rpc_call_sync(server->client_sys, &msg, 0);
+        status = rpc_call_sync(server->nfs_client->cl_rpcclient, &msg, 0);
        dprintk("%s: reply getattr: %d\n", __FUNCTION__, status);
        if (status)
                return status;
        dprintk("%s: call statfs\n", __FUNCTION__);
        msg.rpc_proc = &nfs_procedures[NFSPROC_STATFS];
        msg.rpc_resp = &fsinfo;
-        status = rpc_call_sync(server->client_sys, &msg, 0);
+        status = rpc_call_sync(server->nfs_client->cl_rpcclient, &msg, 0);
        dprintk("%s: reply statfs: %d\n", __FUNCTION__, status);
        if (status)
                return status;
@@ -425,16 +425,17 @@ nfs_proc_link(struct inode *inode, struct inode *dir, struct qstr *name)
 }
 static int
-nfs_proc_symlink(struct inode *dir, struct qstr *name, struct qstr *path,
+nfs_proc_symlink(struct inode *dir, struct dentry *dentry, struct page *page,
-                 struct iattr *sattr, struct nfs_fh *fhandle,
+                 unsigned int len, struct iattr *sattr)
-                 struct nfs_fattr *fattr)
 {
+        struct nfs_fh fhandle;
+        struct nfs_fattr fattr;
        struct nfs_symlinkargs  arg = {
                .fromfh         = NFS_FH(dir),
-                .fromname       = name->name,
+                .fromname       = dentry->d_name.name,
-                .fromlen        = name->len,
+                .fromlen        = dentry->d_name.len,
-                .topath         = path->name,
+                .pages          = &page,
-                .tolen          = path->len,
+                .pathlen        = len,
                .sattr          = sattr
        };
        struct rpc_message msg = {
@@ -443,13 +444,25 @@ nfs_proc_symlink(struct inode *dir, struct qstr *name, struct qstr *path,
        };
        int                     status;
-        if (path->len > NFS2_MAXPATHLEN)
+        if (len > NFS2_MAXPATHLEN)
                return -ENAMETOOLONG;
-        dprintk("NFS call  symlink %s -> %s\n", name->name, path->name);
-        nfs_fattr_init(fattr);
+        dprintk("NFS call  symlink %s\n", dentry->d_name.name);
-        fhandle->size = 0;
        status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0);
        nfs_mark_for_revalidate(dir);
+        /*
+         * V2 SYMLINK requests don't return any attributes.  Setting the
+         * filehandle size to zero indicates to nfs_instantiate that it
+         * should fill in the data with a LOOKUP call on the wire.
+         */
+        if (status == 0) {
+                nfs_fattr_init(&fattr);
+                fhandle.size = 0;
+                status = nfs_instantiate(dentry, &fhandle, &fattr);
+        }
        dprintk("NFS reply symlink: %d\n", status);
        return status;
 }
@@ -671,7 +684,7 @@ nfs_proc_lock(struct file *filp, int cmd, struct file_lock *fl)
 }
-struct nfs_rpc_ops      nfs_v2_clientops = {
+const struct nfs_rpc_ops nfs_v2_clientops = {
        .version        = 2,                   /* protocol version */
        .dentry_ops     = &nfs_dentry_operations,
        .dir_inode_ops  = &nfs_dir_inode_operations,
diff --git a/fs/nfs/read.c b/fs/nfs/read.c
index f0aff824a291..69f1549da2b9 100644
--- a/fs/nfs/read.c
+++ b/fs/nfs/read.c
@@ -171,7 +171,7 @@ static int nfs_readpage_sync(struct nfs_open_context *ctx, struct inode *inode,
                rdata->args.offset = page_offset(page) + rdata->args.pgbase;
                dprintk("NFS: nfs_proc_read(%s, (%s/%Ld), %Lu, %u)\n",
-                        NFS_SERVER(inode)->hostname,
+                        NFS_SERVER(inode)->nfs_client->cl_hostname,
                        inode->i_sb->s_id,
                        (long long)NFS_FILEID(inode),
                        (unsigned long long)rdata->args.pgbase,
@@ -568,8 +568,13 @@ int nfs_readpage_result(struct rpc_task *task, struct nfs_read_data *data)
        nfs_add_stats(data->inode, NFSIOS_SERVERREADBYTES, resp->count);
-        /* Is this a short read? */
+        if (task->tk_status < 0) {
-        if (task->tk_status >= 0 && resp->count < argp->count && !resp->eof) {
+                if (task->tk_status == -ESTALE) {
+                        set_bit(NFS_INO_STALE, &NFS_FLAGS(data->inode));
+                        nfs_mark_for_revalidate(data->inode);
+                }
+        } else if (resp->count < argp->count && !resp->eof) {
+                /* This is a short read! */
                nfs_inc_stats(data->inode, NFSIOS_SHORTREAD);
                /* Has the server at least made some progress? */
                if (resp->count != 0) {
@@ -616,6 +621,10 @@ int nfs_readpage(struct file *file, struct page *page)
        if (error)
                goto out_error;
+        error = -ESTALE;
+        if (NFS_STALE(inode))
+                goto out_error;
        if (file == NULL) {
                ctx = nfs_find_open_context(inode, NULL, FMODE_READ);
                if (ctx == NULL)
@@ -678,7 +687,7 @@ int nfs_readpages(struct file *filp, struct address_space *mapping,
        };
        struct inode *inode = mapping->host;
        struct nfs_server *server = NFS_SERVER(inode);
-        int ret;
+        int ret = -ESTALE;
        dprintk("NFS: nfs_readpages (%s/%Ld %d)\n",
                        inode->i_sb->s_id,
@@ -686,6 +695,9 @@ int nfs_readpages(struct file *filp, struct address_space *mapping,
                        nr_pages);
        nfs_inc_stats(inode, NFSIOS_VFSREADPAGES);
+        if (NFS_STALE(inode))
+                goto out;
        if (filp == NULL) {
                desc.ctx = nfs_find_open_context(inode, NULL, FMODE_READ);
                if (desc.ctx == NULL)
@@ -701,6 +713,7 @@ int nfs_readpages(struct file *filp, struct address_space *mapping,
                        ret = err;
        }
        put_nfs_open_context(desc.ctx);
+out:
        return ret;
 }
diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index e8a9bee74d9d..e8d40030cab4 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -13,6 +13,11 @@
 *
 *  Split from inode.c by David Howells <dhowells@redhat.com>
 *
+ * - superblocks are indexed on server only - all inodes, dentries, etc. associated with a
+ *   particular server are held in the same superblock
+ * - NFS superblocks can have several effective roots to the dentry tree
+ * - directory type roots are spliced into the tree when a path from one root reaches the root
+ *   of another (see nfs_lookup())
 */
 #include <linux/config.h>
@@ -52,66 +57,12 @@
 #define NFSDBG_FACILITY         NFSDBG_VFS
-/* 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)
-/*
- * RPC cruft for NFS
- */
-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
-};
-static struct rpc_program nfs_program = {
-        .name                   = "nfs",
-        .number                 = NFS_PROGRAM,
-        .nrvers                 = ARRAY_SIZE(nfs_version),
-        .version                = nfs_version,
-        .stats                  = &nfs_rpcstat,
-        .pipe_dir_name          = "/nfs",
-};
-struct rpc_stat nfs_rpcstat = {
-        .program                = &nfs_program
-};
-#ifdef CONFIG_NFS_V3_ACL
-static struct rpc_stat          nfsacl_rpcstat = { &nfsacl_program };
-static struct rpc_version *     nfsacl_version[] = {
-        [3]                     = &nfsacl_version3,
-};
-struct rpc_program              nfsacl_program = {
-        .name =                 "nfsacl",
-        .number =               NFS_ACL_PROGRAM,
-        .nrvers =               ARRAY_SIZE(nfsacl_version),
-        .version =              nfsacl_version,
-        .stats =                &nfsacl_rpcstat,
-};
-#endif  /* CONFIG_NFS_V3_ACL */
 static void nfs_umount_begin(struct vfsmount *, int);
 static int  nfs_statfs(struct dentry *, struct kstatfs *);
 static int  nfs_show_options(struct seq_file *, struct vfsmount *);
 static int  nfs_show_stats(struct seq_file *, struct vfsmount *);
 static int nfs_get_sb(struct file_system_type *, int, const char *, void *, struct vfsmount *);
-static int nfs_clone_nfs_sb(struct file_system_type *fs_type,
+static int nfs_xdev_get_sb(struct file_system_type *fs_type,
                int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt);
 static void nfs_kill_super(struct super_block *);
@@ -120,15 +71,15 @@ static struct file_system_type nfs_fs_type = {
        .name           = "nfs",
        .get_sb         = nfs_get_sb,
        .kill_sb        = nfs_kill_super,
-        .fs_flags       = FS_ODD_RENAME|FS_REVAL_DOT|FS_BINARY_MOUNTDATA,
+        .fs_flags       = FS_RENAME_DOES_D_MOVE|FS_REVAL_DOT|FS_BINARY_MOUNTDATA,
 };
-struct file_system_type clone_nfs_fs_type = {
+struct file_system_type nfs_xdev_fs_type = {
        .owner          = THIS_MODULE,
        .name           = "nfs",
-        .get_sb         = nfs_clone_nfs_sb,
+        .get_sb         = nfs_xdev_get_sb,
        .kill_sb        = nfs_kill_super,
-        .fs_flags       = FS_ODD_RENAME|FS_REVAL_DOT|FS_BINARY_MOUNTDATA,
+        .fs_flags       = FS_RENAME_DOES_D_MOVE|FS_REVAL_DOT|FS_BINARY_MOUNTDATA,
 };
 static struct super_operations nfs_sops = {
@@ -145,10 +96,10 @@ static struct super_operations nfs_sops = {
 #ifdef CONFIG_NFS_V4
 static int nfs4_get_sb(struct file_system_type *fs_type,
        int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt);
-static int nfs_clone_nfs4_sb(struct file_system_type *fs_type,
+static int nfs4_xdev_get_sb(struct file_system_type *fs_type,
-                int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt);
+        int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt);
-static int nfs_referral_nfs4_sb(struct file_system_type *fs_type,
+static int nfs4_referral_get_sb(struct file_system_type *fs_type,
-                int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt);
+        int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt);
 static void nfs4_kill_super(struct super_block *sb);
 static struct file_system_type nfs4_fs_type = {
@@ -156,23 +107,23 @@ static struct file_system_type nfs4_fs_type = {
        .name           = "nfs4",
        .get_sb         = nfs4_get_sb,
        .kill_sb        = nfs4_kill_super,
-        .fs_flags       = FS_ODD_RENAME|FS_REVAL_DOT|FS_BINARY_MOUNTDATA,
+        .fs_flags       = FS_RENAME_DOES_D_MOVE|FS_REVAL_DOT|FS_BINARY_MOUNTDATA,
 };
-struct file_system_type clone_nfs4_fs_type = {
+struct file_system_type nfs4_xdev_fs_type = {
        .owner          = THIS_MODULE,
        .name           = "nfs4",
-        .get_sb         = nfs_clone_nfs4_sb,
+        .get_sb         = nfs4_xdev_get_sb,
        .kill_sb        = nfs4_kill_super,
-        .fs_flags       = FS_ODD_RENAME|FS_REVAL_DOT|FS_BINARY_MOUNTDATA,
+        .fs_flags       = FS_RENAME_DOES_D_MOVE|FS_REVAL_DOT|FS_BINARY_MOUNTDATA,
 };
-struct file_system_type nfs_referral_nfs4_fs_type = {
+struct file_system_type nfs4_referral_fs_type = {
        .owner          = THIS_MODULE,
        .name           = "nfs4",
-        .get_sb         = nfs_referral_nfs4_sb,
+        .get_sb         = nfs4_referral_get_sb,
        .kill_sb        = nfs4_kill_super,
-        .fs_flags       = FS_ODD_RENAME|FS_REVAL_DOT|FS_BINARY_MOUNTDATA,
+        .fs_flags       = FS_RENAME_DOES_D_MOVE|FS_REVAL_DOT|FS_BINARY_MOUNTDATA,
 };
 static struct super_operations nfs4_sops = {
@@ -187,39 +138,7 @@ static struct super_operations nfs4_sops = {
 };
 #endif
-#ifdef CONFIG_NFS_V4
+static struct shrinker *acl_shrinker;
-static const int nfs_set_port_min = 0;
-static const int nfs_set_port_max = 65535;
-static int param_set_port(const char *val, struct kernel_param *kp)
-{
-        char *endp;
-        int num = simple_strtol(val, &endp, 0);
-        if (endp == val || *endp || num < nfs_set_port_min || num > nfs_set_port_max)
-                return -EINVAL;
-        *((int *)kp->arg) = num;
-        return 0;
-}
-module_param_call(callback_tcpport, param_set_port, param_get_int,
-                 &nfs_callback_set_tcpport, 0644);
-#endif
-#ifdef CONFIG_NFS_V4
-static int param_set_idmap_timeout(const char *val, struct kernel_param *kp)
-{
-        char *endp;
-        int num = simple_strtol(val, &endp, 0);
-        int jif = num * HZ;
-        if (endp == val || *endp || num < 0 || jif < num)
-                return -EINVAL;
-        *((int *)kp->arg) = jif;
-        return 0;
-}
-module_param_call(idmap_cache_timeout, param_set_idmap_timeout, param_get_int,
-                 &nfs_idmap_cache_timeout, 0644);
-#endif
 /*
 * Register the NFS filesystems
@@ -240,6 +159,7 @@ int __init register_nfs_fs(void)
        if (ret < 0)
                goto error_2;
 #endif
+        acl_shrinker = set_shrinker(DEFAULT_SEEKS, nfs_access_cache_shrinker);
        return 0;
 #ifdef CONFIG_NFS_V4
@@ -257,6 +177,8 @@ error_0:
 */
 void __exit unregister_nfs_fs(void)
 {
+        if (acl_shrinker != NULL)
+                remove_shrinker(acl_shrinker);
 #ifdef CONFIG_NFS_V4
        unregister_filesystem(&nfs4_fs_type);
        nfs_unregister_sysctl();
@@ -269,11 +191,10 @@ void __exit unregister_nfs_fs(void)
 */
 static int nfs_statfs(struct dentry *dentry, struct kstatfs *buf)
 {
-        struct super_block *sb = dentry->d_sb;
+        struct nfs_server *server = NFS_SB(dentry->d_sb);
-        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_fh *fh = NFS_FH(dentry->d_inode);
        struct nfs_fattr fattr;
        struct nfs_fsstat res = {
                        .fattr = &fattr,
@@ -282,7 +203,7 @@ static int nfs_statfs(struct dentry *dentry, struct kstatfs *buf)
        lock_kernel();
-        error = server->rpc_ops->statfs(server, rootfh, &res);
+        error = server->nfs_client->rpc_ops->statfs(server, fh, &res);
        buf->f_type = NFS_SUPER_MAGIC;
        if (error < 0)
                goto out_err;
@@ -292,7 +213,7 @@ static int nfs_statfs(struct dentry *dentry, struct kstatfs *buf)
         * case where f_frsize != f_bsize.  Eventually we want to
         * report the value of wtmult in this field.
         */
-        buf->f_frsize = sb->s_blocksize;
+        buf->f_frsize = dentry->d_sb->s_blocksize;
        /*
         * On most *nix systems, f_blocks, f_bfree, and f_bavail
@@ -301,8 +222,8 @@ static int nfs_statfs(struct dentry *dentry, struct kstatfs *buf)
         * thus historically Linux's sys_statfs reports these
         * fields in units of f_bsize.
         */
-        buf->f_bsize = sb->s_blocksize;
+        buf->f_bsize = dentry->d_sb->s_blocksize;
-        blockbits = sb->s_blocksize_bits;
+        blockbits = dentry->d_sb->s_blocksize_bits;
        blockres = (1 << blockbits) - 1;
        buf->f_blocks = (res.tbytes + blockres) >> blockbits;
        buf->f_bfree = (res.fbytes + blockres) >> blockbits;
@@ -323,9 +244,12 @@ static int nfs_statfs(struct dentry *dentry, struct kstatfs *buf)
 }
+/*
+ * Map the security flavour number to a name
+ */
 static const char *nfs_pseudoflavour_to_name(rpc_authflavor_t flavour)
 {
-        static struct {
+        static const struct {
                rpc_authflavor_t flavour;
                const char *str;
        } sec_flavours[] = {
@@ -356,10 +280,10 @@ static const char *nfs_pseudoflavour_to_name(rpc_authflavor_t flavour)
 */
 static void nfs_show_mount_options(struct seq_file *m, struct nfs_server *nfss, int showdefaults)
 {
-        static struct proc_nfs_info {
+        static const struct proc_nfs_info {
                int flag;
-                char *str;
+                const char *str;
-                char *nostr;
+                const char *nostr;
        } nfs_info[] = {
                { NFS_MOUNT_SOFT, ",soft", ",hard" },
                { NFS_MOUNT_INTR, ",intr", "" },
@@ -369,11 +293,12 @@ static void nfs_show_mount_options(struct seq_file *m, struct nfs_server *nfss,
                { NFS_MOUNT_NOACL, ",noacl", "" },
                { 0, NULL, NULL }
        };
-        struct proc_nfs_info *nfs_infop;
+        const struct proc_nfs_info *nfs_infop;
+        struct nfs_client *clp = nfss->nfs_client;
        char buf[12];
-        char *proto;
+        const char *proto;
-        seq_printf(m, ",vers=%d", nfss->rpc_ops->version);
+        seq_printf(m, ",vers=%d", clp->rpc_ops->version);
        seq_printf(m, ",rsize=%d", nfss->rsize);
        seq_printf(m, ",wsize=%d", nfss->wsize);
        if (nfss->acregmin != 3*HZ || showdefaults)
@@ -402,8 +327,8 @@ static void nfs_show_mount_options(struct seq_file *m, struct nfs_server *nfss,
                        proto = buf;
        }
        seq_printf(m, ",proto=%s", proto);
-        seq_printf(m, ",timeo=%lu", 10U * nfss->retrans_timeo / HZ);
+        seq_printf(m, ",timeo=%lu", 10U * clp->retrans_timeo / HZ);
-        seq_printf(m, ",retrans=%u", nfss->retrans_count);
+        seq_printf(m, ",retrans=%u", clp->retrans_count);
        seq_printf(m, ",sec=%s", nfs_pseudoflavour_to_name(nfss->client->cl_auth->au_flavor));
 }
@@ -417,7 +342,7 @@ static int nfs_show_options(struct seq_file *m, struct vfsmount *mnt)
        nfs_show_mount_options(m, nfss, 0);
        seq_puts(m, ",addr=");
-        seq_escape(m, nfss->hostname, " \t\n\\");
+        seq_escape(m, nfss->nfs_client->cl_hostname, " \t\n\\");
        return 0;
 }
@@ -454,7 +379,7 @@ static int nfs_show_stats(struct seq_file *m, struct vfsmount *mnt)
        seq_printf(m, ",namelen=%d", nfss->namelen);
 #ifdef CONFIG_NFS_V4
-        if (nfss->rpc_ops->version == 4) {
+        if (nfss->nfs_client->cl_nfsversion == 4) {
                seq_printf(m, "\n\tnfsv4:\t");
                seq_printf(m, "bm0=0x%x", nfss->attr_bitmask[0]);
                seq_printf(m, ",bm1=0x%x", nfss->attr_bitmask[1]);
@@ -501,782 +426,353 @@ static int nfs_show_stats(struct seq_file *m, struct vfsmount *mnt)
 /*
 * Begin unmount by attempting to remove all automounted mountpoints we added
- * in response to traversals
+ * in response to xdev traversals and referrals
 */
 static void nfs_umount_begin(struct vfsmount *vfsmnt, int flags)
 {
-        struct nfs_server *server;
-        struct rpc_clnt *rpc;
        shrink_submounts(vfsmnt, &nfs_automount_list);
-        if (!(flags & MNT_FORCE))
-                return;
-        /* -EIO all pending I/O */
-        server = NFS_SB(vfsmnt->mnt_sb);
-        rpc = server->client;
-        if (!IS_ERR(rpc))
-                rpc_killall_tasks(rpc);
-        rpc = server->client_acl;
-        if (!IS_ERR(rpc))
-                rpc_killall_tasks(rpc);
 }
 /*
- * Obtain the root inode of the file system.
+ * Validate the NFS2/NFS3 mount data
+ * - fills in the mount root filehandle
 */
-static struct inode *
+static int nfs_validate_mount_data(struct nfs_mount_data *data,
-nfs_get_root(struct super_block *sb, struct nfs_fh *rootfh, struct nfs_fsinfo *fsinfo)
+                                   struct nfs_fh *mntfh)
 {
-        struct nfs_server       *server = NFS_SB(sb);
+        if (data == NULL) {
-        int                     error;
+                dprintk("%s: missing data argument\n", __FUNCTION__);
+                return -EINVAL;
-        error = server->rpc_ops->getroot(server, rootfh, fsinfo);
-        if (error < 0) {
-                dprintk("nfs_get_root: getattr error = %d\n", -error);
-                return ERR_PTR(error);
        }
-        server->fsid = fsinfo->fattr->fsid;
+        if (data->version <= 0 || data->version > NFS_MOUNT_VERSION) {
-        return nfs_fhget(sb, rootfh, fsinfo->fattr);
+                dprintk("%s: bad mount version\n", __FUNCTION__);
-}
+                return -EINVAL;
+        }
-/*
- * 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;
-        unsigned long max_rpc_payload;
-        /* 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;
+        switch (data->version) {
+                case 1:
+                        data->namlen = 0;
+                case 2:
+                        data->bsize  = 0;
+                case 3:
+                        if (data->flags & NFS_MOUNT_VER3) {
+                                dprintk("%s: mount structure version %d does not support NFSv3\n",
+                                                __FUNCTION__,
+                                                data->version);
+                                return -EINVAL;
+                        }
+                        data->root.size = NFS2_FHSIZE;
+                        memcpy(data->root.data, data->old_root.data, NFS2_FHSIZE);
+                case 4:
+                        if (data->flags & NFS_MOUNT_SECFLAVOUR) {
+                                dprintk("%s: mount structure version %d does not support strong security\n",
+                                                __FUNCTION__,
+                                                data->version);
+                                return -EINVAL;
+                        }
+                case 5:
+                        memset(data->context, 0, sizeof(data->context));
+        }
-        server->io_stats = nfs_alloc_iostats();
+        /* Set the pseudoflavor */
-        if (server->io_stats == NULL)
+        if (!(data->flags & NFS_MOUNT_SECFLAVOUR))
-                return -ENOMEM;
+                data->pseudoflavor = RPC_AUTH_UNIX;
-        root_inode = nfs_get_root(sb, &server->fh, &fsinfo);
+#ifndef CONFIG_NFS_V3
-        /* Did getting the root inode fail? */
+        /* If NFSv3 is not compiled in, return -EPROTONOSUPPORT */
-        if (IS_ERR(root_inode)) {
+        if (data->flags & NFS_MOUNT_VER3) {
-                no_root_error = PTR_ERR(root_inode);
+                dprintk("%s: NFSv3 not compiled into kernel\n", __FUNCTION__);
-                goto out_no_root;
+                return -EPROTONOSUPPORT;
-        }
-        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;
+#endif /* CONFIG_NFS_V3 */
-        /* mount time stamp, in seconds */
-        server->mount_time = jiffies;
-        /* 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);
-        max_rpc_payload = nfs_block_size(rpc_max_payload(server->client), NULL);
-        if (server->rsize > max_rpc_payload)
-                server->rsize = max_rpc_payload;
-        if (server->rsize > NFS_MAX_FILE_IO_SIZE)
-                server->rsize = NFS_MAX_FILE_IO_SIZE;
-        server->rpages = (server->rsize + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
-        if (server->wsize > max_rpc_payload)
-                server->wsize = max_rpc_payload;
-        if (server->wsize > NFS_MAX_FILE_IO_SIZE)
-                server->wsize = NFS_MAX_FILE_IO_SIZE;
-        server->wpages = (server->wsize + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
-        if (sb->s_blocksize == 0)
+        /* We now require that the mount process passes the remote address */
-                sb->s_blocksize = nfs_block_bits(server->wsize,
+        if (data->addr.sin_addr.s_addr == INADDR_ANY) {
-                                                         &sb->s_blocksize_bits);
+                dprintk("%s: mount program didn't pass remote address!\n",
-        server->wtmult = nfs_block_bits(fsinfo.wtmult, NULL);
+                        __FUNCTION__);
+                return -EINVAL;
-        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;
-        nfs_super_set_maxbytes(sb, fsinfo.maxfilesize);
+        /* Prepare the root filehandle */
+        if (data->flags & NFS_MOUNT_VER3)
+                mntfh->size = data->root.size;
+        else
+                mntfh->size = NFS2_FHSIZE;
+        if (mntfh->size > sizeof(mntfh->data)) {
+                dprintk("%s: invalid root filehandle\n", __FUNCTION__);
+                return -EINVAL;
+        }
-        server->client->cl_intr = (server->flags & NFS_MOUNT_INTR) ? 1 : 0;
+        memcpy(mntfh->data, data->root.data, mntfh->size);
-        server->client->cl_softrtry = (server->flags & NFS_MOUNT_SOFT) ? 1 : 0;
+        if (mntfh->size < sizeof(mntfh->data))
+                memset(mntfh->data + mntfh->size, 0,
+                       sizeof(mntfh->data) - mntfh->size);
-        /* 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;
 }
 /*
- * Initialise the timeout values for a connection
+ * Initialise the common bits of the superblock
 */
-static void nfs_init_timeout_values(struct rpc_timeout *to, int proto, unsigned int timeo, unsigned int retrans)
+static inline void nfs_initialise_sb(struct super_block *sb)
 {
-        to->to_initval = timeo * HZ / 10;
+        struct nfs_server *server = NFS_SB(sb);
-        to->to_retries = retrans;
-        if (!to->to_retries)
-                to->to_retries = 2;
-        switch (proto) {
-        case IPPROTO_TCP:
-                if (!to->to_initval)
-                        to->to_initval = 60 * HZ;
-                if (to->to_initval > NFS_MAX_TCP_TIMEOUT)
-                        to->to_initval = NFS_MAX_TCP_TIMEOUT;
-                to->to_increment = to->to_initval;
-                to->to_maxval = to->to_initval + (to->to_increment * to->to_retries);
-                to->to_exponential = 0;
-                break;
-        case IPPROTO_UDP:
-        default:
-                if (!to->to_initval)
-                        to->to_initval = 11 * HZ / 10;
-                if (to->to_initval > NFS_MAX_UDP_TIMEOUT)
-                        to->to_initval = NFS_MAX_UDP_TIMEOUT;
-                to->to_maxval = NFS_MAX_UDP_TIMEOUT;
-                to->to_exponential = 1;
-                break;
-        }
-}
-/*
+        sb->s_magic = NFS_SUPER_MAGIC;
- * Create an RPC client handle.
- */
-static struct rpc_clnt *
-nfs_create_client(struct nfs_server *server, const struct nfs_mount_data *data)
-{
-        struct rpc_timeout      timeparms;
-        struct rpc_xprt         *xprt = NULL;
-        struct rpc_clnt         *clnt = NULL;
-        int                     proto = (data->flags & NFS_MOUNT_TCP) ? IPPROTO_TCP : IPPROTO_UDP;
-        nfs_init_timeout_values(&timeparms, proto, data->timeo, data->retrans);
-        server->retrans_timeo = timeparms.to_initval;
-        server->retrans_count = timeparms.to_retries;
-        /* create transport and client */
-        xprt = xprt_create_proto(proto, &server->addr, &timeparms);
-        if (IS_ERR(xprt)) {
-                dprintk("%s: cannot create RPC transport. Error = %ld\n",
-                                __FUNCTION__, PTR_ERR(xprt));
-                return (struct rpc_clnt *)xprt;
-        }
-        clnt = rpc_create_client(xprt, server->hostname, &nfs_program,
-                                 server->rpc_ops->version, data->pseudoflavor);
-        if (IS_ERR(clnt)) {
-                dprintk("%s: cannot create RPC client. Error = %ld\n",
-                                __FUNCTION__, PTR_ERR(xprt));
-                goto out_fail;
-        }
-        clnt->cl_intr     = 1;
+        /* We probably want something more informative here */
-        clnt->cl_softrtry = 1;
+        snprintf(sb->s_id, sizeof(sb->s_id),
+                 "%x:%x", MAJOR(sb->s_dev), MINOR(sb->s_dev));
+        if (sb->s_blocksize == 0)
+                sb->s_blocksize = nfs_block_bits(server->wsize,
+                                                 &sb->s_blocksize_bits);
-        return clnt;
+        if (server->flags & NFS_MOUNT_NOAC)
+                sb->s_flags |= MS_SYNCHRONOUS;
-out_fail:
+        nfs_super_set_maxbytes(sb, server->maxfilesize);
-        return clnt;
 }
 /*
- * Clone a server record
+ * Finish setting up an NFS2/3 superblock
 */
-static struct nfs_server *nfs_clone_server(struct super_block *sb, struct nfs_clone_mount *data)
+static void nfs_fill_super(struct super_block *sb, struct nfs_mount_data *data)
 {
        struct nfs_server *server = NFS_SB(sb);
-        struct nfs_server *parent = NFS_SB(data->sb);
-        struct inode *root_inode;
-        struct nfs_fsinfo fsinfo;
-        void *err = ERR_PTR(-ENOMEM);
-        sb->s_op = data->sb->s_op;
-        sb->s_blocksize = data->sb->s_blocksize;
-        sb->s_blocksize_bits = data->sb->s_blocksize_bits;
-        sb->s_maxbytes = data->sb->s_maxbytes;
-        server->client_sys = server->client_acl = ERR_PTR(-EINVAL);
-        server->io_stats = nfs_alloc_iostats();
-        if (server->io_stats == NULL)
-                goto out;
-        server->client = rpc_clone_client(parent->client);
-        if (IS_ERR((err = server->client)))
-                goto out;
-        if (!IS_ERR(parent->client_sys)) {
-                server->client_sys = rpc_clone_client(parent->client_sys);
-                if (IS_ERR((err = server->client_sys)))
-                        goto out;
-        }
-        if (!IS_ERR(parent->client_acl)) {
-                server->client_acl = rpc_clone_client(parent->client_acl);
-                if (IS_ERR((err = server->client_acl)))
-                        goto out;
-        }
-        root_inode = nfs_fhget(sb, data->fh, data->fattr);
-        if (!root_inode)
-                goto out;
-        sb->s_root = d_alloc_root(root_inode);
-        if (!sb->s_root)
-                goto out_put_root;
-        fsinfo.fattr = data->fattr;
-        if (NFS_PROTO(root_inode)->fsinfo(server, data->fh, &fsinfo) == 0)
-                nfs_super_set_maxbytes(sb, fsinfo.maxfilesize);
-        sb->s_root->d_op = server->rpc_ops->dentry_ops;
-        sb->s_flags |= MS_ACTIVE;
-        return server;
-out_put_root:
-        iput(root_inode);
-out:
-        return err;
-}
-/*
+        sb->s_blocksize_bits = 0;
- * Copy an existing superblock and attach revised data
+        sb->s_blocksize = 0;
- */
+        if (data->bsize)
-static int nfs_clone_generic_sb(struct nfs_clone_mount *data,
+                sb->s_blocksize = nfs_block_size(data->bsize, &sb->s_blocksize_bits);
-                struct super_block *(*fill_sb)(struct nfs_server *, struct nfs_clone_mount *),
-                struct nfs_server *(*fill_server)(struct super_block *, struct nfs_clone_mount *),
-                struct vfsmount *mnt)
-{
-        struct nfs_server *server;
-        struct nfs_server *parent = NFS_SB(data->sb);
-        struct super_block *sb = ERR_PTR(-EINVAL);
-        char *hostname;
-        int error = -ENOMEM;
-        int len;
-        server = kmalloc(sizeof(struct nfs_server), GFP_KERNEL);
-        if (server == NULL)
-                goto out_err;
-        memcpy(server, parent, sizeof(*server));
-        hostname = (data->hostname != NULL) ? data->hostname : parent->hostname;
-        len = strlen(hostname) + 1;
-        server->hostname = kmalloc(len, GFP_KERNEL);
-        if (server->hostname == NULL)
-                goto free_server;
-        memcpy(server->hostname, hostname, len);
-        error = rpciod_up();
-        if (error != 0)
-                goto free_hostname;
-        sb = fill_sb(server, data);
-        if (IS_ERR(sb)) {
-                error = PTR_ERR(sb);
-                goto kill_rpciod;
-        }
-                
-        if (sb->s_root)
-                goto out_rpciod_down;
-        server = fill_server(sb, data);
+        if (server->flags & NFS_MOUNT_VER3) {
-        if (IS_ERR(server)) {
+                /* The VFS shouldn't apply the umask to mode bits. We will do
-                error = PTR_ERR(server);
+                 * so ourselves when necessary.
-                goto out_deactivate;
+                 */
+                sb->s_flags |= MS_POSIXACL;
+                sb->s_time_gran = 1;
        }
-        return simple_set_mnt(mnt, sb);
-out_deactivate:
+        sb->s_op = &nfs_sops;
-        up_write(&sb->s_umount);
+        nfs_initialise_sb(sb);
-        deactivate_super(sb);
-        return error;
-out_rpciod_down:
-        rpciod_down();
-        kfree(server->hostname);
-        kfree(server);
-        return simple_set_mnt(mnt, sb);
-kill_rpciod:
-        rpciod_down();
-free_hostname:
-        kfree(server->hostname);
-free_server:
-        kfree(server);
-out_err:
-        return error;
 }
 /*
- * Set up an NFS2/3 superblock
+ * Finish setting up a cloned NFS2/3 superblock
- *
- * 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.
 */
-static int
+static void nfs_clone_super(struct super_block *sb,
-nfs_fill_super(struct super_block *sb, struct nfs_mount_data *data, int silent)
+                            const struct super_block *old_sb)
 {
-        struct nfs_server       *server;
+        struct nfs_server *server = NFS_SB(sb);
-        rpc_authflavor_t        authflavor;
-        server           = NFS_SB(sb);
+        sb->s_blocksize_bits = old_sb->s_blocksize_bits;
-        sb->s_blocksize_bits = 0;
+        sb->s_blocksize = old_sb->s_blocksize;
-        sb->s_blocksize = 0;
+        sb->s_maxbytes = old_sb->s_maxbytes;
-        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. */
-#ifdef CONFIG_NFS_V3
-        if (server->flags & NFS_MOUNT_VER3) {
-                server->rpc_ops = &nfs_v3_clientops;
-                server->caps |= NFS_CAP_READDIRPLUS;
-        } else {
-                server->rpc_ops = &nfs_v2_clientops;
-        }
-#else
-        server->rpc_ops = &nfs_v2_clientops;
-#endif
-        /* 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) {
-                struct rpc_auth *auth;
-                server->client_sys = rpc_clone_client(server->client);
-                if (IS_ERR(server->client_sys))
-                        return PTR_ERR(server->client_sys);
-                auth = rpcauth_create(RPC_AUTH_UNIX, server->client_sys);
-                if (IS_ERR(auth))
-                        return PTR_ERR(auth);
-        } else {
-                atomic_inc(&server->client->cl_count);
-                server->client_sys = server->client;
-        }
        if (server->flags & NFS_MOUNT_VER3) {
-#ifdef CONFIG_NFS_V3_ACL
+                /* The VFS shouldn't apply the umask to mode bits. We will do
-                if (!(server->flags & NFS_MOUNT_NOACL)) {
+                 * so ourselves when necessary.
-                        server->client_acl = rpc_bind_new_program(server->client, &nfsacl_program, 3);
-                        /* No errors! Assume that Sun nfsacls are supported */
-                        if (!IS_ERR(server->client_acl))
-                                server->caps |= NFS_CAP_ACLS;
-                }
-#else
-                server->flags &= ~NFS_MOUNT_NOACL;
-#endif /* CONFIG_NFS_V3_ACL */
-                /*
-                 * The VFS shouldn't apply the umask to mode bits. We will
-                 * do so ourselves when necessary.
                 */
                sb->s_flags |= MS_POSIXACL;
-                if (server->namelen == 0 || server->namelen > NFS3_MAXNAMLEN)
-                        server->namelen = NFS3_MAXNAMLEN;
                sb->s_time_gran = 1;
-        } else {
-                if (server->namelen == 0 || server->namelen > NFS2_MAXNAMLEN)
-                        server->namelen = NFS2_MAXNAMLEN;
        }
-        sb->s_op = &nfs_sops;
+        sb->s_op = old_sb->s_op;
-        return nfs_sb_init(sb, authflavor);
+        nfs_initialise_sb(sb);
 }
-static int nfs_set_super(struct super_block *s, void *data)
+static int nfs_set_super(struct super_block *s, void *_server)
 {
-        s->s_fs_info = data;
+        struct nfs_server *server = _server;
-        return set_anon_super(s, data);
+        int ret;
+        s->s_fs_info = server;
+        ret = set_anon_super(s, server);
+        if (ret == 0)
+                server->s_dev = s->s_dev;
+        return ret;
 }
 static int nfs_compare_super(struct super_block *sb, void *data)
 {
-        struct nfs_server *server = data;
+        struct nfs_server *server = data, *old = NFS_SB(sb);
-        struct nfs_server *old = NFS_SB(sb);
-        if (old->addr.sin_addr.s_addr != server->addr.sin_addr.s_addr)
+        if (old->nfs_client != server->nfs_client)
                return 0;
-        if (old->addr.sin_port != server->addr.sin_port)
+        if (memcmp(&old->fsid, &server->fsid, sizeof(old->fsid)) != 0)
                return 0;
-        return !nfs_compare_fh(&old->fh, &server->fh);
+        return 1;
 }
 static int nfs_get_sb(struct file_system_type *fs_type,
        int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt)
 {
-        int error;
        struct nfs_server *server = NULL;
        struct super_block *s;
-        struct nfs_fh *root;
+        struct nfs_fh mntfh;
        struct nfs_mount_data *data = raw_data;
+        struct dentry *mntroot;
+        int error;
-        error = -EINVAL;
+        /* Validate the mount data */
-        if (data == NULL) {
+        error = nfs_validate_mount_data(data, &mntfh);
-                dprintk("%s: missing data argument\n", __FUNCTION__);
+        if (error < 0)
-                goto out_err_noserver;
+                return error;
-        }
-        if (data->version <= 0 || data->version > NFS_MOUNT_VERSION) {
-                dprintk("%s: bad mount version\n", __FUNCTION__);
-                goto out_err_noserver;
-        }
-        switch (data->version) {
-                case 1:
-                        data->namlen = 0;
-                case 2:
-                        data->bsize  = 0;
-                case 3:
-                        if (data->flags & NFS_MOUNT_VER3) {
-                                dprintk("%s: mount structure version %d does not support NFSv3\n",
-                                                __FUNCTION__,
-                                                data->version);
-                                goto out_err_noserver;
-                        }
-                        data->root.size = NFS2_FHSIZE;
-                        memcpy(data->root.data, data->old_root.data, NFS2_FHSIZE);
-                case 4:
-                        if (data->flags & NFS_MOUNT_SECFLAVOUR) {
-                                dprintk("%s: mount structure version %d does not support strong security\n",
-                                                __FUNCTION__,
-                                                data->version);
-                                goto out_err_noserver;
-                        }
-                case 5:
-                        memset(data->context, 0, sizeof(data->context));
-        }
-#ifndef CONFIG_NFS_V3
-        /* If NFSv3 is not compiled in, return -EPROTONOSUPPORT */
-        error = -EPROTONOSUPPORT;
-        if (data->flags & NFS_MOUNT_VER3) {
-                dprintk("%s: NFSv3 not compiled into kernel\n", __FUNCTION__);
-                goto out_err_noserver;
-        }
-#endif /* CONFIG_NFS_V3 */
-        error = -ENOMEM;
+        /* Get a volume representation */
-        server = kzalloc(sizeof(struct nfs_server), GFP_KERNEL);
+        server = nfs_create_server(data, &mntfh);
-        if (!server)
+        if (IS_ERR(server)) {
+                error = PTR_ERR(server);
                goto out_err_noserver;
-        /* Zero out the NFS state stuff */
-        init_nfsv4_state(server);
-        server->client = server->client_sys = server->client_acl = ERR_PTR(-EINVAL);
-        root = &server->fh;
-        if (data->flags & NFS_MOUNT_VER3)
-                root->size = data->root.size;
-        else
-                root->size = NFS2_FHSIZE;
-        error = -EINVAL;
-        if (root->size > sizeof(root->data)) {
-                dprintk("%s: invalid root filehandle\n", __FUNCTION__);
-                goto out_err;
-        }
-        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) {
-                dprintk("%s: mount program didn't pass remote address!\n",
-                                __FUNCTION__);
-                goto out_err;
-        }
-        /* Fire up rpciod if not yet running */
-        error = rpciod_up();
-        if (error < 0) {
-                dprintk("%s: couldn't start rpciod! Error = %d\n",
-                                __FUNCTION__, error);
-                goto out_err;
        }
+        /* Get a superblock - note that we may end up sharing one that already exists */
        s = sget(fs_type, nfs_compare_super, nfs_set_super, server);
        if (IS_ERR(s)) {
                error = PTR_ERR(s);
-                goto out_err_rpciod;
+                goto out_err_nosb;
        }
-        if (s->s_root)
+        if (s->s_fs_info != server) {
-                goto out_rpciod_down;
+                nfs_free_server(server);
+                server = NULL;
+        }
-        s->s_flags = flags;
+        if (!s->s_root) {
+                /* initial superblock/root creation */
+                s->s_flags = flags;
+                nfs_fill_super(s, data);
+        }
-        error = nfs_fill_super(s, data, flags & MS_SILENT ? 1 : 0);
+        mntroot = nfs_get_root(s, &mntfh);
-        if (error) {
+        if (IS_ERR(mntroot)) {
-                up_write(&s->s_umount);
+                error = PTR_ERR(mntroot);
-                deactivate_super(s);
+                goto error_splat_super;
-                return error;
        }
-        s->s_flags |= MS_ACTIVE;
-        return simple_set_mnt(mnt, s);
-out_rpciod_down:
+        s->s_flags |= MS_ACTIVE;
-        rpciod_down();
+        mnt->mnt_sb = s;
-        kfree(server);
+        mnt->mnt_root = mntroot;
-        return simple_set_mnt(mnt, s);
+        return 0;
-out_err_rpciod:
+out_err_nosb:
-        rpciod_down();
+        nfs_free_server(server);
-out_err:
-        kfree(server);
 out_err_noserver:
        return error;
+error_splat_super:
+        up_write(&s->s_umount);
+        deactivate_super(s);
+        return error;
 }
+/*
+ * Destroy an NFS2/3 superblock
+ */
 static void nfs_kill_super(struct super_block *s)
 {
        struct nfs_server *server = NFS_SB(s);
        kill_anon_super(s);
+        nfs_free_server(server);
-        if (!IS_ERR(server->client))
-                rpc_shutdown_client(server->client);
-        if (!IS_ERR(server->client_sys))
-                rpc_shutdown_client(server->client_sys);
-        if (!IS_ERR(server->client_acl))
-                rpc_shutdown_client(server->client_acl);
-        if (!(server->flags & NFS_MOUNT_NONLM))
-                lockd_down();   /* release rpc.lockd */
-        rpciod_down();          /* release rpciod */
-        nfs_free_iostats(server->io_stats);
-        kfree(server->hostname);
-        kfree(server);
-        nfs_release_automount_timer();
-}
-static struct super_block *nfs_clone_sb(struct nfs_server *server, struct nfs_clone_mount *data)
-{
-        struct super_block *sb;
-        server->fsid = data->fattr->fsid;
-        nfs_copy_fh(&server->fh, data->fh);
-        sb = sget(&nfs_fs_type, nfs_compare_super, nfs_set_super, server);
-        if (!IS_ERR(sb) && sb->s_root == NULL && !(server->flags & NFS_MOUNT_NONLM))
-                lockd_up();
-        return sb;
 }
-static int nfs_clone_nfs_sb(struct file_system_type *fs_type,
+/*
-                int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt)
+ * Clone an NFS2/3 server record on xdev traversal (FSID-change)
+ */
+static int nfs_xdev_get_sb(struct file_system_type *fs_type, int flags,
+                           const char *dev_name, void *raw_data,
+                           struct vfsmount *mnt)
 {
        struct nfs_clone_mount *data = raw_data;
-        return nfs_clone_generic_sb(data, nfs_clone_sb, nfs_clone_server, mnt);
+        struct super_block *s;
-}
+        struct nfs_server *server;
+        struct dentry *mntroot;
+        int error;
-#ifdef CONFIG_NFS_V4
+        dprintk("--> nfs_xdev_get_sb()\n");
-static struct rpc_clnt *nfs4_create_client(struct nfs_server *server,
-        struct rpc_timeout *timeparms, int proto, rpc_authflavor_t flavor)
-{
-        struct nfs4_client *clp;
-        struct rpc_xprt *xprt = NULL;
-        struct rpc_clnt *clnt = NULL;
-        int err = -EIO;
-        clp = nfs4_get_client(&server->addr.sin_addr);
-        if (!clp) {
-                dprintk("%s: failed to create NFS4 client.\n", __FUNCTION__);
-                return ERR_PTR(err);
-        }
-        /* Now create transport and client */
+        /* create a new volume representation */
-        down_write(&clp->cl_sem);
+        server = nfs_clone_server(NFS_SB(data->sb), data->fh, data->fattr);
-        if (IS_ERR(clp->cl_rpcclient)) {
+        if (IS_ERR(server)) {
-                xprt = xprt_create_proto(proto, &server->addr, timeparms);
+                error = PTR_ERR(server);
-                if (IS_ERR(xprt)) {
+                goto out_err_noserver;
-                        up_write(&clp->cl_sem);
-                        err = PTR_ERR(xprt);
-                        dprintk("%s: cannot create RPC transport. Error = %d\n",
-                                        __FUNCTION__, err);
-                        goto out_fail;
-                }
-                /* Bind to a reserved port! */
-                xprt->resvport = 1;
-                clnt = rpc_create_client(xprt, server->hostname, &nfs_program,
-                                server->rpc_ops->version, flavor);
-                if (IS_ERR(clnt)) {
-                        up_write(&clp->cl_sem);
-                        err = PTR_ERR(clnt);
-                        dprintk("%s: cannot create RPC client. Error = %d\n",
-                                        __FUNCTION__, err);
-                        goto out_fail;
-                }
-                clnt->cl_intr     = 1;
-                clnt->cl_softrtry = 1;
-                clp->cl_rpcclient = clnt;
-                memcpy(clp->cl_ipaddr, server->ip_addr, sizeof(clp->cl_ipaddr));
-                nfs_idmap_new(clp);
-        }
-        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)) {
-                dprintk("%s: cannot create RPC client. Error = %d\n",
-                                __FUNCTION__, err);
-                return clnt;
        }
-        if (server->nfs4_state->cl_idmap == NULL) {
+        /* Get a superblock - note that we may end up sharing one that already exists */
-                dprintk("%s: failed to create idmapper.\n", __FUNCTION__);
+        s = sget(&nfs_fs_type, nfs_compare_super, nfs_set_super, server);
-                return ERR_PTR(-ENOMEM);
+        if (IS_ERR(s)) {
+                error = PTR_ERR(s);
+                goto out_err_nosb;
        }
-        if (clnt->cl_auth->au_flavor != flavor) {
+        if (s->s_fs_info != server) {
-                struct rpc_auth *auth;
+                nfs_free_server(server);
+                server = NULL;
-                auth = rpcauth_create(flavor, clnt);
-                if (IS_ERR(auth)) {
-                        dprintk("%s: couldn't create credcache!\n", __FUNCTION__);
-                        return (struct rpc_clnt *)auth;
-                }
        }
-        return clnt;
- out_fail:
-        if (clp)
-                nfs4_put_client(clp);
-        return ERR_PTR(err);
-}
-/*
- * Set up an NFS4 superblock
- */
-static int nfs4_fill_super(struct super_block *sb, struct nfs4_mount_data *data, int silent)
-{
-        struct nfs_server *server;
-        struct rpc_timeout timeparms;
-        rpc_authflavor_t authflavour;
-        int err = -EIO;
-        sb->s_blocksize_bits = 0;
+        if (!s->s_root) {
-        sb->s_blocksize = 0;
+                /* initial superblock/root creation */
-        server = NFS_SB(sb);
+                s->s_flags = flags;
-        if (data->rsize != 0)
+                nfs_clone_super(s, data->sb);
-                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;
-        server->caps = NFS_CAP_ATOMIC_OPEN;
-        server->acregmin = data->acregmin*HZ;
+        mntroot = nfs_get_root(s, data->fh);
-        server->acregmax = data->acregmax*HZ;
+        if (IS_ERR(mntroot)) {
-        server->acdirmin = data->acdirmin*HZ;
+                error = PTR_ERR(mntroot);
-        server->acdirmax = data->acdirmax*HZ;
+                goto error_splat_super;
+        }
-        server->rpc_ops = &nfs_v4_clientops;
+        s->s_flags |= MS_ACTIVE;
+        mnt->mnt_sb = s;
+        mnt->mnt_root = mntroot;
-        nfs_init_timeout_values(&timeparms, data->proto, data->timeo, data->retrans);
+        dprintk("<-- nfs_xdev_get_sb() = 0\n");
+        return 0;
-        server->retrans_timeo = timeparms.to_initval;
+out_err_nosb:
-        server->retrans_count = timeparms.to_retries;
+        nfs_free_server(server);
+out_err_noserver:
+        dprintk("<-- nfs_xdev_get_sb() = %d [error]\n", error);
+        return error;
-        /* Now create transport and client */
+error_splat_super:
-        authflavour = RPC_AUTH_UNIX;
+        up_write(&s->s_umount);
-        if (data->auth_flavourlen != 0) {
+        deactivate_super(s);
-                if (data->auth_flavourlen != 1) {
+        dprintk("<-- nfs_xdev_get_sb() = %d [splat]\n", error);
-                        dprintk("%s: Invalid number of RPC auth flavours %d.\n",
+        return error;
-                                        __FUNCTION__, data->auth_flavourlen);
+}
-                        err = -EINVAL;
-                        goto out_fail;
-                }
-                if (copy_from_user(&authflavour, data->auth_flavours, sizeof(authflavour))) {
-                        err = -EFAULT;
-                        goto out_fail;
-                }
-        }
-        server->client = nfs4_create_client(server, &timeparms, data->proto, authflavour);
+#ifdef CONFIG_NFS_V4
-        if (IS_ERR(server->client)) {
-                err = PTR_ERR(server->client);
-                        dprintk("%s: cannot create RPC client. Error = %d\n",
-                                        __FUNCTION__, err);
-                        goto out_fail;
-        }
+/*
+ * Finish setting up a cloned NFS4 superblock
+ */
+static void nfs4_clone_super(struct super_block *sb,
+                            const struct super_block *old_sb)
+{
+        sb->s_blocksize_bits = old_sb->s_blocksize_bits;
+        sb->s_blocksize = old_sb->s_blocksize;
+        sb->s_maxbytes = old_sb->s_maxbytes;
        sb->s_time_gran = 1;
+        sb->s_op = old_sb->s_op;
-        sb->s_op = &nfs4_sops;
+        nfs_initialise_sb(sb);
-        err = nfs_sb_init(sb, authflavour);
- out_fail:
-        return err;
 }
-static int nfs4_compare_super(struct super_block *sb, void *data)
+/*
+ * Set up an NFS4 superblock
+ */
+static void nfs4_fill_super(struct super_block *sb)
 {
-        struct nfs_server *server = data;
+        sb->s_time_gran = 1;
-        struct nfs_server *old = NFS_SB(sb);
+        sb->s_op = &nfs4_sops;
+        nfs_initialise_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 *
+static void *nfs_copy_user_string(char *dst, struct nfs_string *src, int maxlen)
-nfs_copy_user_string(char *dst, struct nfs_string *src, int maxlen)
 {
        void *p = NULL;
@@ -1297,14 +793,22 @@ nfs_copy_user_string(char *dst, struct nfs_string *src, int maxlen)
        return dst;
 }
+/*
+ * Get the superblock for an NFS4 mountpoint
+ */
 static int nfs4_get_sb(struct file_system_type *fs_type,
        int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt)
 {
-        int error;
-        struct nfs_server *server;
-        struct super_block *s;
        struct nfs4_mount_data *data = raw_data;
+        struct super_block *s;
+        struct nfs_server *server;
+        struct sockaddr_in addr;
+        rpc_authflavor_t authflavour;
+        struct nfs_fh mntfh;
+        struct dentry *mntroot;
+        char *mntpath = NULL, *hostname = NULL, ip_addr[16];
        void *p;
+        int error;
        if (data == NULL) {
                dprintk("%s: missing data argument\n", __FUNCTION__);
@@ -1315,84 +819,112 @@ static int nfs4_get_sb(struct file_system_type *fs_type,
                return -EINVAL;
        }
-        server = kzalloc(sizeof(struct nfs_server), GFP_KERNEL);
+        /* We now require that the mount process passes the remote address */
-        if (!server)
+        if (data->host_addrlen != sizeof(addr))
-                return -ENOMEM;
+                return -EINVAL;
-        /* Zero out the NFS state stuff */
-        init_nfsv4_state(server);
+        if (copy_from_user(&addr, data->host_addr, sizeof(addr)))
-        server->client = server->client_sys = server->client_acl = ERR_PTR(-EINVAL);
+                return -EFAULT;
+        if (addr.sin_family != AF_INET ||
+            addr.sin_addr.s_addr == INADDR_ANY
+            ) {
+                dprintk("%s: mount program didn't pass remote IP address!\n",
+                                __FUNCTION__);
+                return -EINVAL;
+        }
+        /* RFC3530: The default port for NFS is 2049 */
+        if (addr.sin_port == 0)
+                addr.sin_port = NFS_PORT;
+        /* Grab the authentication type */
+        authflavour = RPC_AUTH_UNIX;
+        if (data->auth_flavourlen != 0) {
+                if (data->auth_flavourlen != 1) {
+                        dprintk("%s: Invalid number of RPC auth flavours %d.\n",
+                                        __FUNCTION__, data->auth_flavourlen);
+                        error = -EINVAL;
+                        goto out_err_noserver;
+                }
+                if (copy_from_user(&authflavour, data->auth_flavours,
+                                   sizeof(authflavour))) {
+                        error = -EFAULT;
+                        goto out_err_noserver;
+                }
+        }
        p = nfs_copy_user_string(NULL, &data->hostname, 256);
        if (IS_ERR(p))
                goto out_err;
-        server->hostname = p;
+        hostname = p;
        p = nfs_copy_user_string(NULL, &data->mnt_path, 1024);
        if (IS_ERR(p))
                goto out_err;
-        server->mnt_path = p;
+        mntpath = p;
+        dprintk("MNTPATH: %s\n", mntpath);
-        p = nfs_copy_user_string(server->ip_addr, &data->client_addr,
+        p = nfs_copy_user_string(ip_addr, &data->client_addr,
-                        sizeof(server->ip_addr) - 1);
+                                 sizeof(ip_addr) - 1);
        if (IS_ERR(p))
                goto out_err;
-        /* We now require that the mount process passes the remote address */
+        /* Get a volume representation */
-        if (data->host_addrlen != sizeof(server->addr)) {
+        server = nfs4_create_server(data, hostname, &addr, mntpath, ip_addr,
-                error = -EINVAL;
+                                    authflavour, &mntfh);
-                goto out_free;
+        if (IS_ERR(server)) {
-        }
+                error = PTR_ERR(server);
-        if (copy_from_user(&server->addr, data->host_addr, sizeof(server->addr))) {
+                goto out_err_noserver;
-                error = -EFAULT;
-                goto out_free;
-        }
-        if (server->addr.sin_family != AF_INET ||
-            server->addr.sin_addr.s_addr == INADDR_ANY) {
-                dprintk("%s: mount program didn't pass remote IP address!\n",
-                                __FUNCTION__);
-                error = -EINVAL;
-                goto out_free;
-        }
-        /* Fire up rpciod if not yet running */
-        error = rpciod_up();
-        if (error < 0) {
-                dprintk("%s: couldn't start rpciod! Error = %d\n",
-                                __FUNCTION__, error);
-                goto out_free;
        }
-        s = sget(fs_type, nfs4_compare_super, nfs_set_super, server);
+        /* Get a superblock - note that we may end up sharing one that already exists */
+        s = sget(fs_type, nfs_compare_super, nfs_set_super, server);
        if (IS_ERR(s)) {
                error = PTR_ERR(s);
                goto out_free;
        }
-        if (s->s_root) {
+        if (s->s_fs_info != server) {
-                kfree(server->mnt_path);
+                nfs_free_server(server);
-                kfree(server->hostname);
+                server = NULL;
-                kfree(server);
-                return simple_set_mnt(mnt, s);
        }
-        s->s_flags = flags;
+        if (!s->s_root) {
+                /* initial superblock/root creation */
+                s->s_flags = flags;
+                nfs4_fill_super(s);
+        }
-        error = nfs4_fill_super(s, data, flags & MS_SILENT ? 1 : 0);
+        mntroot = nfs4_get_root(s, &mntfh);
-        if (error) {
+        if (IS_ERR(mntroot)) {
-                up_write(&s->s_umount);
+                error = PTR_ERR(mntroot);
-                deactivate_super(s);
+                goto error_splat_super;
-                return error;
        }
        s->s_flags |= MS_ACTIVE;
-        return simple_set_mnt(mnt, s);
+        mnt->mnt_sb = s;
+        mnt->mnt_root = mntroot;
+        kfree(mntpath);
+        kfree(hostname);
+        return 0;
 out_err:
        error = PTR_ERR(p);
+        goto out_err_noserver;
 out_free:
-        kfree(server->mnt_path);
+        nfs_free_server(server);
-        kfree(server->hostname);
+out_err_noserver:
-        kfree(server);
+        kfree(mntpath);
+        kfree(hostname);
        return error;
+error_splat_super:
+        up_write(&s->s_umount);
+        deactivate_super(s);
+        goto out_err_noserver;
 }
 static void nfs4_kill_super(struct super_block *sb)
@@ -1403,135 +935,140 @@ static void nfs4_kill_super(struct super_block *sb)
        kill_anon_super(sb);
        nfs4_renewd_prepare_shutdown(server);
+        nfs_free_server(server);
+}
+/*
+ * Clone an NFS4 server record on xdev traversal (FSID-change)
+ */
+static int nfs4_xdev_get_sb(struct file_system_type *fs_type, int flags,
+                            const char *dev_name, void *raw_data,
+                            struct vfsmount *mnt)
+{
+        struct nfs_clone_mount *data = raw_data;
+        struct super_block *s;
+        struct nfs_server *server;
+        struct dentry *mntroot;
+        int error;
+        dprintk("--> nfs4_xdev_get_sb()\n");
+        /* create a new volume representation */
+        server = nfs_clone_server(NFS_SB(data->sb), data->fh, data->fattr);
+        if (IS_ERR(server)) {
+                error = PTR_ERR(server);
+                goto out_err_noserver;
+        }
+        /* Get a superblock - note that we may end up sharing one that already exists */
+        s = sget(&nfs_fs_type, nfs_compare_super, nfs_set_super, server);
+        if (IS_ERR(s)) {
+                error = PTR_ERR(s);
+                goto out_err_nosb;
+        }
-        if (server->client != NULL && !IS_ERR(server->client))
+        if (s->s_fs_info != server) {
-                rpc_shutdown_client(server->client);
+                nfs_free_server(server);
+                server = NULL;
+        }
-        destroy_nfsv4_state(server);
+        if (!s->s_root) {
+                /* initial superblock/root creation */
+                s->s_flags = flags;
+                nfs4_clone_super(s, data->sb);
+        }
+        mntroot = nfs4_get_root(s, data->fh);
+        if (IS_ERR(mntroot)) {
+                error = PTR_ERR(mntroot);
+                goto error_splat_super;
+        }
-        rpciod_down();
+        s->s_flags |= MS_ACTIVE;
+        mnt->mnt_sb = s;
+        mnt->mnt_root = mntroot;
+        dprintk("<-- nfs4_xdev_get_sb() = 0\n");
+        return 0;
+out_err_nosb:
+        nfs_free_server(server);
+out_err_noserver:
+        dprintk("<-- nfs4_xdev_get_sb() = %d [error]\n", error);
+        return error;
-        nfs_free_iostats(server->io_stats);
+error_splat_super:
-        kfree(server->hostname);
+        up_write(&s->s_umount);
-        kfree(server);
+        deactivate_super(s);
-        nfs_release_automount_timer();
+        dprintk("<-- nfs4_xdev_get_sb() = %d [splat]\n", error);
+        return error;
 }
 /*
- * Constructs the SERVER-side path
+ * Create an NFS4 server record on referral traversal
 */
-static inline char *nfs4_dup_path(const struct dentry *dentry)
+static int nfs4_referral_get_sb(struct file_system_type *fs_type, int flags,
+                                const char *dev_name, void *raw_data,
+                                struct vfsmount *mnt)
 {
-        char *page = (char *) __get_free_page(GFP_USER);
+        struct nfs_clone_mount *data = raw_data;
-        char *path;
+        struct super_block *s;
+        struct nfs_server *server;
+        struct dentry *mntroot;
+        struct nfs_fh mntfh;
+        int error;
-        path = nfs4_path(dentry, page, PAGE_SIZE);
+        dprintk("--> nfs4_referral_get_sb()\n");
-        if (!IS_ERR(path)) {
-                int len = PAGE_SIZE + page - path;
-                char *tmp = path;
-                path = kmalloc(len, GFP_KERNEL);
+        /* create a new volume representation */
-                if (path)
+        server = nfs4_create_referral_server(data, &mntfh);
-                        memcpy(path, tmp, len);
+        if (IS_ERR(server)) {
-                else
+                error = PTR_ERR(server);
-                        path = ERR_PTR(-ENOMEM);
+                goto out_err_noserver;
        }
-        free_page((unsigned long)page);
-        return path;
-}
-static struct super_block *nfs4_clone_sb(struct nfs_server *server, struct nfs_clone_mount *data)
+        /* Get a superblock - note that we may end up sharing one that already exists */
-{
+        s = sget(&nfs_fs_type, nfs_compare_super, nfs_set_super, server);
-        const struct dentry *dentry = data->dentry;
+        if (IS_ERR(s)) {
-        struct nfs4_client *clp = server->nfs4_state;
+                error = PTR_ERR(s);
-        struct super_block *sb;
+                goto out_err_nosb;
-        server->fsid = data->fattr->fsid;
-        nfs_copy_fh(&server->fh, data->fh);
-        server->mnt_path = nfs4_dup_path(dentry);
-        if (IS_ERR(server->mnt_path)) {
-                sb = (struct super_block *)server->mnt_path;
-                goto err;
        }
-        sb = sget(&nfs4_fs_type, nfs4_compare_super, nfs_set_super, server);
-        if (IS_ERR(sb) || sb->s_root)
-                goto free_path;
-        nfs4_server_capabilities(server, &server->fh);
-        down_write(&clp->cl_sem);
-        atomic_inc(&clp->cl_count);
-        list_add_tail(&server->nfs4_siblings, &clp->cl_superblocks);
-        up_write(&clp->cl_sem);
-        return sb;
-free_path:
-        kfree(server->mnt_path);
-err:
-        server->mnt_path = NULL;
-        return sb;
-}
-static int nfs_clone_nfs4_sb(struct file_system_type *fs_type,
+        if (s->s_fs_info != server) {
-                int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt)
+                nfs_free_server(server);
-{
+                server = NULL;
-        struct nfs_clone_mount *data = raw_data;
+        }
-        return nfs_clone_generic_sb(data, nfs4_clone_sb, nfs_clone_server, mnt);
-}
-static struct super_block *nfs4_referral_sb(struct nfs_server *server, struct nfs_clone_mount *data)
+        if (!s->s_root) {
-{
+                /* initial superblock/root creation */
-        struct super_block *sb = ERR_PTR(-ENOMEM);
+                s->s_flags = flags;
-        int len;
+                nfs4_fill_super(s);
+        }
-        len = strlen(data->mnt_path) + 1;
-        server->mnt_path = kmalloc(len, GFP_KERNEL);
-        if (server->mnt_path == NULL)
-                goto err;
-        memcpy(server->mnt_path, data->mnt_path, len);
-        memcpy(&server->addr, data->addr, sizeof(struct sockaddr_in));
-        sb = sget(&nfs4_fs_type, nfs4_compare_super, nfs_set_super, server);
-        if (IS_ERR(sb) || sb->s_root)
-                goto free_path;
-        return sb;
-free_path:
-        kfree(server->mnt_path);
-err:
-        server->mnt_path = NULL;
-        return sb;
-}
-static struct nfs_server *nfs4_referral_server(struct super_block *sb, struct nfs_clone_mount *data)
+        mntroot = nfs4_get_root(s, data->fh);
-{
+        if (IS_ERR(mntroot)) {
-        struct nfs_server *server = NFS_SB(sb);
+                error = PTR_ERR(mntroot);
-        struct rpc_timeout timeparms;
+                goto error_splat_super;
-        int proto, timeo, retrans;
+        }
-        void *err;
-        proto = IPPROTO_TCP;
-        /* Since we are following a referral and there may be alternatives,
-           set the timeouts and retries to low values */
-        timeo = 2;
-        retrans = 1;
-        nfs_init_timeout_values(&timeparms, proto, timeo, retrans);
-        server->client = nfs4_create_client(server, &timeparms, proto, data->authflavor);
-        if (IS_ERR((err = server->client)))
-                goto out_err;
-        sb->s_time_gran = 1;
+        s->s_flags |= MS_ACTIVE;
-        sb->s_op = &nfs4_sops;
+        mnt->mnt_sb = s;
-        err = ERR_PTR(nfs_sb_init(sb, data->authflavor));
+        mnt->mnt_root = mntroot;
-        if (!IS_ERR(err))
-                return server;
-out_err:
-        return (struct nfs_server *)err;
-}
-static int nfs_referral_nfs4_sb(struct file_system_type *fs_type,
+        dprintk("<-- nfs4_referral_get_sb() = 0\n");
-                int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt)
+        return 0;
-{
-        struct nfs_clone_mount *data = raw_data;
+out_err_nosb:
-        return nfs_clone_generic_sb(data, nfs4_referral_sb, nfs4_referral_server, mnt);
+        nfs_free_server(server);
+out_err_noserver:
+        dprintk("<-- nfs4_referral_get_sb() = %d [error]\n", error);
+        return error;
+error_splat_super:
+        up_write(&s->s_umount);
+        deactivate_super(s);
+        dprintk("<-- nfs4_referral_get_sb() = %d [splat]\n", error);
+        return error;
 }
-#endif
+#endif /* CONFIG_NFS_V4 */
diff --git a/fs/nfs/write.c b/fs/nfs/write.c
index 7084ac9a6455..c12effb46fe5 100644
--- a/fs/nfs/write.c
+++ b/fs/nfs/write.c
@@ -396,6 +396,7 @@ int nfs_writepages(struct address_space *mapping, struct writeback_control *wbc)
 out:
        clear_bit(BDI_write_congested, &bdi->state);
        wake_up_all(&nfs_write_congestion);
+        writeback_congestion_end();
        return err;
 }
@@ -1252,7 +1253,13 @@ int nfs_writeback_done(struct rpc_task *task, struct nfs_write_data *data)
        dprintk("NFS: %4d nfs_writeback_done (status %d)\n",
                task->tk_pid, task->tk_status);
-        /* Call the NFS version-specific code */
+        /*
+         * ->write_done will attempt to use post-op attributes to detect
+         * conflicting writes by other clients.  A strict interpretation
+         * of close-to-open would allow us to continue caching even if
+         * another writer had changed the file, but some applications
+         * depend on tighter cache coherency when writing.
+         */
        status = NFS_PROTO(data->inode)->write_done(task, data);
        if (status != 0)
                return status;
@@ -1273,7 +1280,7 @@ int nfs_writeback_done(struct rpc_task *task, struct nfs_write_data *data)
                if (time_before(complain, jiffies)) {
                        dprintk("NFS: faulty NFS server %s:"
                                " (committed = %d) != (stable = %d)\n",
-                                NFS_SERVER(data->inode)->hostname,
+                                NFS_SERVER(data->inode)->nfs_client->cl_hostname,
                                resp->verf->committed, argp->stable);
                        complain = jiffies + 300 * HZ;
                }
diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c
index 54b37b1d2e3a..8583d99ee740 100644
--- a/fs/nfsd/nfs4callback.c
+++ b/fs/nfsd/nfs4callback.c
@@ -375,16 +375,28 @@ nfsd4_probe_callback(struct nfs4_client *clp)
 {
        struct sockaddr_in      addr;
        struct nfs4_callback    *cb = &clp->cl_callback;
-        struct rpc_timeout      timeparms;
+        struct rpc_timeout      timeparms = {
-        struct rpc_xprt *       xprt;
+                .to_initval     = (NFSD_LEASE_TIME/4) * HZ,
+                .to_retries     = 5,
+                .to_maxval      = (NFSD_LEASE_TIME/2) * HZ,
+                .to_exponential = 1,
+        };
        struct rpc_program *    program = &cb->cb_program;
-        struct rpc_stat *       stat = &cb->cb_stat;
+        struct rpc_create_args args = {
-        struct rpc_clnt *       clnt;
+                .protocol       = IPPROTO_TCP,
+                .address        = (struct sockaddr *)&addr,
+                .addrsize       = sizeof(addr),
+                .timeout        = &timeparms,
+                .servername     = clp->cl_name.data,
+                .program        = program,
+                .version        = nfs_cb_version[1]->number,
+                .authflavor     = RPC_AUTH_UNIX,        /* XXX: need AUTH_GSS... */
+                .flags          = (RPC_CLNT_CREATE_NOPING),
+        };
        struct rpc_message msg = {
                .rpc_proc       = &nfs4_cb_procedures[NFSPROC4_CLNT_CB_NULL],
                .rpc_argp       = clp,
        };
-        char                    hostname[32];
        int status;
        if (atomic_read(&cb->cb_set))
@@ -396,51 +408,27 @@ nfsd4_probe_callback(struct nfs4_client *clp)
        addr.sin_port = htons(cb->cb_port);
        addr.sin_addr.s_addr = htonl(cb->cb_addr);
-        /* Initialize timeout */
-        timeparms.to_initval = (NFSD_LEASE_TIME/4) * HZ;
-        timeparms.to_retries = 0;
-        timeparms.to_maxval = (NFSD_LEASE_TIME/2) * HZ;
-        timeparms.to_exponential = 1;
-        /* Create RPC transport */
-        xprt = xprt_create_proto(IPPROTO_TCP, &addr, &timeparms);
-        if (IS_ERR(xprt)) {
-                dprintk("NFSD: couldn't create callback transport!\n");
-                goto out_err;
-        }
        /* Initialize rpc_program */
        program->name = "nfs4_cb";
        program->number = cb->cb_prog;
        program->nrvers = ARRAY_SIZE(nfs_cb_version);
        program->version = nfs_cb_version;
-        program->stats = stat;
+        program->stats = &cb->cb_stat;
        /* Initialize rpc_stat */
-        memset(stat, 0, sizeof(struct rpc_stat));
+        memset(program->stats, 0, sizeof(cb->cb_stat));
-        stat->program = program;
+        program->stats->program = program;
-        /* Create RPC client
+        /* Create RPC client */
-         *
+        cb->cb_client = rpc_create(&args);
-         * XXX AUTH_UNIX only - need AUTH_GSS....
+        if (!cb->cb_client) {
-         */
-        sprintf(hostname, "%u.%u.%u.%u", NIPQUAD(addr.sin_addr.s_addr));
-        clnt = rpc_new_client(xprt, hostname, program, 1, RPC_AUTH_UNIX);
-        if (IS_ERR(clnt)) {
                dprintk("NFSD: couldn't create callback client\n");
                goto out_err;
        }
-        clnt->cl_intr = 0;
-        clnt->cl_softrtry = 1;
        /* Kick rpciod, put the call on the wire. */
+        if (rpciod_up() != 0)
-        if (rpciod_up() != 0) {
-                dprintk("nfsd: couldn't start rpciod for callbacks!\n");
                goto out_clnt;
-        }
-        cb->cb_client = clnt;
        /* the task holds a reference to the nfs4_client struct */
        atomic_inc(&clp->cl_count);
@@ -448,7 +436,7 @@ nfsd4_probe_callback(struct nfs4_client *clp)
        msg.rpc_cred = nfsd4_lookupcred(clp,0);
        if (IS_ERR(msg.rpc_cred))
                goto out_rpciod;
-        status = rpc_call_async(clnt, &msg, RPC_TASK_ASYNC, &nfs4_cb_null_ops, NULL);
+        status = rpc_call_async(cb->cb_client, &msg, RPC_TASK_ASYNC, &nfs4_cb_null_ops, NULL);
        put_rpccred(msg.rpc_cred);
        if (status != 0) {
@@ -462,7 +450,7 @@ out_rpciod:
        rpciod_down();
        cb->cb_client = NULL;
 out_clnt:
-        rpc_shutdown_client(clnt);
+        rpc_shutdown_client(cb->cb_client);
 out_err:
        dprintk("NFSD: warning: no callback path to client %.*s\n",
                (int)clp->cl_name.len, clp->cl_name.data);
diff --git a/fs/nfsd/nfs4recover.c b/fs/nfsd/nfs4recover.c
index 06da7506363c..e35d7e52fdeb 100644
--- a/fs/nfsd/nfs4recover.c
+++ b/fs/nfsd/nfs4recover.c
@@ -33,7 +33,7 @@
 *
 */
+#include <linux/err.h>
 #include <linux/sunrpc/svc.h>
 #include <linux/nfsd/nfsd.h>
 #include <linux/nfs4.h>
@@ -87,34 +87,35 @@ int
 nfs4_make_rec_clidname(char *dname, struct xdr_netobj *clname)
 {
        struct xdr_netobj cksum;
-        struct crypto_tfm *tfm;
+        struct hash_desc desc;
        struct scatterlist sg[1];
        int status = nfserr_resource;
        dprintk("NFSD: nfs4_make_rec_clidname for %.*s\n",
                        clname->len, clname->data);
-        tfm = crypto_alloc_tfm("md5", CRYPTO_TFM_REQ_MAY_SLEEP);
+        desc.flags = CRYPTO_TFM_REQ_MAY_SLEEP;
-        if (tfm == NULL)
+        desc.tfm = crypto_alloc_hash("md5", 0, CRYPTO_ALG_ASYNC);
-                goto out;
+        if (IS_ERR(desc.tfm))
-        cksum.len = crypto_tfm_alg_digestsize(tfm);
+                goto out_no_tfm;
+        cksum.len = crypto_hash_digestsize(desc.tfm);
        cksum.data = kmalloc(cksum.len, GFP_KERNEL);
        if (cksum.data == NULL)
                goto out;
-        crypto_digest_init(tfm);
        sg[0].page = virt_to_page(clname->data);
        sg[0].offset = offset_in_page(clname->data);
        sg[0].length = clname->len;
-        crypto_digest_update(tfm, sg, 1);
+        if (crypto_hash_digest(&desc, sg, sg->length, cksum.data))
-        crypto_digest_final(tfm, cksum.data);
+                goto out;
        md5_to_hex(dname, cksum.data);
        kfree(cksum.data);
        status = nfs_ok;
 out:
-        crypto_free_tfm(tfm);
+        crypto_free_hash(desc.tfm);
+out_no_tfm:
        return status;
 }
diff --git a/fs/ocfs2/Makefile b/fs/ocfs2/Makefile
index 7d3be845a614..9fb8132f19b0 100644
--- a/fs/ocfs2/Makefile
+++ b/fs/ocfs2/Makefile
@@ -16,6 +16,7 @@ ocfs2-objs := \
        file.o                  \
        heartbeat.o             \
        inode.o                 \
+        ioctl.o                 \
        journal.o               \
        localalloc.o            \
        mmap.o                  \
diff --git a/fs/ocfs2/alloc.c b/fs/ocfs2/alloc.c
index edaab05a93e0..f43bc5f18a35 100644
--- a/fs/ocfs2/alloc.c
+++ b/fs/ocfs2/alloc.c
@@ -1717,17 +1717,29 @@ static int ocfs2_do_truncate(struct ocfs2_super *osb,
                        ocfs2_remove_from_cache(inode, eb_bh);
-                        BUG_ON(eb->h_suballoc_slot);
                        BUG_ON(el->l_recs[0].e_clusters);
                        BUG_ON(el->l_recs[0].e_cpos);
                        BUG_ON(el->l_recs[0].e_blkno);
-                        status = ocfs2_free_extent_block(handle,
+                        if (eb->h_suballoc_slot == 0) {
-                                                         tc->tc_ext_alloc_inode,
+                                /*
-                                                         tc->tc_ext_alloc_bh,
+                                 * This code only understands how to
-                                                         eb);
+                                 * lock the suballocator in slot 0,
-                        if (status < 0) {
+                                 * which is fine because allocation is
-                                mlog_errno(status);
+                                 * only ever done out of that
-                                goto bail;
+                                 * suballocator too. A future version
+                                 * might change that however, so avoid
+                                 * a free if we don't know how to
+                                 * handle it. This way an fs incompat
+                                 * bit will not be necessary.
+                                 */
+                                status = ocfs2_free_extent_block(handle,
+                                                                 tc->tc_ext_alloc_inode,
+                                                                 tc->tc_ext_alloc_bh,
+                                                                 eb);
+                                if (status < 0) {
+                                        mlog_errno(status);
+                                        goto bail;
+                                }
                        }
                }
                brelse(eb_bh);
diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c
index f1d1c342ce01..3d7c082a8f58 100644
--- a/fs/ocfs2/aops.c
+++ b/fs/ocfs2/aops.c
@@ -391,31 +391,28 @@ out:
 static int ocfs2_commit_write(struct file *file, struct page *page,
                              unsigned from, unsigned to)
 {
-        int ret, extending = 0, locklevel = 0;
+        int ret;
-        loff_t new_i_size;
        struct buffer_head *di_bh = NULL;
        struct inode *inode = page->mapping->host;
        struct ocfs2_journal_handle *handle = NULL;
+        struct ocfs2_dinode *di;
        mlog_entry("(0x%p, 0x%p, %u, %u)\n", file, page, from, to);
        /* NOTE: ocfs2_file_aio_write has ensured that it's safe for
-         * us to sample inode->i_size here without the metadata lock:
+         * us to continue here without rechecking the I/O against
+         * changed inode values.
         *
         * 1) We're currently holding the inode alloc lock, so no
         *    nodes can change it underneath us.
         *
         * 2) We've had to take the metadata lock at least once
-         *    already to check for extending writes, hence insuring
+         *    already to check for extending writes, suid removal, etc.
-         *    that our current copy is also up to date.
+         *    The meta data update code then ensures that we don't get a
+         *    stale inode allocation image (i_size, i_clusters, etc).
         */
-        new_i_size = ((loff_t)page->index << PAGE_CACHE_SHIFT) + to;
-        if (new_i_size > i_size_read(inode)) {
-                extending = 1;
-                locklevel = 1;
-        }
-        ret = ocfs2_meta_lock_with_page(inode, NULL, &di_bh, locklevel, page);
+        ret = ocfs2_meta_lock_with_page(inode, NULL, &di_bh, 1, page);
        if (ret != 0) {
                mlog_errno(ret);
                goto out;
@@ -427,23 +424,20 @@ static int ocfs2_commit_write(struct file *file, struct page *page,
                goto out_unlock_meta;
        }
-        if (extending) {
+        handle = ocfs2_start_walk_page_trans(inode, page, from, to);
-                handle = ocfs2_start_walk_page_trans(inode, page, from, to);
+        if (IS_ERR(handle)) {
-                if (IS_ERR(handle)) {
+                ret = PTR_ERR(handle);
-                        ret = PTR_ERR(handle);
+                goto out_unlock_data;
-                        handle = NULL;
+        }
-                        goto out_unlock_data;
-                }
-                /* Mark our buffer early. We'd rather catch this error up here
+        /* Mark our buffer early. We'd rather catch this error up here
-                 * as opposed to after a successful commit_write which would
+         * as opposed to after a successful commit_write which would
-                 * require us to set back inode->i_size. */
+         * require us to set back inode->i_size. */
-                ret = ocfs2_journal_access(handle, inode, di_bh,
+        ret = ocfs2_journal_access(handle, inode, di_bh,
-                                           OCFS2_JOURNAL_ACCESS_WRITE);
+                                   OCFS2_JOURNAL_ACCESS_WRITE);
-                if (ret < 0) {
+        if (ret < 0) {
-                        mlog_errno(ret);
+                mlog_errno(ret);
-                        goto out_commit;
+                goto out_commit;
-                }
        }
        /* might update i_size */
@@ -453,37 +447,28 @@ static int ocfs2_commit_write(struct file *file, struct page *page,
                goto out_commit;
        }
-        if (extending) {
+        di = (struct ocfs2_dinode *)di_bh->b_data;
-                loff_t size = (u64) i_size_read(inode);
-                struct ocfs2_dinode *di =
-                        (struct ocfs2_dinode *)di_bh->b_data;
-                /* ocfs2_mark_inode_dirty is too heavy to use here. */
+        /* ocfs2_mark_inode_dirty() is too heavy to use here. */
-                inode->i_blocks = ocfs2_align_bytes_to_sectors(size);
+        inode->i_mtime = inode->i_ctime = CURRENT_TIME;
-                inode->i_ctime = inode->i_mtime = CURRENT_TIME;
+        di->i_mtime = di->i_ctime = cpu_to_le64(inode->i_mtime.tv_sec);
+        di->i_mtime_nsec = di->i_ctime_nsec = cpu_to_le32(inode->i_mtime.tv_nsec);
-                di->i_size = cpu_to_le64(size);
+        inode->i_blocks = ocfs2_align_bytes_to_sectors((u64)(i_size_read(inode)));
-                di->i_ctime = di->i_mtime = 
+        di->i_size = cpu_to_le64((u64)i_size_read(inode));
-                                cpu_to_le64(inode->i_mtime.tv_sec);
-                di->i_ctime_nsec = di->i_mtime_nsec = 
-                                cpu_to_le32(inode->i_mtime.tv_nsec);
-                ret = ocfs2_journal_dirty(handle, di_bh);
+        ret = ocfs2_journal_dirty(handle, di_bh);
-                if (ret < 0) {
+        if (ret < 0) {
-                        mlog_errno(ret);
+                mlog_errno(ret);
-                        goto out_commit;
+                goto out_commit;
-                }
        }
-        BUG_ON(extending && (i_size_read(inode) != new_i_size));
 out_commit:
-        if (handle)
+        ocfs2_commit_trans(handle);
-                ocfs2_commit_trans(handle);
 out_unlock_data:
        ocfs2_data_unlock(inode, 1);
 out_unlock_meta:
-        ocfs2_meta_unlock(inode, locklevel);
+        ocfs2_meta_unlock(inode, 1);
 out:
        if (di_bh)
                brelse(di_bh);
diff --git a/fs/ocfs2/buffer_head_io.c b/fs/ocfs2/buffer_head_io.c
index 9a24adf9be6e..c9037414f4f6 100644
--- a/fs/ocfs2/buffer_head_io.c
+++ b/fs/ocfs2/buffer_head_io.c
@@ -100,6 +100,9 @@ int ocfs2_read_blocks(struct ocfs2_super *osb, u64 block, int nr,
        mlog_entry("(block=(%llu), nr=(%d), flags=%d, inode=%p)\n",
                   (unsigned long long)block, nr, flags, inode);
+        BUG_ON((flags & OCFS2_BH_READAHEAD) &&
+               (!inode || !(flags & OCFS2_BH_CACHED)));
        if (osb == NULL || osb->sb == NULL || bhs == NULL) {
                status = -EINVAL;
                mlog_errno(status);
@@ -140,6 +143,30 @@ int ocfs2_read_blocks(struct ocfs2_super *osb, u64 block, int nr,
                bh = bhs[i];
                ignore_cache = 0;
+                /* There are three read-ahead cases here which we need to
+                 * be concerned with. All three assume a buffer has
+                 * previously been submitted with OCFS2_BH_READAHEAD
+                 * and it hasn't yet completed I/O.
+                 *
+                 * 1) The current request is sync to disk. This rarely
+                 *    happens these days, and never when performance
+                 *    matters - the code can just wait on the buffer
+                 *    lock and re-submit.
+                 *
+                 * 2) The current request is cached, but not
+                 *    readahead. ocfs2_buffer_uptodate() will return
+                 *    false anyway, so we'll wind up waiting on the
+                 *    buffer lock to do I/O. We re-check the request
+                 *    with after getting the lock to avoid a re-submit.
+                 *
+                 * 3) The current request is readahead (and so must
+                 *    also be a caching one). We short circuit if the
+                 *    buffer is locked (under I/O) and if it's in the
+                 *    uptodate cache. The re-check from #2 catches the
+                 *    case that the previous read-ahead completes just
+                 *    before our is-it-in-flight check.
+                 */
                if (flags & OCFS2_BH_CACHED &&
                    !ocfs2_buffer_uptodate(inode, bh)) {
                        mlog(ML_UPTODATE,
@@ -169,6 +196,14 @@ int ocfs2_read_blocks(struct ocfs2_super *osb, u64 block, int nr,
                                continue;
                        }
+                        /* A read-ahead request was made - if the
+                         * buffer is already under read-ahead from a
+                         * previously submitted request than we are
+                         * done here. */
+                        if ((flags & OCFS2_BH_READAHEAD)
+                            && ocfs2_buffer_read_ahead(inode, bh))
+                                continue;
                        lock_buffer(bh);
                        if (buffer_jbd(bh)) {
 #ifdef CATCH_BH_JBD_RACES
@@ -181,13 +216,22 @@ int ocfs2_read_blocks(struct ocfs2_super *osb, u64 block, int nr,
                                continue;
 #endif
                        }
+                        /* Re-check ocfs2_buffer_uptodate() as a
+                         * previously read-ahead buffer may have
+                         * completed I/O while we were waiting for the
+                         * buffer lock. */
+                        if ((flags & OCFS2_BH_CACHED)
+                            && !(flags & OCFS2_BH_READAHEAD)
+                            && ocfs2_buffer_uptodate(inode, bh)) {
+                                unlock_buffer(bh);
+                                continue;
+                        }
                        clear_buffer_uptodate(bh);
                        get_bh(bh); /* for end_buffer_read_sync() */
                        bh->b_end_io = end_buffer_read_sync;
-                        if (flags & OCFS2_BH_READAHEAD)
+                        submit_bh(READ, bh);
-                                submit_bh(READA, bh);
-                        else
-                                submit_bh(READ, bh);
                        continue;
                }
        }
@@ -197,34 +241,39 @@ int ocfs2_read_blocks(struct ocfs2_super *osb, u64 block, int nr,
        for (i = (nr - 1); i >= 0; i--) {
                bh = bhs[i];
-                /* We know this can't have changed as we hold the
+                if (!(flags & OCFS2_BH_READAHEAD)) {
-                 * inode sem. Avoid doing any work on the bh if the
+                        /* We know this can't have changed as we hold the
-                 * journal has it. */
+                         * inode sem. Avoid doing any work on the bh if the
-                if (!buffer_jbd(bh))
+                         * journal has it. */
-                        wait_on_buffer(bh);
+                        if (!buffer_jbd(bh))
+                                wait_on_buffer(bh);
-                if (!buffer_uptodate(bh)) {
-                        /* Status won't be cleared from here on out,
+                        if (!buffer_uptodate(bh)) {
-                         * so we can safely record this and loop back
+                                /* Status won't be cleared from here on out,
-                         * to cleanup the other buffers. Don't need to
+                                 * so we can safely record this and loop back
-                         * remove the clustered uptodate information
+                                 * to cleanup the other buffers. Don't need to
-                         * for this bh as it's not marked locally
+                                 * remove the clustered uptodate information
-                         * uptodate. */
+                                 * for this bh as it's not marked locally
-                        status = -EIO;
+                                 * uptodate. */
-                        brelse(bh);
+                                status = -EIO;
-                        bhs[i] = NULL;
+                                brelse(bh);
-                        continue;
+                                bhs[i] = NULL;
+                                continue;
+                        }
                }
+                /* Always set the buffer in the cache, even if it was
+                 * a forced read, or read-ahead which hasn't yet
+                 * completed. */
                if (inode)
                        ocfs2_set_buffer_uptodate(inode, bh);
        }
        if (inode)
                mutex_unlock(&OCFS2_I(inode)->ip_io_mutex);
-        mlog(ML_BH_IO, "block=(%llu), nr=(%d), cached=%s\n", 
+        mlog(ML_BH_IO, "block=(%llu), nr=(%d), cached=%s, flags=0x%x\n", 
             (unsigned long long)block, nr,
-             (!(flags & OCFS2_BH_CACHED) || ignore_cache) ? "no" : "yes");
+             (!(flags & OCFS2_BH_CACHED) || ignore_cache) ? "no" : "yes", flags);
 bail:
diff --git a/fs/ocfs2/buffer_head_io.h b/fs/ocfs2/buffer_head_io.h
index 6ecb90937b68..6cc20930fac3 100644
--- a/fs/ocfs2/buffer_head_io.h
+++ b/fs/ocfs2/buffer_head_io.h
@@ -49,7 +49,7 @@ int ocfs2_read_blocks(struct ocfs2_super          *osb,
 #define OCFS2_BH_CACHED            1
-#define OCFS2_BH_READAHEAD         8    /* use this to pass READA down to submit_bh */
+#define OCFS2_BH_READAHEAD         8
 static inline int ocfs2_read_block(struct ocfs2_super * osb, u64 off,
                                   struct buffer_head **bh, int flags,
diff --git a/fs/ocfs2/cluster/heartbeat.c b/fs/ocfs2/cluster/heartbeat.c
index 504595d6cf65..305cba3681fe 100644
--- a/fs/ocfs2/cluster/heartbeat.c
+++ b/fs/ocfs2/cluster/heartbeat.c
@@ -320,8 +320,12 @@ static int compute_max_sectors(struct block_device *bdev)
                max_pages = q->max_hw_segments;
        max_pages--; /* Handle I/Os that straddle a page */
-        max_sectors = max_pages << (PAGE_SHIFT - 9);
+        if (max_pages) {
+                max_sectors = max_pages << (PAGE_SHIFT - 9);
+        } else {
+                /* If BIO contains 1 or less than 1 page. */
+                max_sectors = q->max_sectors;
+        }
        /* Why is fls() 1-based???? */
        pow_two_sectors = 1 << (fls(max_sectors) - 1);
diff --git a/fs/ocfs2/cluster/tcp_internal.h b/fs/ocfs2/cluster/tcp_internal.h
index ff9e2e2104c2..4b46aac7d243 100644
--- a/fs/ocfs2/cluster/tcp_internal.h
+++ b/fs/ocfs2/cluster/tcp_internal.h
@@ -44,11 +44,17 @@
 * locking semantics of the file system using the protocol.  It should 
 * be somewhere else, I'm sure, but right now it isn't.
 *
+ * New in version 4:
+ *      - Remove i_generation from lock names for better stat performance.
+ *
+ * New in version 3:
+ *      - Replace dentry votes with a cluster lock
+ *
 * New in version 2:
 *      - full 64 bit i_size in the metadata lock lvbs
 *      - introduction of "rw" lock and pushing meta/data locking down
 */
-#define O2NET_PROTOCOL_VERSION 2ULL
+#define O2NET_PROTOCOL_VERSION 4ULL
 struct o2net_handshake {
        __be64  protocol_version;
        __be64  connector_id;
diff --git a/fs/ocfs2/dcache.c b/fs/ocfs2/dcache.c
index 1a01380e3878..014e73978dac 100644
--- a/fs/ocfs2/dcache.c
+++ b/fs/ocfs2/dcache.c
@@ -35,15 +35,17 @@
 #include "alloc.h"
 #include "dcache.h"
+#include "dlmglue.h"
 #include "file.h"
 #include "inode.h"
 static int ocfs2_dentry_revalidate(struct dentry *dentry,
                                   struct nameidata *nd)
 {
        struct inode *inode = dentry->d_inode;
        int ret = 0;    /* if all else fails, just return false */
-        struct ocfs2_super *osb;
+        struct ocfs2_super *osb = OCFS2_SB(dentry->d_sb);
        mlog_entry("(0x%p, '%.*s')\n", dentry,
                   dentry->d_name.len, dentry->d_name.name);
@@ -55,28 +57,31 @@ static int ocfs2_dentry_revalidate(struct dentry *dentry,
                goto bail;
        }
-        osb = OCFS2_SB(inode->i_sb);
        BUG_ON(!osb);
-        if (inode != osb->root_inode) {
+        if (inode == osb->root_inode || is_bad_inode(inode))
-                spin_lock(&OCFS2_I(inode)->ip_lock);
+                goto bail;
-                /* did we or someone else delete this inode? */
-                if (OCFS2_I(inode)->ip_flags & OCFS2_INODE_DELETED) {
+        spin_lock(&OCFS2_I(inode)->ip_lock);
-                        spin_unlock(&OCFS2_I(inode)->ip_lock);
+        /* did we or someone else delete this inode? */
-                        mlog(0, "inode (%llu) deleted, returning false\n",
+        if (OCFS2_I(inode)->ip_flags & OCFS2_INODE_DELETED) {
-                             (unsigned long long)OCFS2_I(inode)->ip_blkno);
-                        goto bail;
-                }
                spin_unlock(&OCFS2_I(inode)->ip_lock);
+                mlog(0, "inode (%llu) deleted, returning false\n",
+                     (unsigned long long)OCFS2_I(inode)->ip_blkno);
+                goto bail;
+        }
+        spin_unlock(&OCFS2_I(inode)->ip_lock);
-                if (!inode->i_nlink) {
+        /*
-                        mlog(0, "Inode %llu orphaned, returning false "
+         * We don't need a cluster lock to test this because once an
-                             "dir = %d\n",
+         * inode nlink hits zero, it never goes back.
-                             (unsigned long long)OCFS2_I(inode)->ip_blkno,
+         */
-                             S_ISDIR(inode->i_mode));
+        if (inode->i_nlink == 0) {
-                        goto bail;
+                mlog(0, "Inode %llu orphaned, returning false "
-                }
+                     "dir = %d\n",
+                     (unsigned long long)OCFS2_I(inode)->ip_blkno,
+                     S_ISDIR(inode->i_mode));
+                goto bail;
        }
        ret = 1;
@@ -87,6 +92,322 @@ bail:
        return ret;
 }
+static int ocfs2_match_dentry(struct dentry *dentry,
+                              u64 parent_blkno,
+                              int skip_unhashed)
+{
+        struct inode *parent;
+        /*
+         * ocfs2_lookup() does a d_splice_alias() _before_ attaching
+         * to the lock data, so we skip those here, otherwise
+         * ocfs2_dentry_attach_lock() will get its original dentry
+         * back.
+         */
+        if (!dentry->d_fsdata)
+                return 0;
+        if (!dentry->d_parent)
+                return 0;
+        if (skip_unhashed && d_unhashed(dentry))
+                return 0;
+        parent = dentry->d_parent->d_inode;
+        /* Negative parent dentry? */
+        if (!parent)
+                return 0;
+        /* Name is in a different directory. */
+        if (OCFS2_I(parent)->ip_blkno != parent_blkno)
+                return 0;
+        return 1;
+}
+/*
+ * Walk the inode alias list, and find a dentry which has a given
+ * parent. ocfs2_dentry_attach_lock() wants to find _any_ alias as it
+ * is looking for a dentry_lock reference. The vote thread is looking
+ * to unhash aliases, so we allow it to skip any that already have
+ * that property.
+ */
+struct dentry *ocfs2_find_local_alias(struct inode *inode,
+                                      u64 parent_blkno,
+                                      int skip_unhashed)
+{
+        struct list_head *p;
+        struct dentry *dentry = NULL;
+        spin_lock(&dcache_lock);
+        list_for_each(p, &inode->i_dentry) {
+                dentry = list_entry(p, struct dentry, d_alias);
+                if (ocfs2_match_dentry(dentry, parent_blkno, skip_unhashed)) {
+                        mlog(0, "dentry found: %.*s\n",
+                             dentry->d_name.len, dentry->d_name.name);
+                        dget_locked(dentry);
+                        break;
+                }
+                dentry = NULL;
+        }
+        spin_unlock(&dcache_lock);
+        return dentry;
+}
+DEFINE_SPINLOCK(dentry_attach_lock);
+/*
+ * Attach this dentry to a cluster lock.
+ *
+ * Dentry locks cover all links in a given directory to a particular
+ * inode. We do this so that ocfs2 can build a lock name which all
+ * nodes in the cluster can agree on at all times. Shoving full names
+ * in the cluster lock won't work due to size restrictions. Covering
+ * links inside of a directory is a good compromise because it still
+ * allows us to use the parent directory lock to synchronize
+ * operations.
+ *
+ * Call this function with the parent dir semaphore and the parent dir
+ * cluster lock held.
+ *
+ * The dir semaphore will protect us from having to worry about
+ * concurrent processes on our node trying to attach a lock at the
+ * same time.
+ *
+ * The dir cluster lock (held at either PR or EX mode) protects us
+ * from unlink and rename on other nodes.
+ *
+ * A dput() can happen asynchronously due to pruning, so we cover
+ * attaching and detaching the dentry lock with a
+ * dentry_attach_lock.
+ *
+ * A node which has done lookup on a name retains a protected read
+ * lock until final dput. If the user requests and unlink or rename,
+ * the protected read is upgraded to an exclusive lock. Other nodes
+ * who have seen the dentry will then be informed that they need to
+ * downgrade their lock, which will involve d_delete on the
+ * dentry. This happens in ocfs2_dentry_convert_worker().
+ */
+int ocfs2_dentry_attach_lock(struct dentry *dentry,
+                             struct inode *inode,
+                             u64 parent_blkno)
+{
+        int ret;
+        struct dentry *alias;
+        struct ocfs2_dentry_lock *dl = dentry->d_fsdata;
+        mlog(0, "Attach \"%.*s\", parent %llu, fsdata: %p\n",
+             dentry->d_name.len, dentry->d_name.name,
+             (unsigned long long)parent_blkno, dl);
+        /*
+         * Negative dentry. We ignore these for now.
+         *
+         * XXX: Could we can improve ocfs2_dentry_revalidate() by
+         * tracking these?
+         */
+        if (!inode)
+                return 0;
+        if (dl) {
+                mlog_bug_on_msg(dl->dl_parent_blkno != parent_blkno,
+                                " \"%.*s\": old parent: %llu, new: %llu\n",
+                                dentry->d_name.len, dentry->d_name.name,
+                                (unsigned long long)parent_blkno,
+                                (unsigned long long)dl->dl_parent_blkno);
+                return 0;
+        }
+        alias = ocfs2_find_local_alias(inode, parent_blkno, 0);
+        if (alias) {
+                /*
+                 * Great, an alias exists, which means we must have a
+                 * dentry lock already. We can just grab the lock off
+                 * the alias and add it to the list.
+                 *
+                 * We're depending here on the fact that this dentry
+                 * was found and exists in the dcache and so must have
+                 * a reference to the dentry_lock because we can't
+                 * race creates. Final dput() cannot happen on it
+                 * since we have it pinned, so our reference is safe.
+                 */
+                dl = alias->d_fsdata;
+                mlog_bug_on_msg(!dl, "parent %llu, ino %llu\n",
+                                (unsigned long long)parent_blkno,
+                                (unsigned long long)OCFS2_I(inode)->ip_blkno);
+                mlog_bug_on_msg(dl->dl_parent_blkno != parent_blkno,
+                                " \"%.*s\": old parent: %llu, new: %llu\n",
+                                dentry->d_name.len, dentry->d_name.name,
+                                (unsigned long long)parent_blkno,
+                                (unsigned long long)dl->dl_parent_blkno);
+                mlog(0, "Found: %s\n", dl->dl_lockres.l_name);
+                goto out_attach;
+        }
+        /*
+         * There are no other aliases
+         */
+        dl = kmalloc(sizeof(*dl), GFP_NOFS);
+        if (!dl) {
+                ret = -ENOMEM;
+                mlog_errno(ret);
+                return ret;
+        }
+        dl->dl_count = 0;
+        /*
+         * Does this have to happen below, for all attaches, in case
+         * the struct inode gets blown away by votes?
+         */
+        dl->dl_inode = igrab(inode);
+        dl->dl_parent_blkno = parent_blkno;
+        ocfs2_dentry_lock_res_init(dl, parent_blkno, inode);
+out_attach:
+        spin_lock(&dentry_attach_lock);
+        dentry->d_fsdata = dl;
+        dl->dl_count++;
+        spin_unlock(&dentry_attach_lock);
+        /*
+         * This actually gets us our PRMODE level lock. From now on,
+         * we'll have a notification if one of these names is
+         * destroyed on another node.
+         */
+        ret = ocfs2_dentry_lock(dentry, 0);
+        if (!ret)
+                ocfs2_dentry_unlock(dentry, 0);
+        else
+                mlog_errno(ret);
+        dput(alias);
+        return ret;
+}
+/*
+ * ocfs2_dentry_iput() and friends.
+ *
+ * At this point, our particular dentry is detached from the inodes
+ * alias list, so there's no way that the locking code can find it.
+ *
+ * The interesting stuff happens when we determine that our lock needs
+ * to go away because this is the last subdir alias in the
+ * system. This function needs to handle a couple things:
+ *
+ * 1) Synchronizing lock shutdown with the downconvert threads. This
+ *    is already handled for us via the lockres release drop function
+ *    called in ocfs2_release_dentry_lock()
+ *
+ * 2) A race may occur when we're doing our lock shutdown and
+ *    another process wants to create a new dentry lock. Right now we
+ *    let them race, which means that for a very short while, this
+ *    node might have two locks on a lock resource. This should be a
+ *    problem though because one of them is in the process of being
+ *    thrown out.
+ */
+static void ocfs2_drop_dentry_lock(struct ocfs2_super *osb,
+                                   struct ocfs2_dentry_lock *dl)
+{
+        ocfs2_simple_drop_lockres(osb, &dl->dl_lockres);
+        ocfs2_lock_res_free(&dl->dl_lockres);
+        iput(dl->dl_inode);
+        kfree(dl);
+}
+void ocfs2_dentry_lock_put(struct ocfs2_super *osb,
+                           struct ocfs2_dentry_lock *dl)
+{
+        int unlock = 0;
+        BUG_ON(dl->dl_count == 0);
+        spin_lock(&dentry_attach_lock);
+        dl->dl_count--;
+        unlock = !dl->dl_count;
+        spin_unlock(&dentry_attach_lock);
+        if (unlock)
+                ocfs2_drop_dentry_lock(osb, dl);
+}
+static void ocfs2_dentry_iput(struct dentry *dentry, struct inode *inode)
+{
+        struct ocfs2_dentry_lock *dl = dentry->d_fsdata;
+        mlog_bug_on_msg(!dl && !(dentry->d_flags & DCACHE_DISCONNECTED),
+                        "dentry: %.*s\n", dentry->d_name.len,
+                        dentry->d_name.name);
+        if (!dl)
+                goto out;
+        mlog_bug_on_msg(dl->dl_count == 0, "dentry: %.*s, count: %u\n",
+                        dentry->d_name.len, dentry->d_name.name,
+                        dl->dl_count);
+        ocfs2_dentry_lock_put(OCFS2_SB(dentry->d_sb), dl);
+out:
+        iput(inode);
+}
+/*
+ * d_move(), but keep the locks in sync.
+ *
+ * When we are done, "dentry" will have the parent dir and name of
+ * "target", which will be thrown away.
+ *
+ * We manually update the lock of "dentry" if need be.
+ *
+ * "target" doesn't have it's dentry lock touched - we allow the later
+ * dput() to handle this for us.
+ *
+ * This is called during ocfs2_rename(), while holding parent
+ * directory locks. The dentries have already been deleted on other
+ * nodes via ocfs2_remote_dentry_delete().
+ *
+ * Normally, the VFS handles the d_move() for the file sytem, after
+ * the ->rename() callback. OCFS2 wants to handle this internally, so
+ * the new lock can be created atomically with respect to the cluster.
+ */
+void ocfs2_dentry_move(struct dentry *dentry, struct dentry *target,
+                       struct inode *old_dir, struct inode *new_dir)
+{
+        int ret;
+        struct ocfs2_super *osb = OCFS2_SB(old_dir->i_sb);
+        struct inode *inode = dentry->d_inode;
+        /*
+         * Move within the same directory, so the actual lock info won't
+         * change.
+         *
+         * XXX: Is there any advantage to dropping the lock here?
+         */
+        if (old_dir == new_dir)
+                goto out_move;
+        ocfs2_dentry_lock_put(osb, dentry->d_fsdata);
+        dentry->d_fsdata = NULL;
+        ret = ocfs2_dentry_attach_lock(dentry, inode, OCFS2_I(new_dir)->ip_blkno);
+        if (ret)
+                mlog_errno(ret);
+out_move:
+        d_move(dentry, target);
+}
 struct dentry_operations ocfs2_dentry_ops = {
        .d_revalidate           = ocfs2_dentry_revalidate,
+        .d_iput                 = ocfs2_dentry_iput,
 };
diff --git a/fs/ocfs2/dcache.h b/fs/ocfs2/dcache.h
index 90072771114b..c091c34d9883 100644
--- a/fs/ocfs2/dcache.h
+++ b/fs/ocfs2/dcache.h
@@ -28,4 +28,31 @@
 extern struct dentry_operations ocfs2_dentry_ops;
+struct ocfs2_dentry_lock {
+        unsigned int            dl_count;
+        u64                     dl_parent_blkno;
+        /*
+         * The ocfs2_dentry_lock keeps an inode reference until
+         * dl_lockres has been destroyed. This is usually done in
+         * ->d_iput() anyway, so there should be minimal impact.
+         */
+        struct inode            *dl_inode;
+        struct ocfs2_lock_res   dl_lockres;
+};
+int ocfs2_dentry_attach_lock(struct dentry *dentry, struct inode *inode,
+                             u64 parent_blkno);
+void ocfs2_dentry_lock_put(struct ocfs2_super *osb,
+                           struct ocfs2_dentry_lock *dl);
+struct dentry *ocfs2_find_local_alias(struct inode *inode, u64 parent_blkno,
+                                      int skip_unhashed);
+void ocfs2_dentry_move(struct dentry *dentry, struct dentry *target,
+                       struct inode *old_dir, struct inode *new_dir);
+extern spinlock_t dentry_attach_lock;
 #endif /* OCFS2_DCACHE_H */
diff --git a/fs/ocfs2/dir.c b/fs/ocfs2/dir.c
index 3d494d1a5f36..04e01915b86e 100644
--- a/fs/ocfs2/dir.c
+++ b/fs/ocfs2/dir.c
@@ -74,14 +74,14 @@ static int ocfs2_extend_dir(struct ocfs2_super *osb,
 int ocfs2_readdir(struct file * filp, void * dirent, filldir_t filldir)
 {
        int error = 0;
-        unsigned long offset, blk;
+        unsigned long offset, blk, last_ra_blk = 0;
-        int i, num, stored;
+        int i, stored;
        struct buffer_head * bh, * tmp;
        struct ocfs2_dir_entry * de;
        int err;
        struct inode *inode = filp->f_dentry->d_inode;
        struct super_block * sb = inode->i_sb;
-        int have_disk_lock = 0;
+        unsigned int ra_sectors = 16;
        mlog_entry("dirino=%llu\n",
                   (unsigned long long)OCFS2_I(inode)->ip_blkno);
@@ -95,9 +95,8 @@ int ocfs2_readdir(struct file * filp, void * dirent, filldir_t filldir)
                        mlog_errno(error);
                /* we haven't got any yet, so propagate the error. */
                stored = error;
-                goto bail;
+                goto bail_nolock;
        }
-        have_disk_lock = 1;
        offset = filp->f_pos & (sb->s_blocksize - 1);
@@ -113,16 +112,21 @@ int ocfs2_readdir(struct file * filp, void * dirent, filldir_t filldir)
                        continue;
                }
-                /*
+                /* The idea here is to begin with 8k read-ahead and to stay
-                 * Do the readahead (8k)
+                 * 4k ahead of our current position.
-                 */
+                 *
-                if (!offset) {
+                 * TODO: Use the pagecache for this. We just need to
-                        for (i = 16 >> (sb->s_blocksize_bits - 9), num = 0;
+                 * make sure it's cluster-safe... */
+                if (!last_ra_blk
+                    || (((last_ra_blk - blk) << 9) <= (ra_sectors / 2))) {
+                        for (i = ra_sectors >> (sb->s_blocksize_bits - 9);
                             i > 0; i--) {
                                tmp = ocfs2_bread(inode, ++blk, &err, 1);
                                if (tmp)
                                        brelse(tmp);
                        }
+                        last_ra_blk = blk;
+                        ra_sectors = 8;
                }
 revalidate:
@@ -194,9 +198,9 @@ revalidate:
        stored = 0;
 bail:
-        if (have_disk_lock)
+        ocfs2_meta_unlock(inode, 0);
-                ocfs2_meta_unlock(inode, 0);
+bail_nolock:
        mlog_exit(stored);
        return stored;
diff --git a/fs/ocfs2/dlm/dlmapi.h b/fs/ocfs2/dlm/dlmapi.h
index 53652f51c0e1..cfd5cb65cab0 100644
--- a/fs/ocfs2/dlm/dlmapi.h
+++ b/fs/ocfs2/dlm/dlmapi.h
@@ -182,6 +182,7 @@ enum dlm_status dlmlock(struct dlm_ctxt *dlm,
                        struct dlm_lockstatus *lksb,
                        int flags,
                        const char *name,
+                        int namelen,
                        dlm_astlockfunc_t *ast,
                        void *data,
                        dlm_bastlockfunc_t *bast);
diff --git a/fs/ocfs2/dlm/dlmast.c b/fs/ocfs2/dlm/dlmast.c
index 42775e2bbe2c..681046d51393 100644
--- a/fs/ocfs2/dlm/dlmast.c
+++ b/fs/ocfs2/dlm/dlmast.c
@@ -320,8 +320,8 @@ int dlm_proxy_ast_handler(struct o2net_msg *msg, u32 len, void *data)
        res = dlm_lookup_lockres(dlm, name, locklen);
        if (!res) {
-                mlog(ML_ERROR, "got %sast for unknown lockres! "
+                mlog(0, "got %sast for unknown lockres! "
-                               "cookie=%u:%llu, name=%.*s, namelen=%u\n",
+                     "cookie=%u:%llu, name=%.*s, namelen=%u\n",
                     past->type == DLM_AST ? "" : "b",
                     dlm_get_lock_cookie_node(cookie),
                     dlm_get_lock_cookie_seq(cookie),
@@ -367,12 +367,10 @@ int dlm_proxy_ast_handler(struct o2net_msg *msg, u32 len, void *data)
                        goto do_ast;
        }
-        mlog(ML_ERROR, "got %sast for unknown lock!  cookie=%u:%llu, "
+        mlog(0, "got %sast for unknown lock!  cookie=%u:%llu, "
-                       "name=%.*s, namelen=%u\n", 
+             "name=%.*s, namelen=%u\n", past->type == DLM_AST ? "" : "b", 
-                       past->type == DLM_AST ? "" : "b", 
+             dlm_get_lock_cookie_node(cookie), dlm_get_lock_cookie_seq(cookie),
-                       dlm_get_lock_cookie_node(cookie),
+             locklen, name, locklen);
-                       dlm_get_lock_cookie_seq(cookie),
-                       locklen, name, locklen);
        ret = DLM_NORMAL;
 unlock_out:
@@ -464,7 +462,7 @@ int dlm_send_proxy_ast_msg(struct dlm_ctxt *dlm, struct dlm_lock_resource *res,
                        mlog(ML_ERROR, "sent AST to node %u, it returned "
                             "DLM_MIGRATING!\n", lock->ml.node);
                        BUG();
-                } else if (status != DLM_NORMAL) {
+                } else if (status != DLM_NORMAL && status != DLM_IVLOCKID) {
                        mlog(ML_ERROR, "AST to node %u returned %d!\n",
                             lock->ml.node, status);
                        /* ignore it */
diff --git a/fs/ocfs2/dlm/dlmcommon.h b/fs/ocfs2/dlm/dlmcommon.h
index 14530ee7e11d..fa968180b072 100644
--- a/fs/ocfs2/dlm/dlmcommon.h
+++ b/fs/ocfs2/dlm/dlmcommon.h
@@ -747,6 +747,7 @@ void dlm_change_lockres_owner(struct dlm_ctxt *dlm,
                              u8 owner);
 struct dlm_lock_resource * dlm_get_lock_resource(struct dlm_ctxt *dlm,
                                                 const char *lockid,
+                                                 int namelen,
                                                 int flags);
 struct dlm_lock_resource *dlm_new_lockres(struct dlm_ctxt *dlm,
                                          const char *name,
diff --git a/fs/ocfs2/dlm/dlmlock.c b/fs/ocfs2/dlm/dlmlock.c
index 5ca57ec650c7..42a1b91979b5 100644
--- a/fs/ocfs2/dlm/dlmlock.c
+++ b/fs/ocfs2/dlm/dlmlock.c
@@ -540,8 +540,8 @@ static inline void dlm_get_next_cookie(u8 node_num, u64 *cookie)
 enum dlm_status dlmlock(struct dlm_ctxt *dlm, int mode,
                        struct dlm_lockstatus *lksb, int flags,
-                        const char *name, dlm_astlockfunc_t *ast, void *data,
+                        const char *name, int namelen, dlm_astlockfunc_t *ast,
-                        dlm_bastlockfunc_t *bast)
+                        void *data, dlm_bastlockfunc_t *bast)
 {
        enum dlm_status status;
        struct dlm_lock_resource *res = NULL;
@@ -571,7 +571,7 @@ enum dlm_status dlmlock(struct dlm_ctxt *dlm, int mode,
        recovery = (flags & LKM_RECOVERY);
        if (recovery &&
-            (!dlm_is_recovery_lock(name, strlen(name)) || convert) ) {
+            (!dlm_is_recovery_lock(name, namelen) || convert) ) {
                dlm_error(status);
                goto error;
        }
@@ -643,7 +643,7 @@ retry_convert:
                }
                status = DLM_IVBUFLEN;
-                if (strlen(name) > DLM_LOCKID_NAME_MAX || strlen(name) < 1) {
+                if (namelen > DLM_LOCKID_NAME_MAX || namelen < 1) {
                        dlm_error(status);
                        goto error;
                }
@@ -659,7 +659,7 @@ retry_convert:
                        dlm_wait_for_recovery(dlm);
                /* find or create the lock resource */
-                res = dlm_get_lock_resource(dlm, name, flags);
+                res = dlm_get_lock_resource(dlm, name, namelen, flags);
                if (!res) {
                        status = DLM_IVLOCKID;
                        dlm_error(status);
diff --git a/fs/ocfs2/dlm/dlmmaster.c b/fs/ocfs2/dlm/dlmmaster.c
index 9503240ef0e5..f784177b6241 100644
--- a/fs/ocfs2/dlm/dlmmaster.c
+++ b/fs/ocfs2/dlm/dlmmaster.c
@@ -740,6 +740,7 @@ struct dlm_lock_resource *dlm_new_lockres(struct dlm_ctxt *dlm,
 */
 struct dlm_lock_resource * dlm_get_lock_resource(struct dlm_ctxt *dlm,
                                          const char *lockid,
+                                          int namelen,
                                          int flags)
 {
        struct dlm_lock_resource *tmpres=NULL, *res=NULL;
@@ -748,13 +749,12 @@ struct dlm_lock_resource * dlm_get_lock_resource(struct dlm_ctxt *dlm,
        int blocked = 0;
        int ret, nodenum;
        struct dlm_node_iter iter;
-        unsigned int namelen, hash;
+        unsigned int hash;
        int tries = 0;
        int bit, wait_on_recovery = 0;
        BUG_ON(!lockid);
-        namelen = strlen(lockid);
        hash = dlm_lockid_hash(lockid, namelen);
        mlog(0, "get lockres %s (len %d)\n", lockid, namelen);
diff --git a/fs/ocfs2/dlm/dlmrecovery.c b/fs/ocfs2/dlm/dlmrecovery.c
index 594745fab0b5..9d950d7cea38 100644
--- a/fs/ocfs2/dlm/dlmrecovery.c
+++ b/fs/ocfs2/dlm/dlmrecovery.c
@@ -2285,7 +2285,8 @@ again:
        memset(&lksb, 0, sizeof(lksb));
        ret = dlmlock(dlm, LKM_EXMODE, &lksb, LKM_NOQUEUE|LKM_RECOVERY,
-                      DLM_RECOVERY_LOCK_NAME, dlm_reco_ast, dlm, dlm_reco_bast);
+                      DLM_RECOVERY_LOCK_NAME, DLM_RECOVERY_LOCK_NAME_LEN,
+                      dlm_reco_ast, dlm, dlm_reco_bast);
        mlog(0, "%s: dlmlock($RECOVERY) returned %d, lksb=%d\n",
             dlm->name, ret, lksb.status);
diff --git a/fs/ocfs2/dlm/userdlm.c b/fs/ocfs2/dlm/userdlm.c
index e641b084b343..eead48bbfac6 100644
--- a/fs/ocfs2/dlm/userdlm.c
+++ b/fs/ocfs2/dlm/userdlm.c
@@ -102,10 +102,10 @@ static inline void user_recover_from_dlm_error(struct user_lock_res *lockres)
        spin_unlock(&lockres->l_lock);
 }
-#define user_log_dlm_error(_func, _stat, _lockres) do {         \
+#define user_log_dlm_error(_func, _stat, _lockres) do {                 \
-        mlog(ML_ERROR, "Dlm error \"%s\" while calling %s on "  \
+        mlog(ML_ERROR, "Dlm error \"%s\" while calling %s on "          \
-                "resource %s: %s\n", dlm_errname(_stat), _func, \
+                "resource %.*s: %s\n", dlm_errname(_stat), _func,       \
-                _lockres->l_name, dlm_errmsg(_stat));           \
+                _lockres->l_namelen, _lockres->l_name, dlm_errmsg(_stat)); \
 } while (0)
 /* WARNING: This function lives in a world where the only three lock
@@ -127,21 +127,22 @@ static void user_ast(void *opaque)
        struct user_lock_res *lockres = opaque;
        struct dlm_lockstatus *lksb;
-        mlog(0, "AST fired for lockres %s\n", lockres->l_name);
+        mlog(0, "AST fired for lockres %.*s\n", lockres->l_namelen,
+             lockres->l_name);
        spin_lock(&lockres->l_lock);
        lksb = &(lockres->l_lksb);
        if (lksb->status != DLM_NORMAL) {
-                mlog(ML_ERROR, "lksb status value of %u on lockres %s\n",
+                mlog(ML_ERROR, "lksb status value of %u on lockres %.*s\n",
-                     lksb->status, lockres->l_name);
+                     lksb->status, lockres->l_namelen, lockres->l_name);
                spin_unlock(&lockres->l_lock);
                return;
        }
        mlog_bug_on_msg(lockres->l_requested == LKM_IVMODE,
-                        "Lockres %s, requested ivmode. flags 0x%x\n",
+                        "Lockres %.*s, requested ivmode. flags 0x%x\n",
-                        lockres->l_name, lockres->l_flags);
+                        lockres->l_namelen, lockres->l_name, lockres->l_flags);
        /* we're downconverting. */
        if (lockres->l_requested < lockres->l_level) {
@@ -213,8 +214,8 @@ static void user_bast(void *opaque, int level)
 {
        struct user_lock_res *lockres = opaque;
-        mlog(0, "Blocking AST fired for lockres %s. Blocking level %d\n",
+        mlog(0, "Blocking AST fired for lockres %.*s. Blocking level %d\n",
-                lockres->l_name, level);
+             lockres->l_namelen, lockres->l_name, level);
        spin_lock(&lockres->l_lock);
        lockres->l_flags |= USER_LOCK_BLOCKED;
@@ -231,7 +232,8 @@ static void user_unlock_ast(void *opaque, enum dlm_status status)
 {
        struct user_lock_res *lockres = opaque;
-        mlog(0, "UNLOCK AST called on lock %s\n", lockres->l_name);
+        mlog(0, "UNLOCK AST called on lock %.*s\n", lockres->l_namelen,
+             lockres->l_name);
        if (status != DLM_NORMAL && status != DLM_CANCELGRANT)
                mlog(ML_ERROR, "Dlm returns status %d\n", status);
@@ -244,8 +246,6 @@ static void user_unlock_ast(void *opaque, enum dlm_status status)
            && !(lockres->l_flags & USER_LOCK_IN_CANCEL)) {
                lockres->l_level = LKM_IVMODE;
        } else if (status == DLM_CANCELGRANT) {
-                mlog(0, "Lock %s, cancel fails, flags 0x%x\n",
-                     lockres->l_name, lockres->l_flags);
                /* We tried to cancel a convert request, but it was
                 * already granted. Don't clear the busy flag - the
                 * ast should've done this already. */
@@ -255,8 +255,6 @@ static void user_unlock_ast(void *opaque, enum dlm_status status)
        } else {
                BUG_ON(!(lockres->l_flags & USER_LOCK_IN_CANCEL));
                /* Cancel succeeded, we want to re-queue */
-                mlog(0, "Lock %s, cancel succeeds, flags 0x%x\n",
-                     lockres->l_name, lockres->l_flags);
                lockres->l_requested = LKM_IVMODE; /* cancel an
                                                    * upconvert
                                                    * request. */
@@ -287,13 +285,14 @@ static void user_dlm_unblock_lock(void *opaque)
        struct user_lock_res *lockres = (struct user_lock_res *) opaque;
        struct dlm_ctxt *dlm = dlm_ctxt_from_user_lockres(lockres);
-        mlog(0, "processing lockres %s\n", lockres->l_name);
+        mlog(0, "processing lockres %.*s\n", lockres->l_namelen,
+             lockres->l_name);
        spin_lock(&lockres->l_lock);
        mlog_bug_on_msg(!(lockres->l_flags & USER_LOCK_QUEUED),
-                        "Lockres %s, flags 0x%x\n",
+                        "Lockres %.*s, flags 0x%x\n",
-                        lockres->l_name, lockres->l_flags);
+                        lockres->l_namelen, lockres->l_name, lockres->l_flags);
        /* notice that we don't clear USER_LOCK_BLOCKED here. If it's
         * set, we want user_ast clear it. */
@@ -305,22 +304,16 @@ static void user_dlm_unblock_lock(void *opaque)
         * flag, and finally we might get another bast which re-queues
         * us before our ast for the downconvert is called. */
        if (!(lockres->l_flags & USER_LOCK_BLOCKED)) {
-                mlog(0, "Lockres %s, flags 0x%x: queued but not blocking\n",
-                        lockres->l_name, lockres->l_flags);
                spin_unlock(&lockres->l_lock);
                goto drop_ref;
        }
        if (lockres->l_flags & USER_LOCK_IN_TEARDOWN) {
-                mlog(0, "lock is in teardown so we do nothing\n");
                spin_unlock(&lockres->l_lock);
                goto drop_ref;
        }
        if (lockres->l_flags & USER_LOCK_BUSY) {
-                mlog(0, "Cancel lock %s, flags 0x%x\n",
-                     lockres->l_name, lockres->l_flags);
                if (lockres->l_flags & USER_LOCK_IN_CANCEL) {
                        spin_unlock(&lockres->l_lock);
                        goto drop_ref;
@@ -372,6 +365,7 @@ static void user_dlm_unblock_lock(void *opaque)
                         &lockres->l_lksb,
                         LKM_CONVERT|LKM_VALBLK,
                         lockres->l_name,
+                         lockres->l_namelen,
                         user_ast,
                         lockres,
                         user_bast);
@@ -420,16 +414,16 @@ int user_dlm_cluster_lock(struct user_lock_res *lockres,
        if (level != LKM_EXMODE &&
            level != LKM_PRMODE) {
-                mlog(ML_ERROR, "lockres %s: invalid request!\n",
+                mlog(ML_ERROR, "lockres %.*s: invalid request!\n",
-                     lockres->l_name);
+                     lockres->l_namelen, lockres->l_name);
                status = -EINVAL;
                goto bail;
        }
-        mlog(0, "lockres %s: asking for %s lock, passed flags = 0x%x\n",
+        mlog(0, "lockres %.*s: asking for %s lock, passed flags = 0x%x\n",
-                lockres->l_name,
+             lockres->l_namelen, lockres->l_name,
-                (level == LKM_EXMODE) ? "LKM_EXMODE" : "LKM_PRMODE",
+             (level == LKM_EXMODE) ? "LKM_EXMODE" : "LKM_PRMODE",
-                lkm_flags);
+             lkm_flags);
 again:
        if (signal_pending(current)) {
@@ -474,15 +468,13 @@ again:
                BUG_ON(level == LKM_IVMODE);
                BUG_ON(level == LKM_NLMODE);
-                mlog(0, "lock %s, get lock from %d to level = %d\n",
-                        lockres->l_name, lockres->l_level, level);
                /* call dlm_lock to upgrade lock now */
                status = dlmlock(dlm,
                                 level,
                                 &lockres->l_lksb,
                                 local_flags,
                                 lockres->l_name,
+                                 lockres->l_namelen,
                                 user_ast,
                                 lockres,
                                 user_bast);
@@ -498,9 +490,6 @@ again:
                        goto bail;
                }
-                mlog(0, "lock %s, successfull return from dlmlock\n",
-                        lockres->l_name);
                user_wait_on_busy_lock(lockres);
                goto again;
        }
@@ -508,9 +497,6 @@ again:
        user_dlm_inc_holders(lockres, level);
        spin_unlock(&lockres->l_lock);
-        mlog(0, "lockres %s: Got %s lock!\n", lockres->l_name,
-                (level == LKM_EXMODE) ? "LKM_EXMODE" : "LKM_PRMODE");
        status = 0;
 bail:
        return status;
@@ -538,13 +524,11 @@ void user_dlm_cluster_unlock(struct user_lock_res *lockres,
 {
        if (level != LKM_EXMODE &&
            level != LKM_PRMODE) {
-                mlog(ML_ERROR, "lockres %s: invalid request!\n", lockres->l_name);
+                mlog(ML_ERROR, "lockres %.*s: invalid request!\n",
+                     lockres->l_namelen, lockres->l_name);
                return;
        }
-        mlog(0, "lockres %s: dropping %s lock\n", lockres->l_name,
-                (level == LKM_EXMODE) ? "LKM_EXMODE" : "LKM_PRMODE");
        spin_lock(&lockres->l_lock);
        user_dlm_dec_holders(lockres, level);
        __user_dlm_cond_queue_lockres(lockres);
@@ -602,6 +586,7 @@ void user_dlm_lock_res_init(struct user_lock_res *lockres,
        memcpy(lockres->l_name,
               dentry->d_name.name,
               dentry->d_name.len);
+        lockres->l_namelen = dentry->d_name.len;
 }
 int user_dlm_destroy_lock(struct user_lock_res *lockres)
@@ -609,11 +594,10 @@ int user_dlm_destroy_lock(struct user_lock_res *lockres)
        int status = -EBUSY;
        struct dlm_ctxt *dlm = dlm_ctxt_from_user_lockres(lockres);
-        mlog(0, "asked to destroy %s\n", lockres->l_name);
+        mlog(0, "asked to destroy %.*s\n", lockres->l_namelen, lockres->l_name);
        spin_lock(&lockres->l_lock);
        if (lockres->l_flags & USER_LOCK_IN_TEARDOWN) {
-                mlog(0, "Lock is already torn down\n");
                spin_unlock(&lockres->l_lock);
                return 0;
        }
@@ -623,8 +607,6 @@ int user_dlm_destroy_lock(struct user_lock_res *lockres)
        while (lockres->l_flags & USER_LOCK_BUSY) {
                spin_unlock(&lockres->l_lock);
-                mlog(0, "lock %s is busy\n", lockres->l_name);
                user_wait_on_busy_lock(lockres);
                spin_lock(&lockres->l_lock);
@@ -632,14 +614,12 @@ int user_dlm_destroy_lock(struct user_lock_res *lockres)
        if (lockres->l_ro_holders || lockres->l_ex_holders) {
                spin_unlock(&lockres->l_lock);
-                mlog(0, "lock %s has holders\n", lockres->l_name);
                goto bail;
        }
        status = 0;
        if (!(lockres->l_flags & USER_LOCK_ATTACHED)) {
                spin_unlock(&lockres->l_lock);
-                mlog(0, "lock %s is not attached\n", lockres->l_name);
                goto bail;
        }
@@ -647,7 +627,6 @@ int user_dlm_destroy_lock(struct user_lock_res *lockres)
        lockres->l_flags |= USER_LOCK_BUSY;
        spin_unlock(&lockres->l_lock);
-        mlog(0, "unlocking lockres %s\n", lockres->l_name);
        status = dlmunlock(dlm,
                           &lockres->l_lksb,
                           LKM_VALBLK,
diff --git a/fs/ocfs2/dlm/userdlm.h b/fs/ocfs2/dlm/userdlm.h
index 04178bc40b76..c400e93bbf79 100644
--- a/fs/ocfs2/dlm/userdlm.h
+++ b/fs/ocfs2/dlm/userdlm.h
@@ -53,6 +53,7 @@ struct user_lock_res {
 #define USER_DLM_LOCK_ID_MAX_LEN  32
        char                     l_name[USER_DLM_LOCK_ID_MAX_LEN];
+        int                      l_namelen;
        int                      l_level;
        unsigned int             l_ro_holders;
        unsigned int             l_ex_holders;
diff --git a/fs/ocfs2/dlmglue.c b/fs/ocfs2/dlmglue.c
index 762eb1fbb34d..de887063dcfc 100644
--- a/fs/ocfs2/dlmglue.c
+++ b/fs/ocfs2/dlmglue.c
@@ -46,6 +46,7 @@
 #include "ocfs2.h"
 #include "alloc.h"
+#include "dcache.h"
 #include "dlmglue.h"
 #include "extent_map.h"
 #include "heartbeat.h"
@@ -66,78 +67,161 @@ struct ocfs2_mask_waiter {
        unsigned long           mw_goal;
 };
-static void ocfs2_inode_ast_func(void *opaque);
+static struct ocfs2_super *ocfs2_get_dentry_osb(struct ocfs2_lock_res *lockres);
-static void ocfs2_inode_bast_func(void *opaque,
+static struct ocfs2_super *ocfs2_get_inode_osb(struct ocfs2_lock_res *lockres);
-                                  int level);
-static void ocfs2_super_ast_func(void *opaque);
-static void ocfs2_super_bast_func(void *opaque,
-                                  int level);
-static void ocfs2_rename_ast_func(void *opaque);
-static void ocfs2_rename_bast_func(void *opaque,
-                                   int level);
-/* so far, all locks have gotten along with the same unlock ast */
-static void ocfs2_unlock_ast_func(void *opaque,
-                                  enum dlm_status status);
-static int ocfs2_do_unblock_meta(struct inode *inode,
-                                 int *requeue);
-static int ocfs2_unblock_meta(struct ocfs2_lock_res *lockres,
-                              int *requeue);
-static int ocfs2_unblock_data(struct ocfs2_lock_res *lockres,
-                              int *requeue);
-static int ocfs2_unblock_inode_lock(struct ocfs2_lock_res *lockres,
-                              int *requeue);
-static int ocfs2_unblock_osb_lock(struct ocfs2_lock_res *lockres,
-                                  int *requeue);
-typedef void (ocfs2_convert_worker_t)(struct ocfs2_lock_res *, int);
-static int ocfs2_generic_unblock_lock(struct ocfs2_super *osb,
-                                      struct ocfs2_lock_res *lockres,
-                                      int *requeue,
-                                      ocfs2_convert_worker_t *worker);
+/*
+ * Return value from ->downconvert_worker functions.
+ *
+ * These control the precise actions of ocfs2_unblock_lock()
+ * and ocfs2_process_blocked_lock()
+ *
+ */
+enum ocfs2_unblock_action {
+        UNBLOCK_CONTINUE        = 0, /* Continue downconvert */
+        UNBLOCK_CONTINUE_POST   = 1, /* Continue downconvert, fire
+                                      * ->post_unlock callback */
+        UNBLOCK_STOP_POST       = 2, /* Do not downconvert, fire
+                                      * ->post_unlock() callback. */
+};
+struct ocfs2_unblock_ctl {
+        int requeue;
+        enum ocfs2_unblock_action unblock_action;
+};
+static int ocfs2_check_meta_downconvert(struct ocfs2_lock_res *lockres,
+                                        int new_level);
+static void ocfs2_set_meta_lvb(struct ocfs2_lock_res *lockres);
+static int ocfs2_data_convert_worker(struct ocfs2_lock_res *lockres,
+                                     int blocking);
+static int ocfs2_dentry_convert_worker(struct ocfs2_lock_res *lockres,
+                                       int blocking);
+static void ocfs2_dentry_post_unlock(struct ocfs2_super *osb,
+                                     struct ocfs2_lock_res *lockres);
+/*
+ * OCFS2 Lock Resource Operations
+ *
+ * These fine tune the behavior of the generic dlmglue locking infrastructure.
+ *
+ * The most basic of lock types can point ->l_priv to their respective
+ * struct ocfs2_super and allow the default actions to manage things.
+ *
+ * Right now, each lock type also needs to implement an init function,
+ * and trivial lock/unlock wrappers. ocfs2_simple_drop_lockres()
+ * should be called when the lock is no longer needed (i.e., object
+ * destruction time).
+ */
 struct ocfs2_lock_res_ops {
-        void (*ast)(void *);
+        /*
-        void (*bast)(void *, int);
+         * Translate an ocfs2_lock_res * into an ocfs2_super *. Define
-        void (*unlock_ast)(void *, enum dlm_status);
+         * this callback if ->l_priv is not an ocfs2_super pointer
-        int  (*unblock)(struct ocfs2_lock_res *, int *);
+         */
+        struct ocfs2_super * (*get_osb)(struct ocfs2_lock_res *);
+        /*
+         * Optionally called in the downconvert (or "vote") thread
+         * after a successful downconvert. The lockres will not be
+         * referenced after this callback is called, so it is safe to
+         * free memory, etc.
+         *
+         * The exact semantics of when this is called are controlled
+         * by ->downconvert_worker()
+         */
+        void (*post_unlock)(struct ocfs2_super *, struct ocfs2_lock_res *);
+        /*
+         * Allow a lock type to add checks to determine whether it is
+         * safe to downconvert a lock. Return 0 to re-queue the
+         * downconvert at a later time, nonzero to continue.
+         *
+         * For most locks, the default checks that there are no
+         * incompatible holders are sufficient.
+         *
+         * Called with the lockres spinlock held.
+         */
+        int (*check_downconvert)(struct ocfs2_lock_res *, int);
+        /*
+         * Allows a lock type to populate the lock value block. This
+         * is called on downconvert, and when we drop a lock.
+         *
+         * Locks that want to use this should set LOCK_TYPE_USES_LVB
+         * in the flags field.
+         *
+         * Called with the lockres spinlock held.
+         */
+        void (*set_lvb)(struct ocfs2_lock_res *);
+        /*
+         * Called from the downconvert thread when it is determined
+         * that a lock will be downconverted. This is called without
+         * any locks held so the function can do work that might
+         * schedule (syncing out data, etc).
+         *
+         * This should return any one of the ocfs2_unblock_action
+         * values, depending on what it wants the thread to do.
+         */
+        int (*downconvert_worker)(struct ocfs2_lock_res *, int);
+        /*
+         * LOCK_TYPE_* flags which describe the specific requirements
+         * of a lock type. Descriptions of each individual flag follow.
+         */
+        int flags;
 };
+/*
+ * Some locks want to "refresh" potentially stale data when a
+ * meaningful (PRMODE or EXMODE) lock level is first obtained. If this
+ * flag is set, the OCFS2_LOCK_NEEDS_REFRESH flag will be set on the
+ * individual lockres l_flags member from the ast function. It is
+ * expected that the locking wrapper will clear the
+ * OCFS2_LOCK_NEEDS_REFRESH flag when done.
+ */
+#define LOCK_TYPE_REQUIRES_REFRESH 0x1
+/*
+ * Indicate that a lock type makes use of the lock value block. The
+ * ->set_lvb lock type callback must be defined.
+ */
+#define LOCK_TYPE_USES_LVB              0x2
 static struct ocfs2_lock_res_ops ocfs2_inode_rw_lops = {
-        .ast            = ocfs2_inode_ast_func,
+        .get_osb        = ocfs2_get_inode_osb,
-        .bast           = ocfs2_inode_bast_func,
+        .flags          = 0,
-        .unlock_ast     = ocfs2_unlock_ast_func,
-        .unblock        = ocfs2_unblock_inode_lock,
 };
 static struct ocfs2_lock_res_ops ocfs2_inode_meta_lops = {
-        .ast            = ocfs2_inode_ast_func,
+        .get_osb        = ocfs2_get_inode_osb,
-        .bast           = ocfs2_inode_bast_func,
+        .check_downconvert = ocfs2_check_meta_downconvert,
-        .unlock_ast     = ocfs2_unlock_ast_func,
+        .set_lvb        = ocfs2_set_meta_lvb,
-        .unblock        = ocfs2_unblock_meta,
+        .flags          = LOCK_TYPE_REQUIRES_REFRESH|LOCK_TYPE_USES_LVB,
 };
-static void ocfs2_data_convert_worker(struct ocfs2_lock_res *lockres,
-                                      int blocking);
 static struct ocfs2_lock_res_ops ocfs2_inode_data_lops = {
-        .ast            = ocfs2_inode_ast_func,
+        .get_osb        = ocfs2_get_inode_osb,
-        .bast           = ocfs2_inode_bast_func,
+        .downconvert_worker = ocfs2_data_convert_worker,
-        .unlock_ast     = ocfs2_unlock_ast_func,
+        .flags          = 0,
-        .unblock        = ocfs2_unblock_data,
 };
 static struct ocfs2_lock_res_ops ocfs2_super_lops = {
-        .ast            = ocfs2_super_ast_func,
+        .flags          = LOCK_TYPE_REQUIRES_REFRESH,
-        .bast           = ocfs2_super_bast_func,
-        .unlock_ast     = ocfs2_unlock_ast_func,
-        .unblock        = ocfs2_unblock_osb_lock,
 };
 static struct ocfs2_lock_res_ops ocfs2_rename_lops = {
-        .ast            = ocfs2_rename_ast_func,
+        .flags          = 0,
-        .bast           = ocfs2_rename_bast_func,
+};
-        .unlock_ast     = ocfs2_unlock_ast_func,
-        .unblock        = ocfs2_unblock_osb_lock,
+static struct ocfs2_lock_res_ops ocfs2_dentry_lops = {
+        .get_osb        = ocfs2_get_dentry_osb,
+        .post_unlock    = ocfs2_dentry_post_unlock,
+        .downconvert_worker = ocfs2_dentry_convert_worker,
+        .flags          = 0,
 };
 static inline int ocfs2_is_inode_lock(struct ocfs2_lock_res *lockres)
@@ -147,29 +231,26 @@ static inline int ocfs2_is_inode_lock(struct ocfs2_lock_res *lockres)
                lockres->l_type == OCFS2_LOCK_TYPE_RW;
 }
-static inline int ocfs2_is_super_lock(struct ocfs2_lock_res *lockres)
+static inline struct inode *ocfs2_lock_res_inode(struct ocfs2_lock_res *lockres)
 {
-        return lockres->l_type == OCFS2_LOCK_TYPE_SUPER;
+        BUG_ON(!ocfs2_is_inode_lock(lockres));
-}
-static inline int ocfs2_is_rename_lock(struct ocfs2_lock_res *lockres)
+        return (struct inode *) lockres->l_priv;
-{
-        return lockres->l_type == OCFS2_LOCK_TYPE_RENAME;
 }
-static inline struct ocfs2_super *ocfs2_lock_res_super(struct ocfs2_lock_res *lockres)
+static inline struct ocfs2_dentry_lock *ocfs2_lock_res_dl(struct ocfs2_lock_res *lockres)
 {
-        BUG_ON(!ocfs2_is_super_lock(lockres)
+        BUG_ON(lockres->l_type != OCFS2_LOCK_TYPE_DENTRY);
-               && !ocfs2_is_rename_lock(lockres));
-        return (struct ocfs2_super *) lockres->l_priv;
+        return (struct ocfs2_dentry_lock *)lockres->l_priv;
 }
-static inline struct inode *ocfs2_lock_res_inode(struct ocfs2_lock_res *lockres)
+static inline struct ocfs2_super *ocfs2_get_lockres_osb(struct ocfs2_lock_res *lockres)
 {
-        BUG_ON(!ocfs2_is_inode_lock(lockres));
+        if (lockres->l_ops->get_osb)
+                return lockres->l_ops->get_osb(lockres);
-        return (struct inode *) lockres->l_priv;
+        return (struct ocfs2_super *)lockres->l_priv;
 }
 static int ocfs2_lock_create(struct ocfs2_super *osb,
@@ -200,25 +281,6 @@ static int ocfs2_meta_lock_update(struct inode *inode,
                                  struct buffer_head **bh);
 static void ocfs2_drop_osb_locks(struct ocfs2_super *osb);
 static inline int ocfs2_highest_compat_lock_level(int level);
-static inline int ocfs2_can_downconvert_meta_lock(struct inode *inode,
-                                                  struct ocfs2_lock_res *lockres,
-                                                  int new_level);
-static char *ocfs2_lock_type_strings[] = {
-        [OCFS2_LOCK_TYPE_META] = "Meta",
-        [OCFS2_LOCK_TYPE_DATA] = "Data",
-        [OCFS2_LOCK_TYPE_SUPER] = "Super",
-        [OCFS2_LOCK_TYPE_RENAME] = "Rename",
-        /* Need to differntiate from [R]ename.. serializing writes is the
-         * important job it does, anyway. */
-        [OCFS2_LOCK_TYPE_RW] = "Write/Read",
-};
-static char *ocfs2_lock_type_string(enum ocfs2_lock_type type)
-{
-        mlog_bug_on_msg(type >= OCFS2_NUM_LOCK_TYPES, "%d\n", type);
-        return ocfs2_lock_type_strings[type];
-}
 static void ocfs2_build_lock_name(enum ocfs2_lock_type type,
                                  u64 blkno,
@@ -265,13 +327,9 @@ static void ocfs2_remove_lockres_tracking(struct ocfs2_lock_res *res)
 static void ocfs2_lock_res_init_common(struct ocfs2_super *osb,
                                       struct ocfs2_lock_res *res,
                                       enum ocfs2_lock_type type,
-                                       u64 blkno,
-                                       u32 generation,
                                       struct ocfs2_lock_res_ops *ops,
                                       void *priv)
 {
-        ocfs2_build_lock_name(type, blkno, generation, res->l_name);
        res->l_type          = type;
        res->l_ops           = ops;
        res->l_priv          = priv;
@@ -299,6 +357,7 @@ void ocfs2_lock_res_init_once(struct ocfs2_lock_res *res)
 void ocfs2_inode_lock_res_init(struct ocfs2_lock_res *res,
                               enum ocfs2_lock_type type,
+                               unsigned int generation,
                               struct inode *inode)
 {
        struct ocfs2_lock_res_ops *ops;
@@ -319,9 +378,73 @@ void ocfs2_inode_lock_res_init(struct ocfs2_lock_res *res,
                        break;
        };
-        ocfs2_lock_res_init_common(OCFS2_SB(inode->i_sb), res, type,
+        ocfs2_build_lock_name(type, OCFS2_I(inode)->ip_blkno,
-                                   OCFS2_I(inode)->ip_blkno,
+                              generation, res->l_name);
-                                   inode->i_generation, ops, inode);
+        ocfs2_lock_res_init_common(OCFS2_SB(inode->i_sb), res, type, ops, inode);
+}
+static struct ocfs2_super *ocfs2_get_inode_osb(struct ocfs2_lock_res *lockres)
+{
+        struct inode *inode = ocfs2_lock_res_inode(lockres);
+        return OCFS2_SB(inode->i_sb);
+}
+static __u64 ocfs2_get_dentry_lock_ino(struct ocfs2_lock_res *lockres)
+{
+        __be64 inode_blkno_be;
+        memcpy(&inode_blkno_be, &lockres->l_name[OCFS2_DENTRY_LOCK_INO_START],
+               sizeof(__be64));
+        return be64_to_cpu(inode_blkno_be);
+}
+static struct ocfs2_super *ocfs2_get_dentry_osb(struct ocfs2_lock_res *lockres)
+{
+        struct ocfs2_dentry_lock *dl = lockres->l_priv;
+        return OCFS2_SB(dl->dl_inode->i_sb);
+}
+void ocfs2_dentry_lock_res_init(struct ocfs2_dentry_lock *dl,
+                                u64 parent, struct inode *inode)
+{
+        int len;
+        u64 inode_blkno = OCFS2_I(inode)->ip_blkno;
+        __be64 inode_blkno_be = cpu_to_be64(inode_blkno);
+        struct ocfs2_lock_res *lockres = &dl->dl_lockres;
+        ocfs2_lock_res_init_once(lockres);
+        /*
+         * Unfortunately, the standard lock naming scheme won't work
+         * here because we have two 16 byte values to use. Instead,
+         * we'll stuff the inode number as a binary value. We still
+         * want error prints to show something without garbling the
+         * display, so drop a null byte in there before the inode
+         * number. A future version of OCFS2 will likely use all
+         * binary lock names. The stringified names have been a
+         * tremendous aid in debugging, but now that the debugfs
+         * interface exists, we can mangle things there if need be.
+         *
+         * NOTE: We also drop the standard "pad" value (the total lock
+         * name size stays the same though - the last part is all
+         * zeros due to the memset in ocfs2_lock_res_init_once()
+         */
+        len = snprintf(lockres->l_name, OCFS2_DENTRY_LOCK_INO_START,
+                       "%c%016llx",
+                       ocfs2_lock_type_char(OCFS2_LOCK_TYPE_DENTRY),
+                       (long long)parent);
+        BUG_ON(len != (OCFS2_DENTRY_LOCK_INO_START - 1));
+        memcpy(&lockres->l_name[OCFS2_DENTRY_LOCK_INO_START], &inode_blkno_be,
+               sizeof(__be64));
+        ocfs2_lock_res_init_common(OCFS2_SB(inode->i_sb), lockres,
+                                   OCFS2_LOCK_TYPE_DENTRY, &ocfs2_dentry_lops,
+                                   dl);
 }
 static void ocfs2_super_lock_res_init(struct ocfs2_lock_res *res,
@@ -330,8 +453,9 @@ static void ocfs2_super_lock_res_init(struct ocfs2_lock_res *res,
        /* Superblock lockres doesn't come from a slab so we call init
         * once on it manually.  */
        ocfs2_lock_res_init_once(res);
+        ocfs2_build_lock_name(OCFS2_LOCK_TYPE_SUPER, OCFS2_SUPER_BLOCK_BLKNO,
+                              0, res->l_name);
        ocfs2_lock_res_init_common(osb, res, OCFS2_LOCK_TYPE_SUPER,
-                                   OCFS2_SUPER_BLOCK_BLKNO, 0,
                                   &ocfs2_super_lops, osb);
 }
@@ -341,7 +465,8 @@ static void ocfs2_rename_lock_res_init(struct ocfs2_lock_res *res,
        /* Rename lockres doesn't come from a slab so we call init
         * once on it manually.  */
        ocfs2_lock_res_init_once(res);
-        ocfs2_lock_res_init_common(osb, res, OCFS2_LOCK_TYPE_RENAME, 0, 0,
+        ocfs2_build_lock_name(OCFS2_LOCK_TYPE_RENAME, 0, 0, res->l_name);
+        ocfs2_lock_res_init_common(osb, res, OCFS2_LOCK_TYPE_RENAME,
                                   &ocfs2_rename_lops, osb);
 }
@@ -495,7 +620,8 @@ static inline void ocfs2_generic_handle_convert_action(struct ocfs2_lock_res *lo
         * information is already up to data. Convert from NL to
         * *anything* however should mark ourselves as needing an
         * update */
-        if (lockres->l_level == LKM_NLMODE)
+        if (lockres->l_level == LKM_NLMODE &&
+            lockres->l_ops->flags & LOCK_TYPE_REQUIRES_REFRESH)
                lockres_or_flags(lockres, OCFS2_LOCK_NEEDS_REFRESH);
        lockres->l_level = lockres->l_requested;
@@ -512,7 +638,8 @@ static inline void ocfs2_generic_handle_attach_action(struct ocfs2_lock_res *loc
        BUG_ON(lockres->l_flags & OCFS2_LOCK_ATTACHED);
        if (lockres->l_requested > LKM_NLMODE &&
-            !(lockres->l_flags & OCFS2_LOCK_LOCAL))
+            !(lockres->l_flags & OCFS2_LOCK_LOCAL) &&
+            lockres->l_ops->flags & LOCK_TYPE_REQUIRES_REFRESH)
                lockres_or_flags(lockres, OCFS2_LOCK_NEEDS_REFRESH);
        lockres->l_level = lockres->l_requested;
@@ -522,68 +649,6 @@ static inline void ocfs2_generic_handle_attach_action(struct ocfs2_lock_res *loc
        mlog_exit_void();
 }
-static void ocfs2_inode_ast_func(void *opaque)
-{
-        struct ocfs2_lock_res *lockres = opaque;
-        struct inode *inode;
-        struct dlm_lockstatus *lksb;
-        unsigned long flags;
-        mlog_entry_void();
-        inode = ocfs2_lock_res_inode(lockres);
-        mlog(0, "AST fired for inode %llu, l_action = %u, type = %s\n",
-             (unsigned long long)OCFS2_I(inode)->ip_blkno, lockres->l_action,
-             ocfs2_lock_type_string(lockres->l_type));
-        BUG_ON(!ocfs2_is_inode_lock(lockres));
-        spin_lock_irqsave(&lockres->l_lock, flags);
-        lksb = &(lockres->l_lksb);
-        if (lksb->status != DLM_NORMAL) {
-                mlog(ML_ERROR, "ocfs2_inode_ast_func: lksb status value of %u "
-                     "on inode %llu\n", lksb->status,
-                     (unsigned long long)OCFS2_I(inode)->ip_blkno);
-                spin_unlock_irqrestore(&lockres->l_lock, flags);
-                mlog_exit_void();
-                return;
-        }
-        switch(lockres->l_action) {
-        case OCFS2_AST_ATTACH:
-                ocfs2_generic_handle_attach_action(lockres);
-                lockres_clear_flags(lockres, OCFS2_LOCK_LOCAL);
-                break;
-        case OCFS2_AST_CONVERT:
-                ocfs2_generic_handle_convert_action(lockres);
-                break;
-        case OCFS2_AST_DOWNCONVERT:
-                ocfs2_generic_handle_downconvert_action(lockres);
-                break;
-        default:
-                mlog(ML_ERROR, "lockres %s: ast fired with invalid action: %u "
-                     "lockres flags = 0x%lx, unlock action: %u\n",
-                     lockres->l_name, lockres->l_action, lockres->l_flags,
-                     lockres->l_unlock_action);
-                BUG();
-        }
-        /* data and rw locking ignores refresh flag for now. */
-        if (lockres->l_type != OCFS2_LOCK_TYPE_META)
-                lockres_clear_flags(lockres, OCFS2_LOCK_NEEDS_REFRESH);
-        /* set it to something invalid so if we get called again we
-         * can catch it. */
-        lockres->l_action = OCFS2_AST_INVALID;
-        spin_unlock_irqrestore(&lockres->l_lock, flags);
-        wake_up(&lockres->l_event);
-        mlog_exit_void();
-}
 static int ocfs2_generic_handle_bast(struct ocfs2_lock_res *lockres,
                                     int level)
 {
@@ -610,54 +675,33 @@ static int ocfs2_generic_handle_bast(struct ocfs2_lock_res *lockres,
        return needs_downconvert;
 }
-static void ocfs2_generic_bast_func(struct ocfs2_super *osb,
+static void ocfs2_blocking_ast(void *opaque, int level)
-                                    struct ocfs2_lock_res *lockres,
-                                    int level)
 {
+        struct ocfs2_lock_res *lockres = opaque;
+        struct ocfs2_super *osb = ocfs2_get_lockres_osb(lockres);
        int needs_downconvert;
        unsigned long flags;
-        mlog_entry_void();
        BUG_ON(level <= LKM_NLMODE);
+        mlog(0, "BAST fired for lockres %s, blocking %d, level %d type %s\n",
+             lockres->l_name, level, lockres->l_level,
+             ocfs2_lock_type_string(lockres->l_type));
        spin_lock_irqsave(&lockres->l_lock, flags);
        needs_downconvert = ocfs2_generic_handle_bast(lockres, level);
        if (needs_downconvert)
                ocfs2_schedule_blocked_lock(osb, lockres);
        spin_unlock_irqrestore(&lockres->l_lock, flags);
-        ocfs2_kick_vote_thread(osb);
        wake_up(&lockres->l_event);
-        mlog_exit_void();
-}
-static void ocfs2_inode_bast_func(void *opaque, int level)
-{
-        struct ocfs2_lock_res *lockres = opaque;
-        struct inode *inode;
-        struct ocfs2_super *osb;
-        mlog_entry_void();
-        BUG_ON(!ocfs2_is_inode_lock(lockres));
-        inode = ocfs2_lock_res_inode(lockres);
+        ocfs2_kick_vote_thread(osb);
-        osb = OCFS2_SB(inode->i_sb);
-        mlog(0, "BAST fired for inode %llu, blocking %d, level %d type %s\n",
-             (unsigned long long)OCFS2_I(inode)->ip_blkno, level,
-             lockres->l_level, ocfs2_lock_type_string(lockres->l_type));
-        ocfs2_generic_bast_func(osb, lockres, level);
-        mlog_exit_void();
 }
-static void ocfs2_generic_ast_func(struct ocfs2_lock_res *lockres,
+static void ocfs2_locking_ast(void *opaque)
-                                   int ignore_refresh)
 {
+        struct ocfs2_lock_res *lockres = opaque;
        struct dlm_lockstatus *lksb = &lockres->l_lksb;
        unsigned long flags;
@@ -673,6 +717,7 @@ static void ocfs2_generic_ast_func(struct ocfs2_lock_res *lockres,
        switch(lockres->l_action) {
        case OCFS2_AST_ATTACH:
                ocfs2_generic_handle_attach_action(lockres);
+                lockres_clear_flags(lockres, OCFS2_LOCK_LOCAL);
                break;
        case OCFS2_AST_CONVERT:
                ocfs2_generic_handle_convert_action(lockres);
@@ -681,80 +726,19 @@ static void ocfs2_generic_ast_func(struct ocfs2_lock_res *lockres,
                ocfs2_generic_handle_downconvert_action(lockres);
                break;
        default:
+                mlog(ML_ERROR, "lockres %s: ast fired with invalid action: %u "
+                     "lockres flags = 0x%lx, unlock action: %u\n",
+                     lockres->l_name, lockres->l_action, lockres->l_flags,
+                     lockres->l_unlock_action);
                BUG();
        }
-        if (ignore_refresh)
-                lockres_clear_flags(lockres, OCFS2_LOCK_NEEDS_REFRESH);
        /* set it to something invalid so if we get called again we
         * can catch it. */
        lockres->l_action = OCFS2_AST_INVALID;
-        spin_unlock_irqrestore(&lockres->l_lock, flags);
        wake_up(&lockres->l_event);
-}
+        spin_unlock_irqrestore(&lockres->l_lock, flags);
-static void ocfs2_super_ast_func(void *opaque)
-{
-        struct ocfs2_lock_res *lockres = opaque;
-        mlog_entry_void();
-        mlog(0, "Superblock AST fired\n");
-        BUG_ON(!ocfs2_is_super_lock(lockres));
-        ocfs2_generic_ast_func(lockres, 0);
-        mlog_exit_void();
-}
-static void ocfs2_super_bast_func(void *opaque,
-                                  int level)
-{
-        struct ocfs2_lock_res *lockres = opaque;
-        struct ocfs2_super *osb;
-        mlog_entry_void();
-        mlog(0, "Superblock BAST fired\n");
-        BUG_ON(!ocfs2_is_super_lock(lockres));
-        osb = ocfs2_lock_res_super(lockres);
-        ocfs2_generic_bast_func(osb, lockres, level);
-        mlog_exit_void();
-}
-static void ocfs2_rename_ast_func(void *opaque)
-{
-        struct ocfs2_lock_res *lockres = opaque;
-        mlog_entry_void();
-        mlog(0, "Rename AST fired\n");
-        BUG_ON(!ocfs2_is_rename_lock(lockres));
-        ocfs2_generic_ast_func(lockres, 1);
-        mlog_exit_void();
-}
-static void ocfs2_rename_bast_func(void *opaque,
-                                   int level)
-{
-        struct ocfs2_lock_res *lockres = opaque;
-        struct ocfs2_super *osb;
-        mlog_entry_void();
-        mlog(0, "Rename BAST fired\n");
-        BUG_ON(!ocfs2_is_rename_lock(lockres));
-        osb = ocfs2_lock_res_super(lockres);
-        ocfs2_generic_bast_func(osb, lockres, level);
-        mlog_exit_void();
 }
 static inline void ocfs2_recover_from_dlm_error(struct ocfs2_lock_res *lockres,
@@ -810,9 +794,10 @@ static int ocfs2_lock_create(struct ocfs2_super *osb,
                         &lockres->l_lksb,
                         dlm_flags,
                         lockres->l_name,
-                         lockres->l_ops->ast,
+                         OCFS2_LOCK_ID_MAX_LEN - 1,
+                         ocfs2_locking_ast,
                         lockres,
-                         lockres->l_ops->bast);
+                         ocfs2_blocking_ast);
        if (status != DLM_NORMAL) {
                ocfs2_log_dlm_error("dlmlock", status, lockres);
                ret = -EINVAL;
@@ -930,6 +915,9 @@ static int ocfs2_cluster_lock(struct ocfs2_super *osb,
        ocfs2_init_mask_waiter(&mw);
+        if (lockres->l_ops->flags & LOCK_TYPE_USES_LVB)
+                lkm_flags |= LKM_VALBLK;
 again:
        wait = 0;
@@ -997,11 +985,12 @@ again:
                status = dlmlock(osb->dlm,
                                 level,
                                 &lockres->l_lksb,
-                                 lkm_flags|LKM_CONVERT|LKM_VALBLK,
+                                 lkm_flags|LKM_CONVERT,
                                 lockres->l_name,
-                                 lockres->l_ops->ast,
+                                 OCFS2_LOCK_ID_MAX_LEN - 1,
+                                 ocfs2_locking_ast,
                                 lockres,
-                                 lockres->l_ops->bast);
+                                 ocfs2_blocking_ast);
                if (status != DLM_NORMAL) {
                        if ((lkm_flags & LKM_NOQUEUE) &&
                            (status == DLM_NOTQUEUED))
@@ -1074,18 +1063,21 @@ static void ocfs2_cluster_unlock(struct ocfs2_super *osb,
        mlog_exit_void();
 }
-static int ocfs2_create_new_inode_lock(struct inode *inode,
+int ocfs2_create_new_lock(struct ocfs2_super *osb,
-                                       struct ocfs2_lock_res *lockres)
+                          struct ocfs2_lock_res *lockres,
+                          int ex,
+                          int local)
 {
-        struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
+        int level =  ex ? LKM_EXMODE : LKM_PRMODE;
        unsigned long flags;
+        int lkm_flags = local ? LKM_LOCAL : 0;
        spin_lock_irqsave(&lockres->l_lock, flags);
        BUG_ON(lockres->l_flags & OCFS2_LOCK_ATTACHED);
        lockres_or_flags(lockres, OCFS2_LOCK_LOCAL);
        spin_unlock_irqrestore(&lockres->l_lock, flags);
-        return ocfs2_lock_create(osb, lockres, LKM_EXMODE, LKM_LOCAL);
+        return ocfs2_lock_create(osb, lockres, level, lkm_flags);
 }
 /* Grants us an EX lock on the data and metadata resources, skipping
@@ -1097,6 +1089,7 @@ static int ocfs2_create_new_inode_lock(struct inode *inode,
 int ocfs2_create_new_inode_locks(struct inode *inode)
 {
        int ret;
+        struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
        BUG_ON(!inode);
        BUG_ON(!ocfs2_inode_is_new(inode));
@@ -1113,22 +1106,23 @@ int ocfs2_create_new_inode_locks(struct inode *inode)
         * on a resource which has an invalid one -- we'll set it
         * valid when we release the EX. */
-        ret = ocfs2_create_new_inode_lock(inode,
+        ret = ocfs2_create_new_lock(osb, &OCFS2_I(inode)->ip_rw_lockres, 1, 1);
-                                          &OCFS2_I(inode)->ip_rw_lockres);
        if (ret) {
                mlog_errno(ret);
                goto bail;
        }
-        ret = ocfs2_create_new_inode_lock(inode,
+        /*
-                                          &OCFS2_I(inode)->ip_meta_lockres);
+         * We don't want to use LKM_LOCAL on a meta data lock as they
+         * don't use a generation in their lock names.
+         */
+        ret = ocfs2_create_new_lock(osb, &OCFS2_I(inode)->ip_meta_lockres, 1, 0);
        if (ret) {
                mlog_errno(ret);
                goto bail;
        }
-        ret = ocfs2_create_new_inode_lock(inode,
+        ret = ocfs2_create_new_lock(osb, &OCFS2_I(inode)->ip_data_lockres, 1, 1);
-                                          &OCFS2_I(inode)->ip_data_lockres);
        if (ret) {
                mlog_errno(ret);
                goto bail;
@@ -1317,7 +1311,17 @@ static void __ocfs2_stuff_meta_lvb(struct inode *inode)
        lvb = (struct ocfs2_meta_lvb *) lockres->l_lksb.lvb;
-        lvb->lvb_version   = cpu_to_be32(OCFS2_LVB_VERSION);
+        /*
+         * Invalidate the LVB of a deleted inode - this way other
+         * nodes are forced to go to disk and discover the new inode
+         * status.
+         */
+        if (oi->ip_flags & OCFS2_INODE_DELETED) {
+                lvb->lvb_version = 0;
+                goto out;
+        }
+        lvb->lvb_version   = OCFS2_LVB_VERSION;
        lvb->lvb_isize     = cpu_to_be64(i_size_read(inode));
        lvb->lvb_iclusters = cpu_to_be32(oi->ip_clusters);
        lvb->lvb_iuid      = cpu_to_be32(inode->i_uid);
@@ -1330,7 +1334,10 @@ static void __ocfs2_stuff_meta_lvb(struct inode *inode)
                cpu_to_be64(ocfs2_pack_timespec(&inode->i_ctime));
        lvb->lvb_imtime_packed =
                cpu_to_be64(ocfs2_pack_timespec(&inode->i_mtime));
+        lvb->lvb_iattr    = cpu_to_be32(oi->ip_attr);
+        lvb->lvb_igeneration = cpu_to_be32(inode->i_generation);
+out:
        mlog_meta_lvb(0, lockres);
        mlog_exit_void();
@@ -1360,6 +1367,9 @@ static void ocfs2_refresh_inode_from_lvb(struct inode *inode)
        oi->ip_clusters = be32_to_cpu(lvb->lvb_iclusters);
        i_size_write(inode, be64_to_cpu(lvb->lvb_isize));
+        oi->ip_attr = be32_to_cpu(lvb->lvb_iattr);
+        ocfs2_set_inode_flags(inode);
        /* fast-symlinks are a special case */
        if (S_ISLNK(inode->i_mode) && !oi->ip_clusters)
                inode->i_blocks = 0;
@@ -1382,11 +1392,13 @@ static void ocfs2_refresh_inode_from_lvb(struct inode *inode)
        mlog_exit_void();
 }
-static inline int ocfs2_meta_lvb_is_trustable(struct ocfs2_lock_res *lockres)
+static inline int ocfs2_meta_lvb_is_trustable(struct inode *inode,
+                                              struct ocfs2_lock_res *lockres)
 {
        struct ocfs2_meta_lvb *lvb = (struct ocfs2_meta_lvb *) lockres->l_lksb.lvb;
-        if (be32_to_cpu(lvb->lvb_version) == OCFS2_LVB_VERSION)
+        if (lvb->lvb_version == OCFS2_LVB_VERSION
+            && be32_to_cpu(lvb->lvb_igeneration) == inode->i_generation)
                return 1;
        return 0;
 }
@@ -1483,7 +1495,7 @@ static int ocfs2_meta_lock_update(struct inode *inode,
         * map (directories, bitmap files, etc) */
        ocfs2_extent_map_trunc(inode, 0);
-        if (ocfs2_meta_lvb_is_trustable(lockres)) {
+        if (ocfs2_meta_lvb_is_trustable(inode, lockres)) {
                mlog(0, "Trusting LVB on inode %llu\n",
                     (unsigned long long)oi->ip_blkno);
                ocfs2_refresh_inode_from_lvb(inode);
@@ -1624,6 +1636,18 @@ int ocfs2_meta_lock_full(struct inode *inode,
                wait_event(osb->recovery_event,
                           ocfs2_node_map_is_empty(osb, &osb->recovery_map));
+        /*
+         * We only see this flag if we're being called from
+         * ocfs2_read_locked_inode(). It means we're locking an inode
+         * which hasn't been populated yet, so clear the refresh flag
+         * and let the caller handle it.
+         */
+        if (inode->i_state & I_NEW) {
+                status = 0;
+                ocfs2_complete_lock_res_refresh(lockres, 0);
+                goto bail;
+        }
        /* This is fun. The caller may want a bh back, or it may
         * not. ocfs2_meta_lock_update definitely wants one in, but
         * may or may not read one, depending on what's in the
@@ -1803,6 +1827,34 @@ void ocfs2_rename_unlock(struct ocfs2_super *osb)
        ocfs2_cluster_unlock(osb, lockres, LKM_EXMODE);
 }
+int ocfs2_dentry_lock(struct dentry *dentry, int ex)
+{
+        int ret;
+        int level = ex ? LKM_EXMODE : LKM_PRMODE;
+        struct ocfs2_dentry_lock *dl = dentry->d_fsdata;
+        struct ocfs2_super *osb = OCFS2_SB(dentry->d_sb);
+        BUG_ON(!dl);
+        if (ocfs2_is_hard_readonly(osb))
+                return -EROFS;
+        ret = ocfs2_cluster_lock(osb, &dl->dl_lockres, level, 0, 0);
+        if (ret < 0)
+                mlog_errno(ret);
+        return ret;
+}
+void ocfs2_dentry_unlock(struct dentry *dentry, int ex)
+{
+        int level = ex ? LKM_EXMODE : LKM_PRMODE;
+        struct ocfs2_dentry_lock *dl = dentry->d_fsdata;
+        struct ocfs2_super *osb = OCFS2_SB(dentry->d_sb);
+        ocfs2_cluster_unlock(osb, &dl->dl_lockres, level);
+}
 /* Reference counting of the dlm debug structure. We want this because
 * open references on the debug inodes can live on after a mount, so
 * we can't rely on the ocfs2_super to always exist. */
@@ -1933,9 +1985,16 @@ static int ocfs2_dlm_seq_show(struct seq_file *m, void *v)
        if (!lockres)
                return -EINVAL;
-        seq_printf(m, "0x%x\t"
+        seq_printf(m, "0x%x\t", OCFS2_DLM_DEBUG_STR_VERSION);
-                   "%.*s\t"
-                   "%d\t"
+        if (lockres->l_type == OCFS2_LOCK_TYPE_DENTRY)
+                seq_printf(m, "%.*s%08x\t", OCFS2_DENTRY_LOCK_INO_START - 1,
+                           lockres->l_name,
+                           (unsigned int)ocfs2_get_dentry_lock_ino(lockres));
+        else
+                seq_printf(m, "%.*s\t", OCFS2_LOCK_ID_MAX_LEN, lockres->l_name);
+        seq_printf(m, "%d\t"
                   "0x%lx\t"
                   "0x%x\t"
                   "0x%x\t"
@@ -1943,8 +2002,6 @@ static int ocfs2_dlm_seq_show(struct seq_file *m, void *v)
                   "%u\t"
                   "%d\t"
                   "%d\t",
-                   OCFS2_DLM_DEBUG_STR_VERSION,
-                   OCFS2_LOCK_ID_MAX_LEN, lockres->l_name,
                   lockres->l_level,
                   lockres->l_flags,
                   lockres->l_action,
@@ -2134,7 +2191,7 @@ void ocfs2_dlm_shutdown(struct ocfs2_super *osb)
        mlog_exit_void();
 }
-static void ocfs2_unlock_ast_func(void *opaque, enum dlm_status status)
+static void ocfs2_unlock_ast(void *opaque, enum dlm_status status)
 {
        struct ocfs2_lock_res *lockres = opaque;
        unsigned long flags;
@@ -2190,24 +2247,20 @@ complete_unlock:
        mlog_exit_void();
 }
-typedef void (ocfs2_pre_drop_cb_t)(struct ocfs2_lock_res *, void *);
-struct drop_lock_cb {
-        ocfs2_pre_drop_cb_t     *drop_func;
-        void                    *drop_data;
-};
 static int ocfs2_drop_lock(struct ocfs2_super *osb,
-                           struct ocfs2_lock_res *lockres,
+                           struct ocfs2_lock_res *lockres)
-                           struct drop_lock_cb *dcb)
 {
        enum dlm_status status;
        unsigned long flags;
+        int lkm_flags = 0;
        /* We didn't get anywhere near actually using this lockres. */
        if (!(lockres->l_flags & OCFS2_LOCK_INITIALIZED))
                goto out;
+        if (lockres->l_ops->flags & LOCK_TYPE_USES_LVB)
+                lkm_flags |= LKM_VALBLK;
        spin_lock_irqsave(&lockres->l_lock, flags);
        mlog_bug_on_msg(!(lockres->l_flags & OCFS2_LOCK_FREEING),
@@ -2230,8 +2283,12 @@ static int ocfs2_drop_lock(struct ocfs2_super *osb,
                spin_lock_irqsave(&lockres->l_lock, flags);
        }
-        if (dcb)
+        if (lockres->l_ops->flags & LOCK_TYPE_USES_LVB) {
-                dcb->drop_func(lockres, dcb->drop_data);
+                if (lockres->l_flags & OCFS2_LOCK_ATTACHED &&
+                    lockres->l_level == LKM_EXMODE &&
+                    !(lockres->l_flags & OCFS2_LOCK_NEEDS_REFRESH))
+                        lockres->l_ops->set_lvb(lockres);
+        }
        if (lockres->l_flags & OCFS2_LOCK_BUSY)
                mlog(ML_ERROR, "destroying busy lock: \"%s\"\n",
@@ -2257,8 +2314,8 @@ static int ocfs2_drop_lock(struct ocfs2_super *osb,
        mlog(0, "lock %s\n", lockres->l_name);
-        status = dlmunlock(osb->dlm, &lockres->l_lksb, LKM_VALBLK,
+        status = dlmunlock(osb->dlm, &lockres->l_lksb, lkm_flags,
-                           lockres->l_ops->unlock_ast, lockres);
+                           ocfs2_unlock_ast, lockres);
        if (status != DLM_NORMAL) {
                ocfs2_log_dlm_error("dlmunlock", status, lockres);
                mlog(ML_ERROR, "lockres flags: %lu\n", lockres->l_flags);
@@ -2305,43 +2362,26 @@ void ocfs2_mark_lockres_freeing(struct ocfs2_lock_res *lockres)
        spin_unlock_irqrestore(&lockres->l_lock, flags);
 }
-static void ocfs2_drop_osb_locks(struct ocfs2_super *osb)
+void ocfs2_simple_drop_lockres(struct ocfs2_super *osb,
+                               struct ocfs2_lock_res *lockres)
 {
-        int status;
+        int ret;
-        mlog_entry_void();
-        ocfs2_mark_lockres_freeing(&osb->osb_super_lockres);
-        status = ocfs2_drop_lock(osb, &osb->osb_super_lockres, NULL);
-        if (status < 0)
-                mlog_errno(status);
-        ocfs2_mark_lockres_freeing(&osb->osb_rename_lockres);
-        status = ocfs2_drop_lock(osb, &osb->osb_rename_lockres, NULL);
-        if (status < 0)
-                mlog_errno(status);
-        mlog_exit(status);
+        ocfs2_mark_lockres_freeing(lockres);
+        ret = ocfs2_drop_lock(osb, lockres);
+        if (ret)
+                mlog_errno(ret);
 }
-static void ocfs2_meta_pre_drop(struct ocfs2_lock_res *lockres, void *data)
+static void ocfs2_drop_osb_locks(struct ocfs2_super *osb)
 {
-        struct inode *inode = data;
+        ocfs2_simple_drop_lockres(osb, &osb->osb_super_lockres);
+        ocfs2_simple_drop_lockres(osb, &osb->osb_rename_lockres);
-        /* the metadata lock requires a bit more work as we have an
-         * LVB to worry about. */
-        if (lockres->l_flags & OCFS2_LOCK_ATTACHED &&
-            lockres->l_level == LKM_EXMODE &&
-            !(lockres->l_flags & OCFS2_LOCK_NEEDS_REFRESH))
-                __ocfs2_stuff_meta_lvb(inode);
 }
 int ocfs2_drop_inode_locks(struct inode *inode)
 {
        int status, err;
-        struct drop_lock_cb meta_dcb = { ocfs2_meta_pre_drop, inode, };
        mlog_entry_void();
@@ -2349,24 +2389,21 @@ int ocfs2_drop_inode_locks(struct inode *inode)
         * ocfs2_clear_inode has done it for us. */
        err = ocfs2_drop_lock(OCFS2_SB(inode->i_sb),
-                              &OCFS2_I(inode)->ip_data_lockres,
+                              &OCFS2_I(inode)->ip_data_lockres);
-                              NULL);
        if (err < 0)
                mlog_errno(err);
        status = err;
        err = ocfs2_drop_lock(OCFS2_SB(inode->i_sb),
-                              &OCFS2_I(inode)->ip_meta_lockres,
+                              &OCFS2_I(inode)->ip_meta_lockres);
-                              &meta_dcb);
        if (err < 0)
                mlog_errno(err);
        if (err < 0 && !status)
                status = err;
        err = ocfs2_drop_lock(OCFS2_SB(inode->i_sb),
-                              &OCFS2_I(inode)->ip_rw_lockres,
+                              &OCFS2_I(inode)->ip_rw_lockres);
-                              NULL);
        if (err < 0)
                mlog_errno(err);
        if (err < 0 && !status)
@@ -2415,9 +2452,10 @@ static int ocfs2_downconvert_lock(struct ocfs2_super *osb,
                         &lockres->l_lksb,
                         dlm_flags,
                         lockres->l_name,
-                         lockres->l_ops->ast,
+                         OCFS2_LOCK_ID_MAX_LEN - 1,
+                         ocfs2_locking_ast,
                         lockres,
-                         lockres->l_ops->bast);
+                         ocfs2_blocking_ast);
        if (status != DLM_NORMAL) {
                ocfs2_log_dlm_error("dlmlock", status, lockres);
                ret = -EINVAL;
@@ -2476,7 +2514,7 @@ static int ocfs2_cancel_convert(struct ocfs2_super *osb,
        status = dlmunlock(osb->dlm,
                           &lockres->l_lksb,
                           LKM_CANCEL,
-                           lockres->l_ops->unlock_ast,
+                           ocfs2_unlock_ast,
                           lockres);
        if (status != DLM_NORMAL) {
                ocfs2_log_dlm_error("dlmunlock", status, lockres);
@@ -2490,115 +2528,15 @@ static int ocfs2_cancel_convert(struct ocfs2_super *osb,
        return ret;
 }
-static inline int ocfs2_can_downconvert_meta_lock(struct inode *inode,
+static int ocfs2_unblock_lock(struct ocfs2_super *osb,
-                                                  struct ocfs2_lock_res *lockres,
+                              struct ocfs2_lock_res *lockres,
-                                                  int new_level)
+                              struct ocfs2_unblock_ctl *ctl)
-{
-        int ret;
-        mlog_entry_void();
-        BUG_ON(new_level != LKM_NLMODE && new_level != LKM_PRMODE);
-        if (lockres->l_flags & OCFS2_LOCK_REFRESHING) {
-                ret = 0;
-                mlog(0, "lockres %s currently being refreshed -- backing "
-                     "off!\n", lockres->l_name);
-        } else if (new_level == LKM_PRMODE)
-                ret = !lockres->l_ex_holders &&
-                        ocfs2_inode_fully_checkpointed(inode);
-        else /* Must be NLMODE we're converting to. */
-                ret = !lockres->l_ro_holders && !lockres->l_ex_holders &&
-                        ocfs2_inode_fully_checkpointed(inode);
-        mlog_exit(ret);
-        return ret;
-}
-static int ocfs2_do_unblock_meta(struct inode *inode,
-                                 int *requeue)
-{
-        int new_level;
-        int set_lvb = 0;
-        int ret = 0;
-        struct ocfs2_lock_res *lockres = &OCFS2_I(inode)->ip_meta_lockres;
-        unsigned long flags;
-        struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
-        mlog_entry_void();
-        spin_lock_irqsave(&lockres->l_lock, flags);
-        BUG_ON(!(lockres->l_flags & OCFS2_LOCK_BLOCKED));
-        mlog(0, "l_level=%d, l_blocking=%d\n", lockres->l_level,
-             lockres->l_blocking);
-        BUG_ON(lockres->l_level != LKM_EXMODE &&
-               lockres->l_level != LKM_PRMODE);
-        if (lockres->l_flags & OCFS2_LOCK_BUSY) {
-                *requeue = 1;
-                ret = ocfs2_prepare_cancel_convert(osb, lockres);
-                spin_unlock_irqrestore(&lockres->l_lock, flags);
-                if (ret) {
-                        ret = ocfs2_cancel_convert(osb, lockres);
-                        if (ret < 0)
-                                mlog_errno(ret);
-                }
-                goto leave;
-        }
-        new_level = ocfs2_highest_compat_lock_level(lockres->l_blocking);
-        mlog(0, "l_level=%d, l_blocking=%d, new_level=%d\n",
-             lockres->l_level, lockres->l_blocking, new_level);
-        if (ocfs2_can_downconvert_meta_lock(inode, lockres, new_level)) {
-                if (lockres->l_level == LKM_EXMODE)
-                        set_lvb = 1;
-                /* If the lock hasn't been refreshed yet (rare), then
-                 * our memory inode values are old and we skip
-                 * stuffing the lvb. There's no need to actually clear
-                 * out the lvb here as it's value is still valid. */
-                if (!(lockres->l_flags & OCFS2_LOCK_NEEDS_REFRESH)) {
-                        if (set_lvb)
-                                __ocfs2_stuff_meta_lvb(inode);
-                } else
-                        mlog(0, "lockres %s: downconverting stale lock!\n",
-                             lockres->l_name);
-                mlog(0, "calling ocfs2_downconvert_lock with l_level=%d, "
-                     "l_blocking=%d, new_level=%d\n",
-                     lockres->l_level, lockres->l_blocking, new_level);
-                ocfs2_prepare_downconvert(lockres, new_level);
-                spin_unlock_irqrestore(&lockres->l_lock, flags);
-                ret = ocfs2_downconvert_lock(osb, lockres, new_level, set_lvb);
-                goto leave;
-        }
-        if (!ocfs2_inode_fully_checkpointed(inode))
-                ocfs2_start_checkpoint(osb);
-        *requeue = 1;
-        spin_unlock_irqrestore(&lockres->l_lock, flags);
-        ret = 0;
-leave:
-        mlog_exit(ret);
-        return ret;
-}
-static int ocfs2_generic_unblock_lock(struct ocfs2_super *osb,
-                                      struct ocfs2_lock_res *lockres,
-                                      int *requeue,
-                                      ocfs2_convert_worker_t *worker)
 {
        unsigned long flags;
        int blocking;
        int new_level;
        int ret = 0;
+        int set_lvb = 0;
        mlog_entry_void();
@@ -2608,7 +2546,7 @@ static int ocfs2_generic_unblock_lock(struct ocfs2_super *osb,
 recheck:
        if (lockres->l_flags & OCFS2_LOCK_BUSY) {
-                *requeue = 1;
+                ctl->requeue = 1;
                ret = ocfs2_prepare_cancel_convert(osb, lockres);
                spin_unlock_irqrestore(&lockres->l_lock, flags);
                if (ret) {
@@ -2622,27 +2560,33 @@ recheck:
        /* if we're blocking an exclusive and we have *any* holders,
         * then requeue. */
        if ((lockres->l_blocking == LKM_EXMODE)
-            && (lockres->l_ex_holders || lockres->l_ro_holders)) {
+            && (lockres->l_ex_holders || lockres->l_ro_holders))
-                spin_unlock_irqrestore(&lockres->l_lock, flags);
+                goto leave_requeue;
-                *requeue = 1;
-                ret = 0;
-                goto leave;
-        }
        /* If it's a PR we're blocking, then only
         * requeue if we've got any EX holders */
        if (lockres->l_blocking == LKM_PRMODE &&
-            lockres->l_ex_holders) {
+            lockres->l_ex_holders)
-                spin_unlock_irqrestore(&lockres->l_lock, flags);
+                goto leave_requeue;
-                *requeue = 1;
-                ret = 0;
+        /*
-                goto leave;
+         * Can we get a lock in this state if the holder counts are
-        }
+         * zero? The meta data unblock code used to check this.
+         */
+        if ((lockres->l_ops->flags & LOCK_TYPE_REQUIRES_REFRESH)
+            && (lockres->l_flags & OCFS2_LOCK_REFRESHING))
+                goto leave_requeue;
+        new_level = ocfs2_highest_compat_lock_level(lockres->l_blocking);
+        if (lockres->l_ops->check_downconvert
+            && !lockres->l_ops->check_downconvert(lockres, new_level))
+                goto leave_requeue;
        /* If we get here, then we know that there are no more
         * incompatible holders (and anyone asking for an incompatible
         * lock is blocked). We can now downconvert the lock */
-        if (!worker)
+        if (!lockres->l_ops->downconvert_worker)
                goto downconvert;
        /* Some lockres types want to do a bit of work before
@@ -2652,7 +2596,10 @@ recheck:
        blocking = lockres->l_blocking;
        spin_unlock_irqrestore(&lockres->l_lock, flags);
-        worker(lockres, blocking);
+        ctl->unblock_action = lockres->l_ops->downconvert_worker(lockres, blocking);
+        if (ctl->unblock_action == UNBLOCK_STOP_POST)
+                goto leave;
        spin_lock_irqsave(&lockres->l_lock, flags);
        if (blocking != lockres->l_blocking) {
@@ -2662,25 +2609,43 @@ recheck:
        }
 downconvert:
-        *requeue = 0;
+        ctl->requeue = 0;
-        new_level = ocfs2_highest_compat_lock_level(lockres->l_blocking);
+        if (lockres->l_ops->flags & LOCK_TYPE_USES_LVB) {
+                if (lockres->l_level == LKM_EXMODE)
+                        set_lvb = 1;
+                /*
+                 * We only set the lvb if the lock has been fully
+                 * refreshed - otherwise we risk setting stale
+                 * data. Otherwise, there's no need to actually clear
+                 * out the lvb here as it's value is still valid.
+                 */
+                if (set_lvb && !(lockres->l_flags & OCFS2_LOCK_NEEDS_REFRESH))
+                        lockres->l_ops->set_lvb(lockres);
+        }
        ocfs2_prepare_downconvert(lockres, new_level);
        spin_unlock_irqrestore(&lockres->l_lock, flags);
-        ret = ocfs2_downconvert_lock(osb, lockres, new_level, 0);
+        ret = ocfs2_downconvert_lock(osb, lockres, new_level, set_lvb);
 leave:
        mlog_exit(ret);
        return ret;
+leave_requeue:
+        spin_unlock_irqrestore(&lockres->l_lock, flags);
+        ctl->requeue = 1;
+        mlog_exit(0);
+        return 0;
 }
-static void ocfs2_data_convert_worker(struct ocfs2_lock_res *lockres,
+static int ocfs2_data_convert_worker(struct ocfs2_lock_res *lockres,
-                                      int blocking)
+                                     int blocking)
 {
        struct inode *inode;
        struct address_space *mapping;
-        mlog_entry_void();
        inode = ocfs2_lock_res_inode(lockres);
        mapping = inode->i_mapping;
@@ -2701,116 +2666,159 @@ static void ocfs2_data_convert_worker(struct ocfs2_lock_res *lockres,
                filemap_fdatawait(mapping);
        }
-        mlog_exit_void();
+        return UNBLOCK_CONTINUE;
 }
-int ocfs2_unblock_data(struct ocfs2_lock_res *lockres,
+static int ocfs2_check_meta_downconvert(struct ocfs2_lock_res *lockres,
-                       int *requeue)
+                                        int new_level)
 {
-        int status;
+        struct inode *inode = ocfs2_lock_res_inode(lockres);
-        struct inode *inode;
+        int checkpointed = ocfs2_inode_fully_checkpointed(inode);
-        struct ocfs2_super *osb;
-        mlog_entry_void();
-        inode = ocfs2_lock_res_inode(lockres);
-        osb = OCFS2_SB(inode->i_sb);
-        mlog(0, "unblock inode %llu\n",
-             (unsigned long long)OCFS2_I(inode)->ip_blkno);
-        status = ocfs2_generic_unblock_lock(osb,
+        BUG_ON(new_level != LKM_NLMODE && new_level != LKM_PRMODE);
-                                            lockres,
+        BUG_ON(lockres->l_level != LKM_EXMODE && !checkpointed);
-                                            requeue,
-                                            ocfs2_data_convert_worker);
-        if (status < 0)
-                mlog_errno(status);
-        mlog(0, "inode %llu, requeue = %d\n",
+        if (checkpointed)
-             (unsigned long long)OCFS2_I(inode)->ip_blkno, *requeue);
+                return 1;
-        mlog_exit(status);
+        ocfs2_start_checkpoint(OCFS2_SB(inode->i_sb));
-        return status;
+        return 0;
 }
-static int ocfs2_unblock_inode_lock(struct ocfs2_lock_res *lockres,
+static void ocfs2_set_meta_lvb(struct ocfs2_lock_res *lockres)
-                                    int *requeue)
 {
-        int status;
+        struct inode *inode = ocfs2_lock_res_inode(lockres);
-        struct inode *inode;
-        mlog_entry_void();
-        mlog(0, "Unblock lockres %s\n", lockres->l_name);
-        inode  = ocfs2_lock_res_inode(lockres);
-        status = ocfs2_generic_unblock_lock(OCFS2_SB(inode->i_sb),
+        __ocfs2_stuff_meta_lvb(inode);
-                                            lockres,
-                                            requeue,
-                                            NULL);
-        if (status < 0)
-                mlog_errno(status);
-        mlog_exit(status);
-        return status;
 }
+/*
-int ocfs2_unblock_meta(struct ocfs2_lock_res *lockres,
+ * Does the final reference drop on our dentry lock. Right now this
-                       int *requeue)
+ * happens in the vote thread, but we could choose to simplify the
+ * dlmglue API and push these off to the ocfs2_wq in the future.
+ */
+static void ocfs2_dentry_post_unlock(struct ocfs2_super *osb,
+                                     struct ocfs2_lock_res *lockres)
 {
-        int status;
+        struct ocfs2_dentry_lock *dl = ocfs2_lock_res_dl(lockres);
-        struct inode *inode;
+        ocfs2_dentry_lock_put(osb, dl);
+}
-        mlog_entry_void();
-        inode = ocfs2_lock_res_inode(lockres);
+/*
+ * d_delete() matching dentries before the lock downconvert.
+ *
+ * At this point, any process waiting to destroy the
+ * dentry_lock due to last ref count is stopped by the
+ * OCFS2_LOCK_QUEUED flag.
+ *
+ * We have two potential problems
+ *
+ * 1) If we do the last reference drop on our dentry_lock (via dput)
+ *    we'll wind up in ocfs2_release_dentry_lock(), waiting on
+ *    the downconvert to finish. Instead we take an elevated
+ *    reference and push the drop until after we've completed our
+ *    unblock processing.
+ *
+ * 2) There might be another process with a final reference,
+ *    waiting on us to finish processing. If this is the case, we
+ *    detect it and exit out - there's no more dentries anyway.
+ */
+static int ocfs2_dentry_convert_worker(struct ocfs2_lock_res *lockres,
+                                       int blocking)
+{
+        struct ocfs2_dentry_lock *dl = ocfs2_lock_res_dl(lockres);
+        struct ocfs2_inode_info *oi = OCFS2_I(dl->dl_inode);
+        struct dentry *dentry;
+        unsigned long flags;
+        int extra_ref = 0;
-        mlog(0, "unblock inode %llu\n",
+        /*
-             (unsigned long long)OCFS2_I(inode)->ip_blkno);
+         * This node is blocking another node from getting a read
+         * lock. This happens when we've renamed within a
+         * directory. We've forced the other nodes to d_delete(), but
+         * we never actually dropped our lock because it's still
+         * valid. The downconvert code will retain a PR for this node,
+         * so there's no further work to do.
+         */
+        if (blocking == LKM_PRMODE)
+                return UNBLOCK_CONTINUE;
-        status = ocfs2_do_unblock_meta(inode, requeue);
+        /*
-        if (status < 0)
+         * Mark this inode as potentially orphaned. The code in
-                mlog_errno(status);
+         * ocfs2_delete_inode() will figure out whether it actually
+         * needs to be freed or not.
+         */
+        spin_lock(&oi->ip_lock);
+        oi->ip_flags |= OCFS2_INODE_MAYBE_ORPHANED;
+        spin_unlock(&oi->ip_lock);
-        mlog(0, "inode %llu, requeue = %d\n",
+        /*
-             (unsigned long long)OCFS2_I(inode)->ip_blkno, *requeue);
+         * Yuck. We need to make sure however that the check of
+         * OCFS2_LOCK_FREEING and the extra reference are atomic with
+         * respect to a reference decrement or the setting of that
+         * flag.
+         */
+        spin_lock_irqsave(&lockres->l_lock, flags);
+        spin_lock(&dentry_attach_lock);
+        if (!(lockres->l_flags & OCFS2_LOCK_FREEING)
+            && dl->dl_count) {
+                dl->dl_count++;
+                extra_ref = 1;
+        }
+        spin_unlock(&dentry_attach_lock);
+        spin_unlock_irqrestore(&lockres->l_lock, flags);
-        mlog_exit(status);
+        mlog(0, "extra_ref = %d\n", extra_ref);
-        return status;
-}
-/* Generic unblock function for any lockres whose private data is an
+        /*
- * ocfs2_super pointer. */
+         * We have a process waiting on us in ocfs2_dentry_iput(),
-static int ocfs2_unblock_osb_lock(struct ocfs2_lock_res *lockres,
+         * which means we can't have any more outstanding
-                                  int *requeue)
+         * aliases. There's no need to do any more work.
-{
+         */
-        int status;
+        if (!extra_ref)
-        struct ocfs2_super *osb;
+                return UNBLOCK_CONTINUE;
+        spin_lock(&dentry_attach_lock);
+        while (1) {
+                dentry = ocfs2_find_local_alias(dl->dl_inode,
+                                                dl->dl_parent_blkno, 1);
+                if (!dentry)
+                        break;
+                spin_unlock(&dentry_attach_lock);
-        mlog_entry_void();
+                mlog(0, "d_delete(%.*s);\n", dentry->d_name.len,
+                     dentry->d_name.name);
-        mlog(0, "Unblock lockres %s\n", lockres->l_name);
+                /*
+                 * The following dcache calls may do an
+                 * iput(). Normally we don't want that from the
+                 * downconverting thread, but in this case it's ok
+                 * because the requesting node already has an
+                 * exclusive lock on the inode, so it can't be queued
+                 * for a downconvert.
+                 */
+                d_delete(dentry);
+                dput(dentry);
-        osb = ocfs2_lock_res_super(lockres);
+                spin_lock(&dentry_attach_lock);
+        }
+        spin_unlock(&dentry_attach_lock);
-        status = ocfs2_generic_unblock_lock(osb,
+        /*
-                                            lockres,
+         * If we are the last holder of this dentry lock, there is no
-                                            requeue,
+         * reason to downconvert so skip straight to the unlock.
-                                            NULL);
+         */
-        if (status < 0)
+        if (dl->dl_count == 1)
-                mlog_errno(status);
+                return UNBLOCK_STOP_POST;
-        mlog_exit(status);
+        return UNBLOCK_CONTINUE_POST;
-        return status;
 }
 void ocfs2_process_blocked_lock(struct ocfs2_super *osb,
                                struct ocfs2_lock_res *lockres)
 {
        int status;
-        int requeue = 0;
+        struct ocfs2_unblock_ctl ctl = {0, 0,};
        unsigned long flags;
        /* Our reference to the lockres in this function can be
@@ -2821,7 +2829,6 @@ void ocfs2_process_blocked_lock(struct ocfs2_super *osb,
        BUG_ON(!lockres);
        BUG_ON(!lockres->l_ops);
-        BUG_ON(!lockres->l_ops->unblock);
        mlog(0, "lockres %s blocked.\n", lockres->l_name);
@@ -2835,21 +2842,25 @@ void ocfs2_process_blocked_lock(struct ocfs2_super *osb,
                goto unqueue;
        spin_unlock_irqrestore(&lockres->l_lock, flags);
-        status = lockres->l_ops->unblock(lockres, &requeue);
+        status = ocfs2_unblock_lock(osb, lockres, &ctl);
        if (status < 0)
                mlog_errno(status);
        spin_lock_irqsave(&lockres->l_lock, flags);
 unqueue:
-        if (lockres->l_flags & OCFS2_LOCK_FREEING || !requeue) {
+        if (lockres->l_flags & OCFS2_LOCK_FREEING || !ctl.requeue) {
                lockres_clear_flags(lockres, OCFS2_LOCK_QUEUED);
        } else
                ocfs2_schedule_blocked_lock(osb, lockres);
        mlog(0, "lockres %s, requeue = %s.\n", lockres->l_name,
-             requeue ? "yes" : "no");
+             ctl.requeue ? "yes" : "no");
        spin_unlock_irqrestore(&lockres->l_lock, flags);
+        if (ctl.unblock_action != UNBLOCK_CONTINUE
+            && lockres->l_ops->post_unlock)
+                lockres->l_ops->post_unlock(osb, lockres);
        mlog_exit_void();
 }
@@ -2892,15 +2903,17 @@ void ocfs2_dump_meta_lvb_info(u64 level,
        mlog(level, "LVB information for %s (called from %s:%u):\n",
             lockres->l_name, function, line);
-        mlog(level, "version: %u, clusters: %u\n",
+        mlog(level, "version: %u, clusters: %u, generation: 0x%x\n",
-             be32_to_cpu(lvb->lvb_version), be32_to_cpu(lvb->lvb_iclusters));
+             lvb->lvb_version, be32_to_cpu(lvb->lvb_iclusters),
+             be32_to_cpu(lvb->lvb_igeneration));
        mlog(level, "size: %llu, uid %u, gid %u, mode 0x%x\n",
             (unsigned long long)be64_to_cpu(lvb->lvb_isize),
             be32_to_cpu(lvb->lvb_iuid), be32_to_cpu(lvb->lvb_igid),
             be16_to_cpu(lvb->lvb_imode));
        mlog(level, "nlink %u, atime_packed 0x%llx, ctime_packed 0x%llx, "
-             "mtime_packed 0x%llx\n", be16_to_cpu(lvb->lvb_inlink),
+             "mtime_packed 0x%llx iattr 0x%x\n", be16_to_cpu(lvb->lvb_inlink),
             (long long)be64_to_cpu(lvb->lvb_iatime_packed),
             (long long)be64_to_cpu(lvb->lvb_ictime_packed),
-             (long long)be64_to_cpu(lvb->lvb_imtime_packed));
+             (long long)be64_to_cpu(lvb->lvb_imtime_packed),
+             be32_to_cpu(lvb->lvb_iattr));
 }
diff --git a/fs/ocfs2/dlmglue.h b/fs/ocfs2/dlmglue.h
index 8f2d1db2d9ea..4a2769387229 100644
--- a/fs/ocfs2/dlmglue.h
+++ b/fs/ocfs2/dlmglue.h
@@ -27,10 +27,14 @@
 #ifndef DLMGLUE_H
 #define DLMGLUE_H
-#define OCFS2_LVB_VERSION 2
+#include "dcache.h"
+#define OCFS2_LVB_VERSION 4
 struct ocfs2_meta_lvb {
-        __be32       lvb_version;
+        __u8         lvb_version;
+        __u8         lvb_reserved0;
+        __be16       lvb_reserved1;
        __be32       lvb_iclusters;
        __be32       lvb_iuid;
        __be32       lvb_igid;
@@ -40,7 +44,9 @@ struct ocfs2_meta_lvb {
        __be64       lvb_isize;
        __be16       lvb_imode;
        __be16       lvb_inlink;
-        __be32       lvb_reserved[3];
+        __be32       lvb_iattr;
+        __be32       lvb_igeneration;
+        __be32       lvb_reserved2;
 };
 /* ocfs2_meta_lock_full() and ocfs2_data_lock_full() 'arg_flags' flags */
@@ -56,9 +62,14 @@ void ocfs2_dlm_shutdown(struct ocfs2_super *osb);
 void ocfs2_lock_res_init_once(struct ocfs2_lock_res *res);
 void ocfs2_inode_lock_res_init(struct ocfs2_lock_res *res,
                               enum ocfs2_lock_type type,
+                               unsigned int generation,
                               struct inode *inode);
+void ocfs2_dentry_lock_res_init(struct ocfs2_dentry_lock *dl,
+                                u64 parent, struct inode *inode);
 void ocfs2_lock_res_free(struct ocfs2_lock_res *res);
 int ocfs2_create_new_inode_locks(struct inode *inode);
+int ocfs2_create_new_lock(struct ocfs2_super *osb,
+                          struct ocfs2_lock_res *lockres, int ex, int local);
 int ocfs2_drop_inode_locks(struct inode *inode);
 int ocfs2_data_lock_full(struct inode *inode,
                         int write,
@@ -92,7 +103,12 @@ void ocfs2_super_unlock(struct ocfs2_super *osb,
                        int ex);
 int ocfs2_rename_lock(struct ocfs2_super *osb);
 void ocfs2_rename_unlock(struct ocfs2_super *osb);
+int ocfs2_dentry_lock(struct dentry *dentry, int ex);
+void ocfs2_dentry_unlock(struct dentry *dentry, int ex);
 void ocfs2_mark_lockres_freeing(struct ocfs2_lock_res *lockres);
+void ocfs2_simple_drop_lockres(struct ocfs2_super *osb,
+                               struct ocfs2_lock_res *lockres);
 /* for the vote thread */
 void ocfs2_process_blocked_lock(struct ocfs2_super *osb,
diff --git a/fs/ocfs2/export.c b/fs/ocfs2/export.c
index ec55ab3c1214..fb91089a60a7 100644
--- a/fs/ocfs2/export.c
+++ b/fs/ocfs2/export.c
@@ -33,6 +33,7 @@
 #include "dir.h"
 #include "dlmglue.h"
+#include "dcache.h"
 #include "export.h"
 #include "inode.h"
@@ -57,7 +58,7 @@ static struct dentry *ocfs2_get_dentry(struct super_block *sb, void *vobjp)
                return ERR_PTR(-ESTALE);
        }
-        inode = ocfs2_iget(OCFS2_SB(sb), handle->ih_blkno);
+        inode = ocfs2_iget(OCFS2_SB(sb), handle->ih_blkno, 0);
        if (IS_ERR(inode)) {
                mlog_errno(PTR_ERR(inode));
@@ -77,6 +78,7 @@ static struct dentry *ocfs2_get_dentry(struct super_block *sb, void *vobjp)
                mlog_errno(-ENOMEM);
                return ERR_PTR(-ENOMEM);
        }
+        result->d_op = &ocfs2_dentry_ops;
        mlog_exit_ptr(result);
        return result;
@@ -113,7 +115,7 @@ static struct dentry *ocfs2_get_parent(struct dentry *child)
                goto bail_unlock;
        }
-        inode = ocfs2_iget(OCFS2_SB(dir->i_sb), blkno);
+        inode = ocfs2_iget(OCFS2_SB(dir->i_sb), blkno, 0);
        if (IS_ERR(inode)) {
                mlog(ML_ERROR, "Unable to create inode %llu\n",
                     (unsigned long long)blkno);
@@ -127,6 +129,8 @@ static struct dentry *ocfs2_get_parent(struct dentry *child)
                parent = ERR_PTR(-ENOMEM);
        }
+        parent->d_op = &ocfs2_dentry_ops;
 bail_unlock:
        ocfs2_meta_unlock(dir, 0);
diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c
index a9559c874530..2bbfa17090cf 100644
--- a/fs/ocfs2/file.c
+++ b/fs/ocfs2/file.c
@@ -44,6 +44,7 @@
 #include "file.h"
 #include "sysfile.h"
 #include "inode.h"
+#include "ioctl.h"
 #include "journal.h"
 #include "mmap.h"
 #include "suballoc.h"
@@ -1227,10 +1228,12 @@ const struct file_operations ocfs2_fops = {
        .open           = ocfs2_file_open,
        .aio_read       = ocfs2_file_aio_read,
        .aio_write      = ocfs2_file_aio_write,
+        .ioctl          = ocfs2_ioctl,
 };
 const struct file_operations ocfs2_dops = {
        .read           = generic_read_dir,
        .readdir        = ocfs2_readdir,
        .fsync          = ocfs2_sync_file,
+        .ioctl          = ocfs2_ioctl,
 };
diff --git a/fs/ocfs2/inode.c b/fs/ocfs2/inode.c
index 327a5b7b86ed..69d3db569166 100644
--- a/fs/ocfs2/inode.c
+++ b/fs/ocfs2/inode.c
@@ -54,8 +54,6 @@
 #include "buffer_head_io.h"
-#define OCFS2_FI_FLAG_NOWAIT    0x1
-#define OCFS2_FI_FLAG_DELETE    0x2
 struct ocfs2_find_inode_args
 {
        u64             fi_blkno;
@@ -71,6 +69,26 @@ static int ocfs2_truncate_for_delete(struct ocfs2_super *osb,
                                    struct inode *inode,
                                    struct buffer_head *fe_bh);
+void ocfs2_set_inode_flags(struct inode *inode)
+{
+        unsigned int flags = OCFS2_I(inode)->ip_attr;
+        inode->i_flags &= ~(S_IMMUTABLE |
+                S_SYNC | S_APPEND | S_NOATIME | S_DIRSYNC);
+        if (flags & OCFS2_IMMUTABLE_FL)
+                inode->i_flags |= S_IMMUTABLE;
+        if (flags & OCFS2_SYNC_FL)
+                inode->i_flags |= S_SYNC;
+        if (flags & OCFS2_APPEND_FL)
+                inode->i_flags |= S_APPEND;
+        if (flags & OCFS2_NOATIME_FL)
+                inode->i_flags |= S_NOATIME;
+        if (flags & OCFS2_DIRSYNC_FL)
+                inode->i_flags |= S_DIRSYNC;
+}
 struct inode *ocfs2_ilookup_for_vote(struct ocfs2_super *osb,
                                     u64 blkno,
                                     int delete_vote)
@@ -89,7 +107,7 @@ struct inode *ocfs2_ilookup_for_vote(struct ocfs2_super *osb,
        return ilookup5(osb->sb, args.fi_ino, ocfs2_find_actor, &args);
 }
-struct inode *ocfs2_iget(struct ocfs2_super *osb, u64 blkno)
+struct inode *ocfs2_iget(struct ocfs2_super *osb, u64 blkno, int flags)
 {
        struct inode *inode = NULL;
        struct super_block *sb = osb->sb;
@@ -107,7 +125,7 @@ struct inode *ocfs2_iget(struct ocfs2_super *osb, u64 blkno)
        }
        args.fi_blkno = blkno;
-        args.fi_flags = 0;
+        args.fi_flags = flags;
        args.fi_ino = ino_from_blkno(sb, blkno);
        inode = iget5_locked(sb, args.fi_ino, ocfs2_find_actor,
@@ -260,7 +278,6 @@ int ocfs2_populate_inode(struct inode *inode, struct ocfs2_dinode *fe,
                inode->i_blocks =
                        ocfs2_align_bytes_to_sectors(le64_to_cpu(fe->i_size));
        inode->i_mapping->a_ops = &ocfs2_aops;
-        inode->i_flags |= S_NOATIME;
        inode->i_atime.tv_sec = le64_to_cpu(fe->i_atime);
        inode->i_atime.tv_nsec = le32_to_cpu(fe->i_atime_nsec);
        inode->i_mtime.tv_sec = le64_to_cpu(fe->i_mtime);
@@ -276,16 +293,13 @@ int ocfs2_populate_inode(struct inode *inode, struct ocfs2_dinode *fe,
        OCFS2_I(inode)->ip_clusters = le32_to_cpu(fe->i_clusters);
        OCFS2_I(inode)->ip_orphaned_slot = OCFS2_INVALID_SLOT;
+        OCFS2_I(inode)->ip_attr = le32_to_cpu(fe->i_attr);
-        if (create_ino)
-                inode->i_ino = ino_from_blkno(inode->i_sb,
-                               le64_to_cpu(fe->i_blkno));
-        mlog(0, "blkno = %llu, ino = %lu, create_ino = %s\n",
-             (unsigned long long)fe->i_blkno, inode->i_ino, create_ino ? "true" : "false");
        inode->i_nlink = le16_to_cpu(fe->i_links_count);
+        if (fe->i_flags & cpu_to_le32(OCFS2_SYSTEM_FL))
+                OCFS2_I(inode)->ip_flags |= OCFS2_INODE_SYSTEM_FILE;
        if (fe->i_flags & cpu_to_le32(OCFS2_LOCAL_ALLOC_FL)) {
                OCFS2_I(inode)->ip_flags |= OCFS2_INODE_BITMAP;
                mlog(0, "local alloc inode: i_ino=%lu\n", inode->i_ino);
@@ -323,12 +337,31 @@ int ocfs2_populate_inode(struct inode *inode, struct ocfs2_dinode *fe,
                    break;
        }
+        if (create_ino) {
+                inode->i_ino = ino_from_blkno(inode->i_sb,
+                               le64_to_cpu(fe->i_blkno));
+                /*
+                 * If we ever want to create system files from kernel,
+                 * the generation argument to
+                 * ocfs2_inode_lock_res_init() will have to change.
+                 */
+                BUG_ON(fe->i_flags & cpu_to_le32(OCFS2_SYSTEM_FL));
+                ocfs2_inode_lock_res_init(&OCFS2_I(inode)->ip_meta_lockres,
+                                          OCFS2_LOCK_TYPE_META, 0, inode);
+        }
        ocfs2_inode_lock_res_init(&OCFS2_I(inode)->ip_rw_lockres,
-                                  OCFS2_LOCK_TYPE_RW, inode);
+                                  OCFS2_LOCK_TYPE_RW, inode->i_generation,
-        ocfs2_inode_lock_res_init(&OCFS2_I(inode)->ip_meta_lockres,
+                                  inode);
-                                  OCFS2_LOCK_TYPE_META, inode);
        ocfs2_inode_lock_res_init(&OCFS2_I(inode)->ip_data_lockres,
-                                  OCFS2_LOCK_TYPE_DATA, inode);
+                                  OCFS2_LOCK_TYPE_DATA, inode->i_generation,
+                                  inode);
+        ocfs2_set_inode_flags(inode);
+        inode->i_flags |= S_NOATIME;
        status = 0;
 bail:
@@ -343,15 +376,15 @@ static int ocfs2_read_locked_inode(struct inode *inode,
        struct ocfs2_super *osb;
        struct ocfs2_dinode *fe;
        struct buffer_head *bh = NULL;
-        int status;
+        int status, can_lock;
-        int sysfile = 0;
+        u32 generation = 0;
        mlog_entry("(0x%p, 0x%p)\n", inode, args);
        status = -EINVAL;
        if (inode == NULL || inode->i_sb == NULL) {
                mlog(ML_ERROR, "bad inode\n");
-                goto bail;
+                return status;
        }
        sb = inode->i_sb;
        osb = OCFS2_SB(sb);
@@ -359,50 +392,110 @@ static int ocfs2_read_locked_inode(struct inode *inode,
        if (!args) {
                mlog(ML_ERROR, "bad inode args\n");
                make_bad_inode(inode);
-                goto bail;
+                return status;
+        }
+        /*
+         * To improve performance of cold-cache inode stats, we take
+         * the cluster lock here if possible.
+         *
+         * Generally, OCFS2 never trusts the contents of an inode
+         * unless it's holding a cluster lock, so taking it here isn't
+         * a correctness issue as much as it is a performance
+         * improvement.
+         *
+         * There are three times when taking the lock is not a good idea:
+         *
+         * 1) During startup, before we have initialized the DLM.
+         *
+         * 2) If we are reading certain system files which never get
+         *    cluster locks (local alloc, truncate log).
+         *
+         * 3) If the process doing the iget() is responsible for
+         *    orphan dir recovery. We're holding the orphan dir lock and
+         *    can get into a deadlock with another process on another
+         *    node in ->delete_inode().
+         *
+         * #1 and #2 can be simply solved by never taking the lock
+         * here for system files (which are the only type we read
+         * during mount). It's a heavier approach, but our main
+         * concern is user-accesible files anyway.
+         *
+         * #3 works itself out because we'll eventually take the
+         * cluster lock before trusting anything anyway.
+         */
+        can_lock = !(args->fi_flags & OCFS2_FI_FLAG_SYSFILE)
+                && !(args->fi_flags & OCFS2_FI_FLAG_NOLOCK);
+        /*
+         * To maintain backwards compatibility with older versions of
+         * ocfs2-tools, we still store the generation value for system
+         * files. The only ones that actually matter to userspace are
+         * the journals, but it's easier and inexpensive to just flag
+         * all system files similarly.
+         */
+        if (args->fi_flags & OCFS2_FI_FLAG_SYSFILE)
+                generation = osb->fs_generation;
+        ocfs2_inode_lock_res_init(&OCFS2_I(inode)->ip_meta_lockres,
+                                  OCFS2_LOCK_TYPE_META,
+                                  generation, inode);
+        if (can_lock) {
+                status = ocfs2_meta_lock(inode, NULL, NULL, 0);
+                if (status) {
+                        make_bad_inode(inode);
+                        mlog_errno(status);
+                        return status;
+                }
        }
-        /* Read the FE off disk. This is safe because the kernel only
+        status = ocfs2_read_block(osb, args->fi_blkno, &bh, 0,
-         * does one read_inode2 for a new inode, and if it doesn't
+                                  can_lock ? inode : NULL);
-         * exist yet then nobody can be working on it! */
-        status = ocfs2_read_block(osb, args->fi_blkno, &bh, 0, NULL);
        if (status < 0) {
                mlog_errno(status);
-                make_bad_inode(inode);
                goto bail;
        }
+        status = -EINVAL;
        fe = (struct ocfs2_dinode *) bh->b_data;
        if (!OCFS2_IS_VALID_DINODE(fe)) {
                mlog(ML_ERROR, "Invalid dinode #%llu: signature = %.*s\n",
                     (unsigned long long)fe->i_blkno, 7, fe->i_signature);
-                make_bad_inode(inode);
                goto bail;
        }
-        if (fe->i_flags & cpu_to_le32(OCFS2_SYSTEM_FL))
+        /*
-                sysfile = 1;
+         * This is a code bug. Right now the caller needs to
+         * understand whether it is asking for a system file inode or
+         * not so the proper lock names can be built.
+         */
+        mlog_bug_on_msg(!!(fe->i_flags & cpu_to_le32(OCFS2_SYSTEM_FL)) !=
+                        !!(args->fi_flags & OCFS2_FI_FLAG_SYSFILE),
+                        "Inode %llu: system file state is ambigous\n",
+                        (unsigned long long)args->fi_blkno);
        if (S_ISCHR(le16_to_cpu(fe->i_mode)) ||
            S_ISBLK(le16_to_cpu(fe->i_mode)))
                inode->i_rdev = huge_decode_dev(le64_to_cpu(fe->id1.dev1.i_rdev));
-        status = -EINVAL;
        if (ocfs2_populate_inode(inode, fe, 0) < 0) {
                mlog(ML_ERROR, "populate failed! i_blkno=%llu, i_ino=%lu\n",
                     (unsigned long long)fe->i_blkno, inode->i_ino);
-                make_bad_inode(inode);
                goto bail;
        }
        BUG_ON(args->fi_blkno != le64_to_cpu(fe->i_blkno));
-        if (sysfile)
-               OCFS2_I(inode)->ip_flags |= OCFS2_INODE_SYSTEM_FILE;
        status = 0;
 bail:
+        if (can_lock)
+                ocfs2_meta_unlock(inode, 0);
+        if (status < 0)
+                make_bad_inode(inode);
        if (args && bh)
                brelse(bh);
@@ -875,9 +968,15 @@ void ocfs2_delete_inode(struct inode *inode)
                goto bail_unlock_inode;
        }
-        /* Mark the inode as successfully deleted. This is important
+        /*
-         * for ocfs2_clear_inode as it will check this flag and skip
+         * Mark the inode as successfully deleted.
-         * any checkpointing work */
+         *
+         * This is important for ocfs2_clear_inode() as it will check
+         * this flag and skip any checkpointing work
+         *
+         * ocfs2_stuff_meta_lvb() also uses this flag to invalidate
+         * the LVB for other nodes.
+         */
        OCFS2_I(inode)->ip_flags |= OCFS2_INODE_DELETED;
 bail_unlock_inode:
@@ -1002,12 +1101,10 @@ void ocfs2_drop_inode(struct inode *inode)
        /* Testing ip_orphaned_slot here wouldn't work because we may
         * not have gotten a delete_inode vote from any other nodes
         * yet. */
-        if (oi->ip_flags & OCFS2_INODE_MAYBE_ORPHANED) {
+        if (oi->ip_flags & OCFS2_INODE_MAYBE_ORPHANED)
-                mlog(0, "Inode was orphaned on another node, clearing nlink.\n");
+                generic_delete_inode(inode);
-                inode->i_nlink = 0;
+        else
-        }
+                generic_drop_inode(inode);
-        generic_drop_inode(inode);
        mlog_exit_void();
 }
@@ -1027,12 +1124,8 @@ struct buffer_head *ocfs2_bread(struct inode *inode,
        u64 p_blkno;
        int readflags = OCFS2_BH_CACHED;
-#if 0
-        /* only turn this on if we know we can deal with read_block
-         * returning nothing */
        if (reada)
                readflags |= OCFS2_BH_READAHEAD;
-#endif
        if (((u64)block << inode->i_sb->s_blocksize_bits) >=
            i_size_read(inode)) {
@@ -1131,6 +1224,7 @@ int ocfs2_mark_inode_dirty(struct ocfs2_journal_handle *handle,
        spin_lock(&OCFS2_I(inode)->ip_lock);
        fe->i_clusters = cpu_to_le32(OCFS2_I(inode)->ip_clusters);
+        fe->i_attr = cpu_to_le32(OCFS2_I(inode)->ip_attr);
        spin_unlock(&OCFS2_I(inode)->ip_lock);
        fe->i_size = cpu_to_le64(i_size_read(inode));
@@ -1169,6 +1263,8 @@ void ocfs2_refresh_inode(struct inode *inode,
        spin_lock(&OCFS2_I(inode)->ip_lock);
        OCFS2_I(inode)->ip_clusters = le32_to_cpu(fe->i_clusters);
+        OCFS2_I(inode)->ip_attr = le32_to_cpu(fe->i_attr);
+        ocfs2_set_inode_flags(inode);
        i_size_write(inode, le64_to_cpu(fe->i_size));
        inode->i_nlink = le16_to_cpu(fe->i_links_count);
        inode->i_uid = le32_to_cpu(fe->i_uid);
diff --git a/fs/ocfs2/inode.h b/fs/ocfs2/inode.h
index 35140f6cf840..9957810fdf85 100644
--- a/fs/ocfs2/inode.h
+++ b/fs/ocfs2/inode.h
@@ -56,6 +56,7 @@ struct ocfs2_inode_info
        struct ocfs2_journal_handle     *ip_handle;
        u32                             ip_flags; /* see below */
+        u32                             ip_attr; /* inode attributes */
        /* protected by recovery_lock. */
        struct inode                    *ip_next_orphan;
@@ -121,7 +122,13 @@ struct buffer_head *ocfs2_bread(struct inode *inode, int block,
 void ocfs2_clear_inode(struct inode *inode);
 void ocfs2_delete_inode(struct inode *inode);
 void ocfs2_drop_inode(struct inode *inode);
-struct inode *ocfs2_iget(struct ocfs2_super *osb, u64 feoff);
+/* Flags for ocfs2_iget() */
+#define OCFS2_FI_FLAG_NOWAIT    0x1
+#define OCFS2_FI_FLAG_DELETE    0x2
+#define OCFS2_FI_FLAG_SYSFILE   0x4
+#define OCFS2_FI_FLAG_NOLOCK    0x8
+struct inode *ocfs2_iget(struct ocfs2_super *osb, u64 feoff, int flags);
 struct inode *ocfs2_ilookup_for_vote(struct ocfs2_super *osb,
                                     u64 blkno,
                                     int delete_vote);
@@ -142,4 +149,6 @@ int ocfs2_mark_inode_dirty(struct ocfs2_journal_handle *handle,
 int ocfs2_aio_read(struct file *file, struct kiocb *req, struct iocb *iocb);
 int ocfs2_aio_write(struct file *file, struct kiocb *req, struct iocb *iocb);
+void ocfs2_set_inode_flags(struct inode *inode);
 #endif /* OCFS2_INODE_H */
diff --git a/fs/ocfs2/ioctl.c b/fs/ocfs2/ioctl.c
new file mode 100644
index 000000000000..3663cef80689
--- /dev/null
+++ b/fs/ocfs2/ioctl.c
@@ -0,0 +1,136 @@
+/*
+ * linux/fs/ocfs2/ioctl.c
+ *
+ * Copyright (C) 2006 Herbert Poetzl
+ * adapted from Remy Card's ext2/ioctl.c
+ */
+#include <linux/fs.h>
+#include <linux/mount.h>
+#define MLOG_MASK_PREFIX ML_INODE
+#include <cluster/masklog.h>
+#include "ocfs2.h"
+#include "alloc.h"
+#include "dlmglue.h"
+#include "inode.h"
+#include "journal.h"
+#include "ocfs2_fs.h"
+#include "ioctl.h"
+#include <linux/ext2_fs.h>
+static int ocfs2_get_inode_attr(struct inode *inode, unsigned *flags)
+{
+        int status;
+        status = ocfs2_meta_lock(inode, NULL, NULL, 0);
+        if (status < 0) {
+                mlog_errno(status);
+                return status;
+        }
+        *flags = OCFS2_I(inode)->ip_attr;
+        ocfs2_meta_unlock(inode, 0);
+        mlog_exit(status);
+        return status;
+}
+static int ocfs2_set_inode_attr(struct inode *inode, unsigned flags,
+                                unsigned mask)
+{
+        struct ocfs2_inode_info *ocfs2_inode = OCFS2_I(inode);
+        struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
+        struct ocfs2_journal_handle *handle = NULL;
+        struct buffer_head *bh = NULL;
+        unsigned oldflags;
+        int status;
+        mutex_lock(&inode->i_mutex);
+        status = ocfs2_meta_lock(inode, NULL, &bh, 1);
+        if (status < 0) {
+                mlog_errno(status);
+                goto bail;
+        }
+        status = -EROFS;
+        if (IS_RDONLY(inode))
+                goto bail_unlock;
+        status = -EACCES;
+        if ((current->fsuid != inode->i_uid) && !capable(CAP_FOWNER))
+                goto bail_unlock;
+        if (!S_ISDIR(inode->i_mode))
+                flags &= ~OCFS2_DIRSYNC_FL;
+        handle = ocfs2_start_trans(osb, NULL, OCFS2_INODE_UPDATE_CREDITS);
+        if (IS_ERR(handle)) {
+                status = PTR_ERR(handle);
+                mlog_errno(status);
+                goto bail_unlock;
+        }
+        oldflags = ocfs2_inode->ip_attr;
+        flags = flags & mask;
+        flags |= oldflags & ~mask;
+        /*
+         * The IMMUTABLE and APPEND_ONLY flags can only be changed by
+         * the relevant capability.
+         */
+        status = -EPERM;
+        if ((oldflags & OCFS2_IMMUTABLE_FL) || ((flags ^ oldflags) &
+                (OCFS2_APPEND_FL | OCFS2_IMMUTABLE_FL))) {
+                if (!capable(CAP_LINUX_IMMUTABLE))
+                        goto bail_unlock;
+        }
+        ocfs2_inode->ip_attr = flags;
+        ocfs2_set_inode_flags(inode);
+        status = ocfs2_mark_inode_dirty(handle, inode, bh);
+        if (status < 0)
+                mlog_errno(status);
+        ocfs2_commit_trans(handle);
+bail_unlock:
+        ocfs2_meta_unlock(inode, 1);
+bail:
+        mutex_unlock(&inode->i_mutex);
+        if (bh)
+                brelse(bh);
+        mlog_exit(status);
+        return status;
+}
+int ocfs2_ioctl(struct inode * inode, struct file * filp,
+        unsigned int cmd, unsigned long arg)
+{
+        unsigned int flags;
+        int status;
+        switch (cmd) {
+        case OCFS2_IOC_GETFLAGS:
+                status = ocfs2_get_inode_attr(inode, &flags);
+                if (status < 0)
+                        return status;
+                flags &= OCFS2_FL_VISIBLE;
+                return put_user(flags, (int __user *) arg);
+        case OCFS2_IOC_SETFLAGS:
+                if (get_user(flags, (int __user *) arg))
+                        return -EFAULT;
+                return ocfs2_set_inode_attr(inode, flags,
+                        OCFS2_FL_MODIFIABLE);
+        default:
+                return -ENOTTY;
+        }
+}
diff --git a/fs/ocfs2/ioctl.h b/fs/ocfs2/ioctl.h
new file mode 100644
index 000000000000..4a7c82931dba
--- /dev/null
+++ b/fs/ocfs2/ioctl.h
@@ -0,0 +1,16 @@
+/*
+ * ioctl.h
+ *
+ * Function prototypes
+ *
+ * Copyright (C) 2006 Herbert Poetzl
+ *
+ */
+#ifndef OCFS2_IOCTL_H
+#define OCFS2_IOCTL_H
+int ocfs2_ioctl(struct inode * inode, struct file * filp,
+        unsigned int cmd, unsigned long arg);
+#endif /* OCFS2_IOCTL_H */
diff --git a/fs/ocfs2/journal.c b/fs/ocfs2/journal.c
index f92bf1dd379a..fd9734def551 100644
--- a/fs/ocfs2/journal.c
+++ b/fs/ocfs2/journal.c
@@ -1493,7 +1493,8 @@ static int ocfs2_queue_orphans(struct ocfs2_super *osb,
                        if (de->name_len == 2 && !strncmp("..", de->name, 2))
                                continue;
-                        iter = ocfs2_iget(osb, le64_to_cpu(de->inode));
+                        iter = ocfs2_iget(osb, le64_to_cpu(de->inode),
+                                          OCFS2_FI_FLAG_NOLOCK);
                        if (IS_ERR(iter))
                                continue;
diff --git a/fs/ocfs2/namei.c b/fs/ocfs2/namei.c
index 0673862c8bdd..849c3b4bb94a 100644
--- a/fs/ocfs2/namei.c
+++ b/fs/ocfs2/namei.c
@@ -56,6 +56,7 @@
 #include "journal.h"
 #include "namei.h"
 #include "suballoc.h"
+#include "super.h"
 #include "symlink.h"
 #include "sysfile.h"
 #include "uptodate.h"
@@ -178,7 +179,7 @@ static struct dentry *ocfs2_lookup(struct inode *dir, struct dentry *dentry,
        if (status < 0)
                goto bail_add;
-        inode = ocfs2_iget(OCFS2_SB(dir->i_sb), blkno);
+        inode = ocfs2_iget(OCFS2_SB(dir->i_sb), blkno, 0);
        if (IS_ERR(inode)) {
                mlog(ML_ERROR, "Unable to create inode %llu\n",
                     (unsigned long long)blkno);
@@ -198,10 +199,32 @@ static struct dentry *ocfs2_lookup(struct inode *dir, struct dentry *dentry,
        spin_unlock(&oi->ip_lock);
 bail_add:
        dentry->d_op = &ocfs2_dentry_ops;
        ret = d_splice_alias(inode, dentry);
+        if (inode) {
+                /*
+                 * If d_splice_alias() finds a DCACHE_DISCONNECTED
+                 * dentry, it will d_move() it on top of ourse. The
+                 * return value will indicate this however, so in
+                 * those cases, we switch them around for the locking
+                 * code.
+                 *
+                 * NOTE: This dentry already has ->d_op set from
+                 * ocfs2_get_parent() and ocfs2_get_dentry()
+                 */
+                if (ret)
+                        dentry = ret;
+                status = ocfs2_dentry_attach_lock(dentry, inode,
+                                                  OCFS2_I(dir)->ip_blkno);
+                if (status) {
+                        mlog_errno(status);
+                        ret = ERR_PTR(status);
+                        goto bail_unlock;
+                }
+        }
 bail_unlock:
        /* Don't drop the cluster lock until *after* the d_add --
         * unlink on another node will message us to remove that
@@ -310,13 +333,6 @@ static int ocfs2_mknod(struct inode *dir,
        /* get our super block */
        osb = OCFS2_SB(dir->i_sb);
-        if (S_ISDIR(mode) && (dir->i_nlink >= OCFS2_LINK_MAX)) {
-                mlog(ML_ERROR, "inode %llu has i_nlink of %u\n",
-                     (unsigned long long)OCFS2_I(dir)->ip_blkno, dir->i_nlink);
-                status = -EMLINK;
-                goto leave;
-        }
        handle = ocfs2_alloc_handle(osb);
        if (handle == NULL) {
                status = -ENOMEM;
@@ -331,6 +347,11 @@ static int ocfs2_mknod(struct inode *dir,
                goto leave;
        }
+        if (S_ISDIR(mode) && (dir->i_nlink >= OCFS2_LINK_MAX)) {
+                status = -EMLINK;
+                goto leave;
+        }
        dirfe = (struct ocfs2_dinode *) parent_fe_bh->b_data;
        if (!dirfe->i_links_count) {
                /* can't make a file in a deleted directory. */
@@ -419,6 +440,13 @@ static int ocfs2_mknod(struct inode *dir,
                goto leave;
        }
+        status = ocfs2_dentry_attach_lock(dentry, inode,
+                                          OCFS2_I(dir)->ip_blkno);
+        if (status) {
+                mlog_errno(status);
+                goto leave;
+        }
        insert_inode_hash(inode);
        dentry->d_op = &ocfs2_dentry_ops;
        d_instantiate(dentry, inode);
@@ -643,11 +671,6 @@ static int ocfs2_link(struct dentry *old_dentry,
                goto bail;
        }
-        if (inode->i_nlink >= OCFS2_LINK_MAX) {
-                err = -EMLINK;
-                goto bail;
-        }
        handle = ocfs2_alloc_handle(osb);
        if (handle == NULL) {
                err = -ENOMEM;
@@ -661,6 +684,11 @@ static int ocfs2_link(struct dentry *old_dentry,
                goto bail;
        }
+        if (!dir->i_nlink) {
+                err = -ENOENT;
+                goto bail;
+        }
        err = ocfs2_check_dir_for_entry(dir, dentry->d_name.name,
                                        dentry->d_name.len);
        if (err)
@@ -726,6 +754,12 @@ static int ocfs2_link(struct dentry *old_dentry,
                goto bail;
        }
+        err = ocfs2_dentry_attach_lock(dentry, inode, OCFS2_I(dir)->ip_blkno);
+        if (err) {
+                mlog_errno(err);
+                goto bail;
+        }
        atomic_inc(&inode->i_count);
        dentry->d_op = &ocfs2_dentry_ops;
        d_instantiate(dentry, inode);
@@ -744,6 +778,23 @@ bail:
        return err;
 }
+/*
+ * Takes and drops an exclusive lock on the given dentry. This will
+ * force other nodes to drop it.
+ */
+static int ocfs2_remote_dentry_delete(struct dentry *dentry)
+{
+        int ret;
+        ret = ocfs2_dentry_lock(dentry, 1);
+        if (ret)
+                mlog_errno(ret);
+        else
+                ocfs2_dentry_unlock(dentry, 1);
+        return ret;
+}
 static int ocfs2_unlink(struct inode *dir,
                        struct dentry *dentry)
 {
@@ -833,8 +884,7 @@ static int ocfs2_unlink(struct inode *dir,
        else
                inode->i_nlink--;
-        status = ocfs2_request_unlink_vote(inode, dentry,
+        status = ocfs2_remote_dentry_delete(dentry);
-                                           (unsigned int) inode->i_nlink);
        if (status < 0) {
                /* This vote should succeed under all normal
                 * circumstances. */
@@ -1020,7 +1070,6 @@ static int ocfs2_rename(struct inode *old_dir,
        struct buffer_head *old_inode_de_bh = NULL; // if old_dentry is a dir,
                                                    // this is the 1st dirent bh
        nlink_t old_dir_nlink = old_dir->i_nlink, new_dir_nlink = new_dir->i_nlink;
-        unsigned int links_count;
        /* At some point it might be nice to break this function up a
         * bit. */
@@ -1094,23 +1143,26 @@ static int ocfs2_rename(struct inode *old_dir,
                }
        }
-        if (S_ISDIR(old_inode->i_mode)) {
+        /*
-                /* Directories actually require metadata updates to
+         * Though we don't require an inode meta data update if
-                 * the directory info so we can't get away with not
+         * old_inode is not a directory, we lock anyway here to ensure
-                 * doing node locking on it. */
+         * the vote thread on other nodes won't have to concurrently
-                status = ocfs2_meta_lock(old_inode, handle, NULL, 1);
+         * downconvert the inode and the dentry locks.
-                if (status < 0) {
+         */
-                        if (status != -ENOENT)
+        status = ocfs2_meta_lock(old_inode, handle, NULL, 1);
-                                mlog_errno(status);
+        if (status < 0) {
-                        goto bail;
+                if (status != -ENOENT)
-                }
-                status = ocfs2_request_rename_vote(old_inode, old_dentry);
-                if (status < 0) {
                        mlog_errno(status);
-                        goto bail;
+                goto bail;
-                }
+        }
+        status = ocfs2_remote_dentry_delete(old_dentry);
+        if (status < 0) {
+                mlog_errno(status);
+                goto bail;
+        }
+        if (S_ISDIR(old_inode->i_mode)) {
                status = -EIO;
                old_inode_de_bh = ocfs2_bread(old_inode, 0, &status, 0);
                if (!old_inode_de_bh)
@@ -1124,14 +1176,6 @@ static int ocfs2_rename(struct inode *old_dir,
                if (!new_inode && new_dir!=old_dir &&
                    new_dir->i_nlink >= OCFS2_LINK_MAX)
                        goto bail;
-        } else {
-                /* Ah, the simple case - we're a file so just send a
-                 * message. */
-                status = ocfs2_request_rename_vote(old_inode, old_dentry);
-                if (status < 0) {
-                        mlog_errno(status);
-                        goto bail;
-                }
        }
        status = -ENOENT;
@@ -1203,13 +1247,7 @@ static int ocfs2_rename(struct inode *old_dir,
                        goto bail;
                }
-                if (S_ISDIR(new_inode->i_mode))
+                status = ocfs2_remote_dentry_delete(new_dentry);
-                        links_count = 0;
-                else
-                        links_count = (unsigned int) (new_inode->i_nlink - 1);
-                status = ocfs2_request_unlink_vote(new_inode, new_dentry,
-                                                   links_count);
                if (status < 0) {
                        mlog_errno(status);
                        goto bail;
@@ -1388,6 +1426,7 @@ static int ocfs2_rename(struct inode *old_dir,
                }
        }
+        ocfs2_dentry_move(old_dentry, new_dentry, old_dir, new_dir);
        status = 0;
 bail:
        if (rename_lock)
@@ -1676,6 +1715,12 @@ static int ocfs2_symlink(struct inode *dir,
                goto bail;
        }
+        status = ocfs2_dentry_attach_lock(dentry, inode, OCFS2_I(dir)->ip_blkno);
+        if (status) {
+                mlog_errno(status);
+                goto bail;
+        }
        insert_inode_hash(inode);
        dentry->d_op = &ocfs2_dentry_ops;
        d_instantiate(dentry, inode);
@@ -1964,13 +2009,8 @@ restart:
                                }
                                num++;
-                                /* XXX: questionable readahead stuff here */
                                bh = ocfs2_bread(dir, b++, &err, 1);
                                bh_use[ra_max] = bh;
-#if 0           // ???
-                                if (bh)
-                                        ll_rw_block(READ, 1, &bh);
-#endif
                        }
                }
                if ((bh = bh_use[ra_ptr++]) == NULL)
@@ -1978,6 +2018,10 @@ restart:
                wait_on_buffer(bh);
                if (!buffer_uptodate(bh)) {
                        /* read error, skip block & hope for the best */
+                        ocfs2_error(dir->i_sb, "reading directory %llu, "
+                                    "offset %lu\n",
+                                    (unsigned long long)OCFS2_I(dir)->ip_blkno,
+                                    block);
                        brelse(bh);
                        goto next;
                }
diff --git a/fs/ocfs2/ocfs2_fs.h b/fs/ocfs2/ocfs2_fs.h
index c5b1ac547c15..3330a5dc6be2 100644
--- a/fs/ocfs2/ocfs2_fs.h
+++ b/fs/ocfs2/ocfs2_fs.h
@@ -114,6 +114,26 @@
 #define OCFS2_CHAIN_FL          (0x00000400)    /* Chain allocator */
 #define OCFS2_DEALLOC_FL        (0x00000800)    /* Truncate log */
+/* Inode attributes, keep in sync with EXT2 */
+#define OCFS2_SECRM_FL          (0x00000001)    /* Secure deletion */
+#define OCFS2_UNRM_FL           (0x00000002)    /* Undelete */
+#define OCFS2_COMPR_FL          (0x00000004)    /* Compress file */
+#define OCFS2_SYNC_FL           (0x00000008)    /* Synchronous updates */
+#define OCFS2_IMMUTABLE_FL      (0x00000010)    /* Immutable file */
+#define OCFS2_APPEND_FL         (0x00000020)    /* writes to file may only append */
+#define OCFS2_NODUMP_FL         (0x00000040)    /* do not dump file */
+#define OCFS2_NOATIME_FL        (0x00000080)    /* do not update atime */
+#define OCFS2_DIRSYNC_FL        (0x00010000)    /* dirsync behaviour (directories only) */
+#define OCFS2_FL_VISIBLE        (0x000100FF)    /* User visible flags */
+#define OCFS2_FL_MODIFIABLE     (0x000100FF)    /* User modifiable flags */
+/*
+ * ioctl commands
+ */
+#define OCFS2_IOC_GETFLAGS      _IOR('f', 1, long)
+#define OCFS2_IOC_SETFLAGS      _IOW('f', 2, long)
 /*
 * Journal Flags (ocfs2_dinode.id1.journal1.i_flags)
 */
@@ -399,7 +419,9 @@ struct ocfs2_dinode {
        __le32 i_atime_nsec;
        __le32 i_ctime_nsec;
        __le32 i_mtime_nsec;
-/*70*/  __le64 i_reserved1[9];
+        __le32 i_attr;
+        __le32 i_reserved1;
+/*70*/  __le64 i_reserved2[8];
 /*B8*/  union {
                __le64 i_pad1;          /* Generic way to refer to this
                                           64bit union */
diff --git a/fs/ocfs2/ocfs2_lockid.h b/fs/ocfs2/ocfs2_lockid.h
index 7dd9e1e705b0..4d5d5655c185 100644
--- a/fs/ocfs2/ocfs2_lockid.h
+++ b/fs/ocfs2/ocfs2_lockid.h
@@ -35,12 +35,15 @@
 #define OCFS2_LOCK_ID_MAX_LEN  32
 #define OCFS2_LOCK_ID_PAD "000000"
+#define OCFS2_DENTRY_LOCK_INO_START 18
 enum ocfs2_lock_type {
        OCFS2_LOCK_TYPE_META = 0,
        OCFS2_LOCK_TYPE_DATA,
        OCFS2_LOCK_TYPE_SUPER,
        OCFS2_LOCK_TYPE_RENAME,
        OCFS2_LOCK_TYPE_RW,
+        OCFS2_LOCK_TYPE_DENTRY,
        OCFS2_NUM_LOCK_TYPES
 };
@@ -63,6 +66,9 @@ static inline char ocfs2_lock_type_char(enum ocfs2_lock_type type)
                case OCFS2_LOCK_TYPE_RW:
                        c = 'W';
                        break;
+                case OCFS2_LOCK_TYPE_DENTRY:
+                        c = 'N';
+                        break;
                default:
                        c = '\0';
        }
@@ -70,4 +76,23 @@ static inline char ocfs2_lock_type_char(enum ocfs2_lock_type type)
        return c;
 }
+static char *ocfs2_lock_type_strings[] = {
+        [OCFS2_LOCK_TYPE_META] = "Meta",
+        [OCFS2_LOCK_TYPE_DATA] = "Data",
+        [OCFS2_LOCK_TYPE_SUPER] = "Super",
+        [OCFS2_LOCK_TYPE_RENAME] = "Rename",
+        /* Need to differntiate from [R]ename.. serializing writes is the
+         * important job it does, anyway. */
+        [OCFS2_LOCK_TYPE_RW] = "Write/Read",
+        [OCFS2_LOCK_TYPE_DENTRY] = "Dentry",
+};
+static inline const char *ocfs2_lock_type_string(enum ocfs2_lock_type type)
+{
+#ifdef __KERNEL__
+        mlog_bug_on_msg(type >= OCFS2_NUM_LOCK_TYPES, "%d\n", type);
+#endif
+        return ocfs2_lock_type_strings[type];
+}
 #endif  /* OCFS2_LOCKID_H */
diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c
index d17e33e66a1e..4c29cd7cc8e6 100644
--- a/fs/ocfs2/super.c
+++ b/fs/ocfs2/super.c
@@ -202,7 +202,7 @@ static int ocfs2_init_global_system_inodes(struct ocfs2_super *osb)
        mlog_entry_void();
-        new = ocfs2_iget(osb, osb->root_blkno);
+        new = ocfs2_iget(osb, osb->root_blkno, OCFS2_FI_FLAG_SYSFILE);
        if (IS_ERR(new)) {
                status = PTR_ERR(new);
                mlog_errno(status);
@@ -210,7 +210,7 @@ static int ocfs2_init_global_system_inodes(struct ocfs2_super *osb)
        }
        osb->root_inode = new;
-        new = ocfs2_iget(osb, osb->system_dir_blkno);
+        new = ocfs2_iget(osb, osb->system_dir_blkno, OCFS2_FI_FLAG_SYSFILE);
        if (IS_ERR(new)) {
                status = PTR_ERR(new);
                mlog_errno(status);
@@ -682,7 +682,7 @@ static struct file_system_type ocfs2_fs_type = {
        .kill_sb        = kill_block_super, /* set to the generic one
                                             * right now, but do we
                                             * need to change that? */
-        .fs_flags       = FS_REQUIRES_DEV,
+        .fs_flags       = FS_REQUIRES_DEV|FS_RENAME_DOES_D_MOVE,
        .next           = NULL
 };
diff --git a/fs/ocfs2/sysfile.c b/fs/ocfs2/sysfile.c
index fc29cb7a437d..5df6e35d09b1 100644
--- a/fs/ocfs2/sysfile.c
+++ b/fs/ocfs2/sysfile.c
@@ -28,11 +28,11 @@
 #include <linux/slab.h>
 #include <linux/highmem.h>
-#include "ocfs2.h"
 #define MLOG_MASK_PREFIX ML_INODE
 #include <cluster/masklog.h>
+#include "ocfs2.h"
 #include "alloc.h"
 #include "dir.h"
 #include "inode.h"
@@ -115,7 +115,7 @@ static struct inode * _ocfs2_get_system_file_inode(struct ocfs2_super *osb,
                goto bail;
        }
-        inode = ocfs2_iget(osb, blkno);
+        inode = ocfs2_iget(osb, blkno, OCFS2_FI_FLAG_SYSFILE);
        if (IS_ERR(inode)) {
                mlog_errno(PTR_ERR(inode));
                inode = NULL;
diff --git a/fs/ocfs2/uptodate.c b/fs/ocfs2/uptodate.c
index b8a00a793326..9707ed7a3206 100644
--- a/fs/ocfs2/uptodate.c
+++ b/fs/ocfs2/uptodate.c
@@ -206,7 +206,10 @@ static int ocfs2_buffer_cached(struct ocfs2_inode_info *oi,
 }
 /* Warning: even if it returns true, this does *not* guarantee that
- * the block is stored in our inode metadata cache. */
+ * the block is stored in our inode metadata cache. 
+ * 
+ * This can be called under lock_buffer()
+ */
 int ocfs2_buffer_uptodate(struct inode *inode,
                          struct buffer_head *bh)
 {
@@ -226,6 +229,16 @@ int ocfs2_buffer_uptodate(struct inode *inode,
        return ocfs2_buffer_cached(OCFS2_I(inode), bh);
 }
+/* 
+ * Determine whether a buffer is currently out on a read-ahead request.
+ * ip_io_sem should be held to serialize submitters with the logic here.
+ */
+int ocfs2_buffer_read_ahead(struct inode *inode,
+                            struct buffer_head *bh)
+{
+        return buffer_locked(bh) && ocfs2_buffer_cached(OCFS2_I(inode), bh);
+}
 /* Requires ip_lock */
 static void ocfs2_append_cache_array(struct ocfs2_caching_info *ci,
                                     sector_t block)
@@ -403,7 +416,11 @@ out_free:
 *
 * Note that this function may actually fail to insert the block if
 * memory cannot be allocated. This is not fatal however (but may
- * result in a performance penalty) */
+ * result in a performance penalty)
+ *
+ * Readahead buffers can be passed in here before the I/O request is
+ * completed.
+ */
 void ocfs2_set_buffer_uptodate(struct inode *inode,
                               struct buffer_head *bh)
 {
diff --git a/fs/ocfs2/uptodate.h b/fs/ocfs2/uptodate.h
index 01cd32d26b06..2e73206059a8 100644
--- a/fs/ocfs2/uptodate.h
+++ b/fs/ocfs2/uptodate.h
@@ -40,5 +40,7 @@ void ocfs2_set_new_buffer_uptodate(struct inode *inode,
                                   struct buffer_head *bh);
 void ocfs2_remove_from_cache(struct inode *inode,
                             struct buffer_head *bh);
+int ocfs2_buffer_read_ahead(struct inode *inode,
+                            struct buffer_head *bh);
 #endif /* OCFS2_UPTODATE_H */
diff --git a/fs/ocfs2/vote.c b/fs/ocfs2/vote.c
index cf70fe2075b8..5b4dca79990b 100644
--- a/fs/ocfs2/vote.c
+++ b/fs/ocfs2/vote.c
@@ -74,9 +74,6 @@ struct ocfs2_vote_msg
                __be32 v_orphaned_slot; /* Used during delete votes */
                __be32 v_nlink;         /* Used during unlink votes */
        } md1;                          /* Message type dependant 1 */
-        __be32 v_unlink_namelen;
-        __be64 v_unlink_parent;
-        u8  v_unlink_dirent[OCFS2_VOTE_FILENAME_LEN];
 };
 /* Responses are given these values to maintain backwards
@@ -100,8 +97,6 @@ struct ocfs2_vote_work {
 enum ocfs2_vote_request {
        OCFS2_VOTE_REQ_INVALID = 0,
        OCFS2_VOTE_REQ_DELETE,
-        OCFS2_VOTE_REQ_UNLINK,
-        OCFS2_VOTE_REQ_RENAME,
        OCFS2_VOTE_REQ_MOUNT,
        OCFS2_VOTE_REQ_UMOUNT,
        OCFS2_VOTE_REQ_LAST
@@ -261,103 +256,13 @@ done:
        return response;
 }
-static int ocfs2_match_dentry(struct dentry *dentry,
-                              u64 parent_blkno,
-                              unsigned int namelen,
-                              const char *name)
-{
-        struct inode *parent;
-        if (!dentry->d_parent) {
-                mlog(0, "Detached from parent.\n");
-                return 0;
-        }
-        parent = dentry->d_parent->d_inode;
-        /* Negative parent dentry? */
-        if (!parent)
-                return 0;
-        /* Name is in a different directory. */
-        if (OCFS2_I(parent)->ip_blkno != parent_blkno)
-                return 0;
-        if (dentry->d_name.len != namelen)
-                return 0;
-        /* comparison above guarantees this is safe. */
-        if (memcmp(dentry->d_name.name, name, namelen))
-                return 0;
-        return 1;
-}
-static void ocfs2_process_dentry_request(struct inode *inode,
-                                         int rename,
-                                         unsigned int new_nlink,
-                                         u64 parent_blkno,
-                                         unsigned int namelen,
-                                         const char *name)
-{
-        struct dentry *dentry = NULL;
-        struct list_head *p;
-        struct ocfs2_inode_info *oi = OCFS2_I(inode);
-        mlog(0, "parent %llu, namelen = %u, name = %.*s\n",
-             (unsigned long long)parent_blkno, namelen, namelen, name);
-        spin_lock(&dcache_lock);
-        /* Another node is removing this name from the system. It is
-         * up to us to find the corresponding dentry and if it exists,
-         * unhash it from the dcache. */
-        list_for_each(p, &inode->i_dentry) {
-                dentry = list_entry(p, struct dentry, d_alias);
-                if (ocfs2_match_dentry(dentry, parent_blkno, namelen, name)) {
-                        mlog(0, "dentry found: %.*s\n",
-                             dentry->d_name.len, dentry->d_name.name);
-                        dget_locked(dentry);
-                        break;
-                }
-                dentry = NULL;
-        }
-        spin_unlock(&dcache_lock);
-        if (dentry) {
-                d_delete(dentry);
-                dput(dentry);
-        }
-        /* rename votes don't send link counts */
-        if (!rename) {
-                mlog(0, "new_nlink = %u\n", new_nlink);
-                /* We don't have the proper locks here to directly
-                 * change i_nlink and besides, the vote is sent
-                 * *before* the operation so it may have failed on the
-                 * other node. This passes a hint to ocfs2_drop_inode
-                 * to force ocfs2_delete_inode, who will take the
-                 * proper cluster locks to sort things out. */
-                if (new_nlink == 0) {
-                        spin_lock(&oi->ip_lock);
-                        oi->ip_flags |= OCFS2_INODE_MAYBE_ORPHANED;
-                        spin_unlock(&OCFS2_I(inode)->ip_lock);
-                }
-        }
-}
 static void ocfs2_process_vote(struct ocfs2_super *osb,
                               struct ocfs2_vote_msg *msg)
 {
        int net_status, vote_response;
        int orphaned_slot = 0;
-        int rename = 0;
+        unsigned int node_num, generation;
-        unsigned int node_num, generation, new_nlink, namelen;
+        u64 blkno;
-        u64 blkno, parent_blkno;
        enum ocfs2_vote_request request;
        struct inode *inode = NULL;
        struct ocfs2_msg_hdr *hdr = &msg->v_hdr;
@@ -437,18 +342,6 @@ static void ocfs2_process_vote(struct ocfs2_super *osb,
                vote_response = ocfs2_process_delete_request(inode,
                                                             &orphaned_slot);
                break;
-        case OCFS2_VOTE_REQ_RENAME:
-                rename = 1;
-                /* fall through */
-        case OCFS2_VOTE_REQ_UNLINK:
-                parent_blkno = be64_to_cpu(msg->v_unlink_parent);
-                namelen = be32_to_cpu(msg->v_unlink_namelen);
-                /* new_nlink will be ignored in case of a rename vote */
-                new_nlink = be32_to_cpu(msg->md1.v_nlink);
-                ocfs2_process_dentry_request(inode, rename, new_nlink,
-                                             parent_blkno, namelen,
-                                             msg->v_unlink_dirent);
-                break;
        default:
                mlog(ML_ERROR, "node %u, invalid request: %u\n",
                     node_num, request);
@@ -889,75 +782,6 @@ int ocfs2_request_delete_vote(struct inode *inode)
        return status;
 }
-static void ocfs2_setup_unlink_vote(struct ocfs2_vote_msg *request,
-                                    struct dentry *dentry)
-{
-        struct inode *parent = dentry->d_parent->d_inode;
-        /* We need some values which will uniquely identify a dentry
-         * on the other nodes so that they can find it and run
-         * d_delete against it. Parent directory block and full name
-         * should suffice. */
-        mlog(0, "unlink/rename request: parent: %llu name: %.*s\n",
-             (unsigned long long)OCFS2_I(parent)->ip_blkno, dentry->d_name.len,
-             dentry->d_name.name);
-        request->v_unlink_parent = cpu_to_be64(OCFS2_I(parent)->ip_blkno);
-        request->v_unlink_namelen = cpu_to_be32(dentry->d_name.len);
-        memcpy(request->v_unlink_dirent, dentry->d_name.name,
-               dentry->d_name.len);
-}
-int ocfs2_request_unlink_vote(struct inode *inode,
-                              struct dentry *dentry,
-                              unsigned int nlink)
-{
-        int status;
-        struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
-        struct ocfs2_vote_msg *request;
-        if (dentry->d_name.len > OCFS2_VOTE_FILENAME_LEN)
-                return -ENAMETOOLONG;
-        status = -ENOMEM;
-        request = ocfs2_new_vote_request(osb, OCFS2_I(inode)->ip_blkno,
-                                         inode->i_generation,
-                                         OCFS2_VOTE_REQ_UNLINK, nlink);
-        if (request) {
-                ocfs2_setup_unlink_vote(request, dentry);
-                status = ocfs2_request_vote(inode, request, NULL);
-                kfree(request);
-        }
-        return status;
-}
-int ocfs2_request_rename_vote(struct inode *inode,
-                              struct dentry *dentry)
-{
-        int status;
-        struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
-        struct ocfs2_vote_msg *request;
-        if (dentry->d_name.len > OCFS2_VOTE_FILENAME_LEN)
-                return -ENAMETOOLONG;
-        status = -ENOMEM;
-        request = ocfs2_new_vote_request(osb, OCFS2_I(inode)->ip_blkno,
-                                         inode->i_generation,
-                                         OCFS2_VOTE_REQ_RENAME, 0);
-        if (request) {
-                ocfs2_setup_unlink_vote(request, dentry);
-                status = ocfs2_request_vote(inode, request, NULL);
-                kfree(request);
-        }
-        return status;
-}
 int ocfs2_request_mount_vote(struct ocfs2_super *osb)
 {
        int status;
diff --git a/fs/ocfs2/vote.h b/fs/ocfs2/vote.h
index 9cce60703466..53ebc1c69e56 100644
--- a/fs/ocfs2/vote.h
+++ b/fs/ocfs2/vote.h
@@ -39,11 +39,6 @@ static inline void ocfs2_kick_vote_thread(struct ocfs2_super *osb)
 }
 int ocfs2_request_delete_vote(struct inode *inode);
-int ocfs2_request_unlink_vote(struct inode *inode,
-                              struct dentry *dentry,
-                              unsigned int nlink);
-int ocfs2_request_rename_vote(struct inode *inode,
-                              struct dentry *dentry);
 int ocfs2_request_mount_vote(struct ocfs2_super *osb);
 int ocfs2_request_umount_vote(struct ocfs2_super *osb);
 int ocfs2_register_net_handlers(struct ocfs2_super *osb);
diff --git a/fs/openpromfs/inode.c b/fs/openpromfs/inode.c
index 93a56bd4a2b7..592a6402e851 100644
--- a/fs/openpromfs/inode.c
+++ b/fs/openpromfs/inode.c
@@ -8,10 +8,10 @@
 #include <linux/types.h>
 #include <linux/string.h>
 #include <linux/fs.h>
-#include <linux/openprom_fs.h>
 #include <linux/init.h>
 #include <linux/slab.h>
 #include <linux/seq_file.h>
+#include <linux/magic.h>
 #include <asm/openprom.h>
 #include <asm/oplib.h>