Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6: (46 commits)
  [PATCH] fs: add a sanity check in d_free
  [PATCH] i_version: remount support
  [patch] vfs: make security_inode_setattr() calling consistent
  [patch 1/3] FS_MBCACHE: don't needlessly make it built-in
  [PATCH] move executable checking into ->permission()
  [PATCH] fs/dcache.c: update comment of d_validate()
  [RFC PATCH] touch_mnt_namespace when the mount flags change
  [PATCH] reiserfs: add missing llseek method
  [PATCH] fix ->llseek for more directories
  [PATCH vfs-2.6 6/6] vfs: add LOOKUP_RENAME_TARGET intent
  [PATCH vfs-2.6 5/6] vfs: remove LOOKUP_PARENT from non LOOKUP_PARENT lookup
  [PATCH vfs-2.6 4/6] vfs: remove unnecessary fsnotify_d_instantiate()
  [PATCH vfs-2.6 3/6] vfs: add __d_instantiate() helper
  [PATCH vfs-2.6 2/6] vfs: add d_ancestor()
  [PATCH vfs-2.6 1/6] vfs: replace parent == dentry->d_parent by IS_ROOT()
  [PATCH] get rid of on-stack dentry in udf
  [PATCH 2/2] anondev: switch to IDA
  [PATCH 1/2] anondev: init IDR statically
  [JFFS2] Use d_splice_alias() not d_add() in jffs2_lookup()
  [PATCH] Optimise NFS readdir hack slightly.
  ...

arch/alpha/kernel/osf_sys.c

@@ -165,14 +165,11 @@ osf_getdirentries(unsigned int fd, struct osf_dirent __user *dirent,
 	buf.error = 0;
 
 	error = vfs_readdir(file, osf_filldir, &buf);
-	if (error < 0)
-		goto out_putf;
-
-	error = buf.error;
+	if (error >= 0)
+		error = buf.error;
 	if (count != buf.count)
 		error = count - buf.count;
 
- out_putf:
 	fput(file);
  out:
 	return error;

arch/parisc/hpux/fs.c

@@ -127,9 +127,8 @@ int hpux_getdents(unsigned int fd, struct hpux_dirent __user *dirent, unsigned i
 	buf.error = 0;
 
 	error = vfs_readdir(file, filldir, &buf);
-	if (error < 0)
-		goto out_putf;
-	error = buf.error;
+	if (error >= 0)
+		error = buf.error;
 	lastdirent = buf.previous;
 	if (lastdirent) {
 		if (put_user(file->f_pos, &lastdirent->d_off))

arch/powerpc/platforms/cell/spufs/syscalls.c

@@ -69,9 +69,9 @@ static long do_spu_create(const char __user *pathname, unsigned int flags,
 	if (!IS_ERR(tmp)) {
 		struct nameidata nd;
 
-		ret = path_lookup(tmp, LOOKUP_PARENT|
-				LOOKUP_OPEN|LOOKUP_CREATE, &nd);
+		ret = path_lookup(tmp, LOOKUP_PARENT, &nd);
 		if (!ret) {
+			nd.flags |= LOOKUP_OPEN | LOOKUP_CREATE;
 			ret = spufs_create(&nd, flags, mode, neighbor);
 			path_put(&nd.path);
 		}

drivers/md/dm-table.c

@@ -312,19 +312,6 @@ static inline int check_space(struct dm_table *t)
 	return 0;
 }
 
-/*
- * Convert a device path to a dev_t.
- */
-static int lookup_device(const char *path, dev_t *dev)
-{
-	struct block_device *bdev = lookup_bdev(path);
-	if (IS_ERR(bdev))
-		return PTR_ERR(bdev);
-	*dev = bdev->bd_dev;
-	bdput(bdev);
-	return 0;
-}
-
 /*
  * See if we've already got a device in the list.
  */
@@ -437,8 +424,12 @@ static int __table_get_device(struct dm_table *t, struct dm_target *ti,
 			return -EOVERFLOW;
 	} else {
 		/* convert the path to a device */
-		if ((r = lookup_device(path, &dev)))
-			return r;
+		struct block_device *bdev = lookup_bdev(path);
+
+		if (IS_ERR(bdev))
+			return PTR_ERR(bdev);
+		dev = bdev->bd_dev;
+		bdput(bdev);
 	}
 
 	dd = find_device(&t->devices, dev);

fs/Kconfig

@@ -22,9 +22,10 @@ source "fs/jbd2/Kconfig"
 config FS_MBCACHE
 # Meta block cache for Extended Attributes (ext2/ext3/ext4)
 	tristate
-	depends on EXT2_FS_XATTR || EXT3_FS_XATTR || EXT4_FS_XATTR
-	default y if EXT2_FS=y || EXT3_FS=y || EXT4_FS=y
-	default m if EXT2_FS=m || EXT3_FS=m || EXT4_FS=m
+	default y if EXT2_FS=y && EXT2_FS_XATTR
+	default y if EXT3_FS=y && EXT3_FS_XATTR
+	default y if EXT4_FS=y && EXT4_FS_XATTR
+	default m if EXT2_FS_XATTR || EXT3_FS_XATTR || EXT4_FS_XATTR
 
 config REISERFS_FS
 	tristate "Reiserfs support"

fs/afs/dir.c

@@ -45,6 +45,7 @@ const struct file_operations afs_dir_file_operations = {
 	.release	= afs_release,
 	.readdir	= afs_readdir,
 	.lock		= afs_lock,
+	.llseek		= generic_file_llseek,
 };
 
 const struct inode_operations afs_dir_inode_operations = {

fs/attr.c

@@ -159,17 +159,17 @@ int notify_change(struct dentry * dentry, struct iattr * attr)
 	if (!(attr->ia_valid & ~(ATTR_KILL_SUID | ATTR_KILL_SGID)))
 		return 0;
 
+	error = security_inode_setattr(dentry, attr);
+	if (error)
+		return error;
+
 	if (ia_valid & ATTR_SIZE)
 		down_write(&dentry->d_inode->i_alloc_sem);
 
 	if (inode->i_op && inode->i_op->setattr) {
-		error = security_inode_setattr(dentry, attr);
-		if (!error)
-			error = inode->i_op->setattr(dentry, attr);
+		error = inode->i_op->setattr(dentry, attr);
 	} else {
 		error = inode_change_ok(inode, attr);
-		if (!error)
-			error = security_inode_setattr(dentry, attr);
 		if (!error) {
 			if ((ia_valid & ATTR_UID && attr->ia_uid != inode->i_uid) ||
 			    (ia_valid & ATTR_GID && attr->ia_gid != inode->i_gid))

fs/bfs/dir.c

@@ -80,6 +80,7 @@ const struct file_operations bfs_dir_operations = {
 	.read		= generic_read_dir,
 	.readdir	= bfs_readdir,
 	.fsync		= file_fsync,
+	.llseek		= generic_file_llseek,
 };
 
 extern void dump_imap(const char *, struct super_block *);

fs/block_dev.c

@@ -1268,33 +1268,33 @@ EXPORT_SYMBOL(ioctl_by_bdev);
  * namespace if possible and return it.  Return ERR_PTR(error)
  * otherwise.
  */
-struct block_device *lookup_bdev(const char *path)
+struct block_device *lookup_bdev(const char *pathname)
 {
 	struct block_device *bdev;
 	struct inode *inode;
-	struct nameidata nd;
+	struct path path;
 	int error;
 
-	if (!path || !*path)
+	if (!pathname || !*pathname)
 		return ERR_PTR(-EINVAL);
 
-	error = path_lookup(path, LOOKUP_FOLLOW, &nd);
+	error = kern_path(pathname, LOOKUP_FOLLOW, &path);
 	if (error)
 		return ERR_PTR(error);
 
-	inode = nd.path.dentry->d_inode;
+	inode = path.dentry->d_inode;
 	error = -ENOTBLK;
 	if (!S_ISBLK(inode->i_mode))
 		goto fail;
 	error = -EACCES;
-	if (nd.path.mnt->mnt_flags & MNT_NODEV)
+	if (path.mnt->mnt_flags & MNT_NODEV)
 		goto fail;
 	error = -ENOMEM;
 	bdev = bd_acquire(inode);
 	if (!bdev)
 		goto fail;
 out:
-	path_put(&nd.path);
+	path_put(&path);
 	return bdev;
 fail:
 	bdev = ERR_PTR(error);

fs/char_dev.c

@@ -386,15 +386,22 @@ static int chrdev_open(struct inode *inode, struct file *filp)
 	cdev_put(new);
 	if (ret)
 		return ret;
+
+	ret = -ENXIO;
 	filp->f_op = fops_get(p->ops);
-	if (!filp->f_op) {
-		cdev_put(p);
-		return -ENXIO;
-	}
-	if (filp->f_op->open)
+	if (!filp->f_op)
+		goto out_cdev_put;
+
+	if (filp->f_op->open) {
 		ret = filp->f_op->open(inode,filp);
-	if (ret)
-		cdev_put(p);
+		if (ret)
+			goto out_cdev_put;
+	}
+
+	return 0;
+
+ out_cdev_put:
+	cdev_put(p);
 	return ret;
 }
 

fs/cifs/cifsfs.c

@@ -275,9 +275,12 @@ static int cifs_permission(struct inode *inode, int mask)
 
 	cifs_sb = CIFS_SB(inode->i_sb);
 
-	if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_PERM)
-		return 0;
-	else /* file mode might have been restricted at mount time
+	if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_PERM) {
+		if ((mask & MAY_EXEC) && !execute_ok(inode))
+			return -EACCES;
+		else
+			return 0;
+	} else /* file mode might have been restricted at mount time
 		on the client (above and beyond ACL on servers) for
 		servers which do not support setting and viewing mode bits,
 		so allowing client to check permissions is useful */
@@ -765,6 +768,7 @@ const struct file_operations cifs_dir_ops = {
 	.dir_notify = cifs_dir_notify,
 #endif /* CONFIG_CIFS_EXPERIMENTAL */
 	.unlocked_ioctl  = cifs_ioctl,
+	.llseek = generic_file_llseek,
 };
 
 static void

fs/coda/dir.c

@@ -146,6 +146,9 @@ int coda_permission(struct inode *inode, int mask)
 	if (!mask)
 		return 0; 
 
+	if ((mask & MAY_EXEC) && !execute_ok(inode))
+		return -EACCES;
+
 	lock_kernel();
 
 	if (coda_cache_check(inode, mask))

fs/coda/pioctl.c

@@ -43,7 +43,7 @@ const struct file_operations coda_ioctl_operations = {
 /* the coda pioctl inode ops */
 static int coda_ioctl_permission(struct inode *inode, int mask)
 {
-        return 0;
+	return (mask & MAY_EXEC) ? -EACCES : 0;
 }
 
 static int coda_pioctl(struct inode * inode, struct file * filp, 

fs/compat.c

@@ -869,7 +869,7 @@ asmlinkage long compat_sys_old_readdir(unsigned int fd,
 	buf.dirent = dirent;
 
 	error = vfs_readdir(file, compat_fillonedir, &buf);
-	if (error >= 0)
+	if (buf.result)
 		error = buf.result;
 
 	fput(file);
@@ -956,9 +956,8 @@ asmlinkage long compat_sys_getdents(unsigned int fd,
 	buf.error = 0;
 
 	error = vfs_readdir(file, compat_filldir, &buf);
-	if (error < 0)
-		goto out_putf;
-	error = buf.error;
+	if (error >= 0)
+		error = buf.error;
 	lastdirent = buf.previous;
 	if (lastdirent) {
 		if (put_user(file->f_pos, &lastdirent->d_off))
@@ -966,8 +965,6 @@ asmlinkage long compat_sys_getdents(unsigned int fd,
 		else
 			error = count - buf.count;
 	}
-
-out_putf:
 	fput(file);
 out:
 	return error;
@@ -1047,19 +1044,16 @@ asmlinkage long compat_sys_getdents64(unsigned int fd,
 	buf.error = 0;
 
 	error = vfs_readdir(file, compat_filldir64, &buf);
-	if (error < 0)
-		goto out_putf;
-	error = buf.error;
+	if (error >= 0)
+		error = buf.error;
 	lastdirent = buf.previous;
 	if (lastdirent) {
 		typeof(lastdirent->d_off) d_off = file->f_pos;
-		error = -EFAULT;
 		if (__put_user_unaligned(d_off, &lastdirent->d_off))
-			goto out_putf;
-		error = count - buf.count;
+			error = -EFAULT;
+		else
+			error = count - buf.count;
 	}
-
-out_putf:
 	fput(file);
 out:
 	return error;

fs/configfs/symlink.c

@@ -108,18 +108,18 @@ out:
 }
 
 
-static int get_target(const char *symname, struct nameidata *nd,
+static int get_target(const char *symname, struct path *path,
 		      struct config_item **target)
 {
 	int ret;
 
-	ret = path_lookup(symname, LOOKUP_FOLLOW|LOOKUP_DIRECTORY, nd);
+	ret = kern_path(symname, LOOKUP_FOLLOW|LOOKUP_DIRECTORY, path);
 	if (!ret) {
-		if (nd->path.dentry->d_sb == configfs_sb) {
-			*target = configfs_get_config_item(nd->path.dentry);
+		if (path->dentry->d_sb == configfs_sb) {
+			*target = configfs_get_config_item(path->dentry);
 			if (!*target) {
 				ret = -ENOENT;
-				path_put(&nd->path);
+				path_put(path);
 			}
 		} else
 			ret = -EPERM;
@@ -132,7 +132,7 @@ static int get_target(const char *symname, struct nameidata *nd,
 int configfs_symlink(struct inode *dir, struct dentry *dentry, const char *symname)
 {
 	int ret;
-	struct nameidata nd;
+	struct path path;
 	struct configfs_dirent *sd;
 	struct config_item *parent_item;
 	struct config_item *target_item;
@@ -159,7 +159,7 @@ int configfs_symlink(struct inode *dir, struct dentry *dentry, const char *symna
 	    !type->ct_item_ops->allow_link)
 		goto out_put;
 
-	ret = get_target(symname, &nd, &target_item);
+	ret = get_target(symname, &path, &target_item);
 	if (ret)
 		goto out_put;
 
@@ -174,7 +174,7 @@ int configfs_symlink(struct inode *dir, struct dentry *dentry, const char *symna
 	}
 
 	config_item_put(target_item);
-	path_put(&nd.path);
+	path_put(&path);
 
 out_put:
 	config_item_put(parent_item);

fs/dcache.c

@@ -69,6 +69,7 @@ struct dentry_stat_t dentry_stat = {
 
 static void __d_free(struct dentry *dentry)
 {
+	WARN_ON(!list_empty(&dentry->d_alias));
 	if (dname_external(dentry))
 		kfree(dentry->d_name.name);
 	kmem_cache_free(dentry_cache, dentry); 
@@ -174,9 +175,12 @@ static struct dentry *d_kill(struct dentry *dentry)
 	dentry_stat.nr_dentry--;	/* For d_free, below */
 	/*drops the locks, at that point nobody can reach this dentry */
 	dentry_iput(dentry);
-	parent = dentry->d_parent;
+	if (IS_ROOT(dentry))
+		parent = NULL;
+	else
+		parent = dentry->d_parent;
 	d_free(dentry);
-	return dentry == parent ? NULL : parent;
+	return parent;
 }
 
 /* 
@@ -666,11 +670,12 @@ static void shrink_dcache_for_umount_subtree(struct dentry *dentry)
 				BUG();
 			}
 
-			parent = dentry->d_parent;
-			if (parent == dentry)
+			if (IS_ROOT(dentry))
 				parent = NULL;
-			else
+			else {
+				parent = dentry->d_parent;
 				atomic_dec(&parent->d_count);
+			}
 
 			list_del(&dentry->d_u.d_child);
 			detached++;
@@ -977,6 +982,15 @@ struct dentry *d_alloc_name(struct dentry *parent, const char *name)
 	return d_alloc(parent, &q);
 }
 
+/* the caller must hold dcache_lock */
+static void __d_instantiate(struct dentry *dentry, struct inode *inode)
+{
+	if (inode)
+		list_add(&dentry->d_alias, &inode->i_dentry);
+	dentry->d_inode = inode;
+	fsnotify_d_instantiate(dentry, inode);
+}
+
 /**
  * d_instantiate - fill in inode information for a dentry
  * @entry: dentry to complete
@@ -996,10 +1010,7 @@ void d_instantiate(struct dentry *entry, struct inode * inode)
 {
 	BUG_ON(!list_empty(&entry->d_alias));
 	spin_lock(&dcache_lock);
-	if (inode)
-		list_add(&entry->d_alias, &inode->i_dentry);
-	entry->d_inode = inode;
-	fsnotify_d_instantiate(entry, inode);
+	__d_instantiate(entry, inode);
 	spin_unlock(&dcache_lock);
 	security_d_instantiate(entry, inode);
 }
@@ -1029,7 +1040,7 @@ static struct dentry *__d_instantiate_unique(struct dentry *entry,
 	unsigned int hash = entry->d_name.hash;
 
 	if (!inode) {
-		entry->d_inode = NULL;
+		__d_instantiate(entry, NULL);
 		return NULL;
 	}
 
@@ -1048,9 +1059,7 @@ static struct dentry *__d_instantiate_unique(struct dentry *entry,
 		return alias;
 	}
 
-	list_add(&entry->d_alias, &inode->i_dentry);
-	entry->d_inode = inode;
-	fsnotify_d_instantiate(entry, inode);
+	__d_instantiate(entry, inode);
 	return NULL;
 }
 
@@ -1111,69 +1120,71 @@ static inline struct hlist_head *d_hash(struct dentry *parent,
 }
 
 /**
- * d_alloc_anon - allocate an anonymous dentry
+ * d_obtain_alias - find or allocate a dentry for a given inode
  * @inode: inode to allocate the dentry for
  *
- * This is similar to d_alloc_root.  It is used by filesystems when
- * creating a dentry for a given inode, often in the process of 
- * mapping a filehandle to a dentry.  The returned dentry may be
- * anonymous, or may have a full name (if the inode was already
- * in the cache).  The file system may need to make further
- * efforts to connect this dentry into the dcache properly.
+ * Obtain a dentry for an inode resulting from NFS filehandle conversion or
+ * similar open by handle operations.  The returned dentry may be anonymous,
+ * or may have a full name (if the inode was already in the cache).
  *
- * When called on a directory inode, we must ensure that
- * the inode only ever has one dentry.  If a dentry is
- * found, that is returned instead of allocating a new one.
+ * When called on a directory inode, we must ensure that the inode only ever
+ * has one dentry.  If a dentry is found, that is returned instead of
+ * allocating a new one.
  *
  * On successful return, the reference to the inode has been transferred
- * to the dentry.  If %NULL is returned (indicating kmalloc failure),
- * the reference on the inode has not been released.
+ * to the dentry.  In case of an error the reference on the inode is released.
+ * To make it easier to use in export operations a %NULL or IS_ERR inode may
+ * be passed in and will be the error will be propagate to the return value,
+ * with a %NULL @inode replaced by ERR_PTR(-ESTALE).
  */
-
-struct dentry * d_alloc_anon(struct inode *inode)
+struct dentry *d_obtain_alias(struct inode *inode)
 {
 	static const struct qstr anonstring = { .name = "" };
 	struct dentry *tmp;
 	struct dentry *res;
 
-	if ((res = d_find_alias(inode))) {
-		iput(inode);
-		return res;
-	}
+	if (!inode)
+		return ERR_PTR(-ESTALE);
+	if (IS_ERR(inode))
+		return ERR_CAST(inode);
 
-	tmp = d_alloc(NULL, &anonstring);
-	if (!tmp)
-		return NULL;
+	res = d_find_alias(inode);
+	if (res)
+		goto out_iput;
 
+	tmp = d_alloc(NULL, &anonstring);
+	if (!tmp) {
+		res = ERR_PTR(-ENOMEM);
+		goto out_iput;
+	}
 	tmp->d_parent = tmp; /* make sure dput doesn't croak */
-	
+
 	spin_lock(&dcache_lock);
 	res = __d_find_alias(inode, 0);
-	if (!res) {
-		/* attach a disconnected dentry */
-		res = tmp;
-		tmp = NULL;
-		spin_lock(&res->d_lock);
-		res->d_sb = inode->i_sb;
-		res->d_parent = res;
-		res->d_inode = inode;
-		res->d_flags |= DCACHE_DISCONNECTED;
-		res->d_flags &= ~DCACHE_UNHASHED;
-		list_add(&res->d_alias, &inode->i_dentry);
-		hlist_add_head(&res->d_hash, &inode->i_sb->s_anon);
-		spin_unlock(&res->d_lock);
-
-		inode = NULL; /* don't drop reference */
+	if (res) {
+		spin_unlock(&dcache_lock);
+		dput(tmp);
+		goto out_iput;
 	}
+
+	/* attach a disconnected dentry */
+	spin_lock(&tmp->d_lock);
+	tmp->d_sb = inode->i_sb;
+	tmp->d_inode = inode;
+	tmp->d_flags |= DCACHE_DISCONNECTED;
+	tmp->d_flags &= ~DCACHE_UNHASHED;
+	list_add(&tmp->d_alias, &inode->i_dentry);
+	hlist_add_head(&tmp->d_hash, &inode->i_sb->s_anon);
+	spin_unlock(&tmp->d_lock);
+
 	spin_unlock(&dcache_lock);
+	return tmp;
 
-	if (inode)
-		iput(inode);
-	if (tmp)
-		dput(tmp);
+ out_iput:
+	iput(inode);
 	return res;
 }
-
+EXPORT_SYMBOL_GPL(d_obtain_alias);
 
 /**
  * d_splice_alias - splice a disconnected dentry into the tree if one exists
@@ -1200,17 +1211,14 @@ struct dentry *d_splice_alias(struct inode *inode, struct dentry *dentry)
 		new = __d_find_alias(inode, 1);
 		if (new) {
 			BUG_ON(!(new->d_flags & DCACHE_DISCONNECTED));
-			fsnotify_d_instantiate(new, inode);
 			spin_unlock(&dcache_lock);
 			security_d_instantiate(new, inode);
 			d_rehash(dentry);
 			d_move(new, dentry);
 			iput(inode);
 		} else {
-			/* d_instantiate takes dcache_lock, so we do it by hand */
-			list_add(&dentry->d_alias, &inode->i_dentry);
-			dentry->d_inode = inode;
-			fsnotify_d_instantiate(dentry, inode);
+			/* already taking dcache_lock, so d_add() by hand */
+			__d_instantiate(dentry, inode);
 			spin_unlock(&dcache_lock);
 			security_d_instantiate(dentry, inode);
 			d_rehash(dentry);
@@ -1293,8 +1301,7 @@ struct dentry *d_add_ci(struct dentry *dentry, struct inode *inode,
 		 * d_instantiate() by hand because it takes dcache_lock which
 		 * we already hold.
 		 */
-		list_add(&found->d_alias, &inode->i_dentry);
-		found->d_inode = inode;
+		__d_instantiate(found, inode);
 		spin_unlock(&dcache_lock);
 		security_d_instantiate(found, inode);
 		return found;
@@ -1456,8 +1463,6 @@ out:
  * d_validate - verify dentry provided from insecure source
  * @dentry: The dentry alleged to be valid child of @dparent
  * @dparent: The parent dentry (known to be valid)
- * @hash: Hash of the dentry
- * @len: Length of the name
  *
  * An insecure source has sent us a dentry, here we verify it and dget() it.
  * This is used by ncpfs in its readdir implementation.
@@ -1714,18 +1719,23 @@ void d_move(struct dentry * dentry, struct dentry * target)
 	spin_unlock(&dcache_lock);
 }
 
-/*
- * Helper that returns 1 if p1 is a parent of p2, else 0
+/**
+ * d_ancestor - search for an ancestor
+ * @p1: ancestor dentry
+ * @p2: child dentry
+ *
+ * Returns the ancestor dentry of p2 which is a child of p1, if p1 is
+ * an ancestor of p2, else NULL.
  */
-static int d_isparent(struct dentry *p1, struct dentry *p2)
+struct dentry *d_ancestor(struct dentry *p1, struct dentry *p2)
 {
 	struct dentry *p;
 
-	for (p = p2; p->d_parent != p; p = p->d_parent) {
+	for (p = p2; !IS_ROOT(p); p = p->d_parent) {
 		if (p->d_parent == p1)
-			return 1;
+			return p;
 	}
-	return 0;
+	return NULL;
 }
 
 /*
@@ -1749,7 +1759,7 @@ static struct dentry *__d_unalias(struct dentry *dentry, struct dentry *alias)
 
 	/* Check for loops */
 	ret = ERR_PTR(-ELOOP);
-	if (d_isparent(alias, dentry))
+	if (d_ancestor(alias, dentry))
 		goto out_err;
 
 	/* See lock_rename() */
@@ -1822,7 +1832,7 @@ struct dentry *d_materialise_unique(struct dentry *dentry, struct inode *inode)
 
 	if (!inode) {
 		actual = dentry;
-		dentry->d_inode = NULL;
+		__d_instantiate(dentry, NULL);
 		goto found_lock;
 	}
 
@@ -2149,32 +2159,27 @@ out:
  * Caller must ensure that "new_dentry" is pinned before calling is_subdir()
  */
   
-int is_subdir(struct dentry * new_dentry, struct dentry * old_dentry)
+int is_subdir(struct dentry *new_dentry, struct dentry *old_dentry)
 {
 	int result;
-	struct dentry * saved = new_dentry;
 	unsigned long seq;
 
-	/* need rcu_readlock to protect against the d_parent trashing due to
-	 * d_move
+	/* FIXME: This is old behavior, needed? Please check callers. */
+	if (new_dentry == old_dentry)
+		return 1;
+
+	/*
+	 * Need rcu_readlock to protect against the d_parent trashing
+	 * due to d_move
 	 */
 	rcu_read_lock();
-        do {
+	do {
 		/* for restarting inner loop in case of seq retry */
-		new_dentry = saved;
-		result = 0;
 		seq = read_seqbegin(&rename_lock);
-		for (;;) {
-			if (new_dentry != old_dentry) {
-				struct dentry * parent = new_dentry->d_parent;
-				if (parent == new_dentry)
-					break;
-				new_dentry = parent;
-				continue;
-			}
+		if (d_ancestor(old_dentry, new_dentry))
 			result = 1;
-			break;
-		}
+		else
+			result = 0;
 	} while (read_seqretry(&rename_lock, seq));
 	rcu_read_unlock();
 
@@ -2344,7 +2349,6 @@ void __init vfs_caches_init(unsigned long mempages)
 }
 
 EXPORT_SYMBOL(d_alloc);
-EXPORT_SYMBOL(d_alloc_anon);
 EXPORT_SYMBOL(d_alloc_root);
 EXPORT_SYMBOL(d_delete);
 EXPORT_SYMBOL(d_find_alias);

fs/dquot.c

@@ -1805,19 +1805,19 @@ int vfs_quota_on_path(struct super_block *sb, int type, int format_id,
 }
 
 /* Actual function called from quotactl() */
-int vfs_quota_on(struct super_block *sb, int type, int format_id, char *path,
+int vfs_quota_on(struct super_block *sb, int type, int format_id, char *name,
 		 int remount)
 {
-	struct nameidata nd;
+	struct path path;
 	int error;
 
 	if (remount)
 		return vfs_quota_on_remount(sb, type);
 
-	error = path_lookup(path, LOOKUP_FOLLOW, &nd);
+	error = kern_path(name, LOOKUP_FOLLOW, &path);
 	if (!error) {
-		error = vfs_quota_on_path(sb, type, format_id, &nd.path);
-		path_put(&nd.path);
+		error = vfs_quota_on_path(sb, type, format_id, &path);
+		path_put(&path);
 	}
 	return error;
 }

fs/ecryptfs/main.c

@@ -471,31 +471,26 @@ out:
  */
 static int ecryptfs_read_super(struct super_block *sb, const char *dev_name)
 {
+	struct path path;
 	int rc;
-	struct nameidata nd;
-	struct dentry *lower_root;
-	struct vfsmount *lower_mnt;
 
-	memset(&nd, 0, sizeof(struct nameidata));
-	rc = path_lookup(dev_name, LOOKUP_FOLLOW | LOOKUP_DIRECTORY, &nd);
+	rc = kern_path(dev_name, LOOKUP_FOLLOW | LOOKUP_DIRECTORY, &path);
 	if (rc) {
 		ecryptfs_printk(KERN_WARNING, "path_lookup() failed\n");
 		goto out;
 	}
-	lower_root = nd.path.dentry;
-	lower_mnt = nd.path.mnt;
-	ecryptfs_set_superblock_lower(sb, lower_root->d_sb);
-	sb->s_maxbytes = lower_root->d_sb->s_maxbytes;
-	sb->s_blocksize = lower_root->d_sb->s_blocksize;
-	ecryptfs_set_dentry_lower(sb->s_root, lower_root);
-	ecryptfs_set_dentry_lower_mnt(sb->s_root, lower_mnt);
-	rc = ecryptfs_interpose(lower_root, sb->s_root, sb, 0);
+	ecryptfs_set_superblock_lower(sb, path.dentry->d_sb);
+	sb->s_maxbytes = path.dentry->d_sb->s_maxbytes;
+	sb->s_blocksize = path.dentry->d_sb->s_blocksize;
+	ecryptfs_set_dentry_lower(sb->s_root, path.dentry);
+	ecryptfs_set_dentry_lower_mnt(sb->s_root, path.mnt);
+	rc = ecryptfs_interpose(path.dentry, sb->s_root, sb, 0);
 	if (rc)
 		goto out_free;
 	rc = 0;
 	goto out;
 out_free:
-	path_put(&nd.path);
+	path_put(&path);
 out:
 	return rc;
 }

fs/efs/namei.c

@@ -112,35 +112,14 @@ struct dentry *efs_fh_to_parent(struct super_block *sb, struct fid *fid,
 
 struct dentry *efs_get_parent(struct dentry *child)
 {
-	struct dentry *parent;
-	struct inode *inode;
+	struct dentry *parent = ERR_PTR(-ENOENT);
 	efs_ino_t ino;
-	long error;
 
 	lock_kernel();
-
-	error = -ENOENT;
 	ino = efs_find_entry(child->d_inode, "..", 2);
-	if (!ino)
-		goto fail;
-
-	inode = efs_iget(child->d_inode->i_sb, ino);
-	if (IS_ERR(inode)) {
-		error = PTR_ERR(inode);
-		goto fail;
-	}
-
-	error = -ENOMEM;
-	parent = d_alloc_anon(inode);
-	if (!parent)
-		goto fail_iput;
-
+	if (ino)
+		parent = d_obtain_alias(efs_iget(child->d_inode->i_sb, ino));
 	unlock_kernel();
-	return parent;
 
- fail_iput:
-	iput(inode);
- fail:
-	unlock_kernel();
-	return ERR_PTR(error);
+	return parent;
 }

fs/exportfs/expfs.c

@@ -94,9 +94,8 @@ find_disconnected_root(struct dentry *dentry)
  * It may already be, as the flag isn't always updated when connection happens.
  */
 static int
-reconnect_path(struct vfsmount *mnt, struct dentry *target_dir)
+reconnect_path(struct vfsmount *mnt, struct dentry *target_dir, char *nbuf)
 {
-	char nbuf[NAME_MAX+1];
 	int noprogress = 0;
 	int err = -ESTALE;
 
@@ -281,13 +280,14 @@ static int get_name(struct vfsmount *mnt, struct dentry *dentry,
 		int old_seq = buffer.sequence;
 
 		error = vfs_readdir(file, filldir_one, &buffer);
+		if (buffer.found) {
+			error = 0;
+			break;
+		}
 
 		if (error < 0)
 			break;
 
-		error = 0;
-		if (buffer.found)
-			break;
 		error = -ENOENT;
 		if (old_seq == buffer.sequence)
 			break;
@@ -360,14 +360,13 @@ struct dentry *exportfs_decode_fh(struct vfsmount *mnt, struct fid *fid,
 {
 	const struct export_operations *nop = mnt->mnt_sb->s_export_op;
 	struct dentry *result, *alias;
+	char nbuf[NAME_MAX+1];
 	int err;
 
 	/*
 	 * Try to get any dentry for the given file handle from the filesystem.
 	 */
 	result = nop->fh_to_dentry(mnt->mnt_sb, fid, fh_len, fileid_type);
-	if (!result)
-		result = ERR_PTR(-ESTALE);
 	if (IS_ERR(result))
 		return result;
 
@@ -381,7 +380,7 @@ struct dentry *exportfs_decode_fh(struct vfsmount *mnt, struct fid *fid,
 		 * filesystem root.
 		 */
 		if (result->d_flags & DCACHE_DISCONNECTED) {
-			err = reconnect_path(mnt, result);
+			err = reconnect_path(mnt, result, nbuf);
 			if (err)
 				goto err_result;
 		}
@@ -397,7 +396,6 @@ struct dentry *exportfs_decode_fh(struct vfsmount *mnt, struct fid *fid,
 		 * It's not a directory.  Life is a little more complicated.
 		 */
 		struct dentry *target_dir, *nresult;
-		char nbuf[NAME_MAX+1];
 
 		/*
 		 * See if either the dentry we just got from the filesystem
@@ -422,8 +420,6 @@ struct dentry *exportfs_decode_fh(struct vfsmount *mnt, struct fid *fid,
 
 		target_dir = nop->fh_to_parent(mnt->mnt_sb, fid,
 				fh_len, fileid_type);
-		if (!target_dir)
-			goto err_result;
 		err = PTR_ERR(target_dir);
 		if (IS_ERR(target_dir))
 			goto err_result;
@@ -433,7 +429,7 @@ struct dentry *exportfs_decode_fh(struct vfsmount *mnt, struct fid *fid,
 		 * connected to the filesystem root.  The VFS really doesn't
 		 * like disconnected directories..
 		 */
-		err = reconnect_path(mnt, target_dir);
+		err = reconnect_path(mnt, target_dir, nbuf);
 		if (err) {
 			dput(target_dir);
 			goto err_result;

fs/ext2/dir.c

@@ -354,11 +354,11 @@ ext2_readdir (struct file * filp, void * dirent, filldir_t filldir)
  * (as a parameter - res_dir). Page is returned mapped and unlocked.
  * Entry is guaranteed to be valid.
  */
-struct ext2_dir_entry_2 * ext2_find_entry (struct inode * dir,
-			struct dentry *dentry, struct page ** res_page)
+struct ext2_dir_entry_2 *ext2_find_entry (struct inode * dir,
+			struct qstr *child, struct page ** res_page)
 {
-	const char *name = dentry->d_name.name;
-	int namelen = dentry->d_name.len;
+	const char *name = child->name;
+	int namelen = child->len;
 	unsigned reclen = EXT2_DIR_REC_LEN(namelen);
 	unsigned long start, n;
 	unsigned long npages = dir_pages(dir);
@@ -431,13 +431,13 @@ struct ext2_dir_entry_2 * ext2_dotdot (struct inode *dir, struct page **p)
 	return de;
 }
 
-ino_t ext2_inode_by_name(struct inode * dir, struct dentry *dentry)
+ino_t ext2_inode_by_name(struct inode *dir, struct qstr *child)
 {
 	ino_t res = 0;
-	struct ext2_dir_entry_2 * de;
+	struct ext2_dir_entry_2 *de;
 	struct page *page;
 	
-	de = ext2_find_entry (dir, dentry, &page);
+	de = ext2_find_entry (dir, child, &page);
 	if (de) {
 		res = le32_to_cpu(de->inode);
 		ext2_put_page(page);

fs/ext2/ext2.h

@@ -105,9 +105,9 @@ extern void ext2_rsv_window_add(struct super_block *sb, struct ext2_reserve_wind
 
 /* dir.c */
 extern int ext2_add_link (struct dentry *, struct inode *);
-extern ino_t ext2_inode_by_name(struct inode *, struct dentry *);
+extern ino_t ext2_inode_by_name(struct inode *, struct qstr *);
 extern int ext2_make_empty(struct inode *, struct inode *);
-extern struct ext2_dir_entry_2 * ext2_find_entry (struct inode *,struct dentry *, struct page **);
+extern struct ext2_dir_entry_2 * ext2_find_entry (struct inode *,struct qstr *, struct page **);
 extern int ext2_delete_entry (struct ext2_dir_entry_2 *, struct page *);
 extern int ext2_empty_dir (struct inode *);
 extern struct ext2_dir_entry_2 * ext2_dotdot (struct inode *, struct page **);

fs/ext2/namei.c

@@ -60,7 +60,7 @@ static struct dentry *ext2_lookup(struct inode * dir, struct dentry *dentry, str
 	if (dentry->d_name.len > EXT2_NAME_LEN)
 		return ERR_PTR(-ENAMETOOLONG);
 
-	ino = ext2_inode_by_name(dir, dentry);
+	ino = ext2_inode_by_name(dir, &dentry->d_name);
 	inode = NULL;
 	if (ino) {
 		inode = ext2_iget(dir->i_sb, ino);
@@ -72,27 +72,11 @@ static struct dentry *ext2_lookup(struct inode * dir, struct dentry *dentry, str
 
 struct dentry *ext2_get_parent(struct dentry *child)
 {
-	unsigned long ino;
-	struct dentry *parent;
-	struct inode *inode;
-	struct dentry dotdot;
-
-	dotdot.d_name.name = "..";
-	dotdot.d_name.len = 2;
-
-	ino = ext2_inode_by_name(child->d_inode, &dotdot);
+	struct qstr dotdot = {.name = "..", .len = 2};
+	unsigned long ino = ext2_inode_by_name(child->d_inode, &dotdot);
 	if (!ino)
 		return ERR_PTR(-ENOENT);
-	inode = ext2_iget(child->d_inode->i_sb, ino);
-
-	if (IS_ERR(inode))
-		return ERR_CAST(inode);
-	parent = d_alloc_anon(inode);
-	if (!parent) {
-		iput(inode);
-		parent = ERR_PTR(-ENOMEM);
-	}
-	return parent;
+	return d_obtain_alias(ext2_iget(child->d_inode->i_sb, ino));
 } 
 
 /*
@@ -257,7 +241,7 @@ static int ext2_unlink(struct inode * dir, struct dentry *dentry)
 	struct page * page;
 	int err = -ENOENT;
 
-	de = ext2_find_entry (dir, dentry, &page);
+	de = ext2_find_entry (dir, &dentry->d_name, &page);
 	if (!de)
 		goto out;
 
@@ -299,7 +283,7 @@ static int ext2_rename (struct inode * old_dir, struct dentry * old_dentry,
 	struct ext2_dir_entry_2 * old_de;
 	int err = -ENOENT;
 
-	old_de = ext2_find_entry (old_dir, old_dentry, &old_page);
+	old_de = ext2_find_entry (old_dir, &old_dentry->d_name, &old_page);
 	if (!old_de)
 		goto out;
 
@@ -319,7 +303,7 @@ static int ext2_rename (struct inode * old_dir, struct dentry * old_dentry,
 			goto out_dir;
 
 		err = -ENOENT;
-		new_de = ext2_find_entry (new_dir, new_dentry, &new_page);
+		new_de = ext2_find_entry (new_dir, &new_dentry->d_name, &new_page);
 		if (!new_de)
 			goto out_dir;
 		inode_inc_link_count(old_inode);

fs/ext3/namei.c

@@ -159,7 +159,7 @@ static void dx_set_count (struct dx_entry *entries, unsigned value);
 static void dx_set_limit (struct dx_entry *entries, unsigned value);
 static unsigned dx_root_limit (struct inode *dir, unsigned infosize);
 static unsigned dx_node_limit (struct inode *dir);
-static struct dx_frame *dx_probe(struct dentry *dentry,
+static struct dx_frame *dx_probe(struct qstr *entry,
 				 struct inode *dir,
 				 struct dx_hash_info *hinfo,
 				 struct dx_frame *frame,
@@ -176,8 +176,9 @@ static int ext3_htree_next_block(struct inode *dir, __u32 hash,
 				 struct dx_frame *frame,
 				 struct dx_frame *frames,
 				 __u32 *start_hash);
-static struct buffer_head * ext3_dx_find_entry(struct dentry *dentry,
-		       struct ext3_dir_entry_2 **res_dir, int *err);
+static struct buffer_head * ext3_dx_find_entry(struct inode *dir,
+			struct qstr *entry, struct ext3_dir_entry_2 **res_dir,
+			int *err);
 static int ext3_dx_add_entry(handle_t *handle, struct dentry *dentry,
 			     struct inode *inode);
 
@@ -342,7 +343,7 @@ struct stats dx_show_entries(struct dx_hash_info *hinfo, struct inode *dir,
  * back to userspace.
  */
 static struct dx_frame *
-dx_probe(struct dentry *dentry, struct inode *dir,
+dx_probe(struct qstr *entry, struct inode *dir,
 	 struct dx_hash_info *hinfo, struct dx_frame *frame_in, int *err)
 {
 	unsigned count, indirect;
@@ -353,8 +354,6 @@ dx_probe(struct dentry *dentry, struct inode *dir,
 	u32 hash;
 
 	frame->bh = NULL;
-	if (dentry)
-		dir = dentry->d_parent->d_inode;
 	if (!(bh = ext3_bread (NULL,dir, 0, 0, err)))
 		goto fail;
 	root = (struct dx_root *) bh->b_data;
@@ -370,8 +369,8 @@ dx_probe(struct dentry *dentry, struct inode *dir,
 	}
 	hinfo->hash_version = root->info.hash_version;
 	hinfo->seed = EXT3_SB(dir->i_sb)->s_hash_seed;
-	if (dentry)
-		ext3fs_dirhash(dentry->d_name.name, dentry->d_name.len, hinfo);
+	if (entry)
+		ext3fs_dirhash(entry->name, entry->len, hinfo);
 	hash = hinfo->hash;
 
 	if (root->info.unused_flags & 1) {
@@ -803,15 +802,15 @@ static inline int ext3_match (int len, const char * const name,
  */
 static inline int search_dirblock(struct buffer_head * bh,
 				  struct inode *dir,
-				  struct dentry *dentry,
+				  struct qstr *child,
 				  unsigned long offset,
 				  struct ext3_dir_entry_2 ** res_dir)
 {
 	struct ext3_dir_entry_2 * de;
 	char * dlimit;
 	int de_len;
-	const char *name = dentry->d_name.name;
-	int namelen = dentry->d_name.len;
+	const char *name = child->name;
+	int namelen = child->len;
 
 	de = (struct ext3_dir_entry_2 *) bh->b_data;
 	dlimit = bh->b_data + dir->i_sb->s_blocksize;
@@ -850,8 +849,9 @@ static inline int search_dirblock(struct buffer_head * bh,
  * The returned buffer_head has ->b_count elevated.  The caller is expected
  * to brelse() it when appropriate.
  */
-static struct buffer_head * ext3_find_entry (struct dentry *dentry,
-					struct ext3_dir_entry_2 ** res_dir)
+static struct buffer_head *ext3_find_entry(struct inode *dir,
+					struct qstr *entry,
+					struct ext3_dir_entry_2 **res_dir)
 {
 	struct super_block * sb;
 	struct buffer_head * bh_use[NAMEI_RA_SIZE];
@@ -863,16 +863,15 @@ static struct buffer_head * ext3_find_entry (struct dentry *dentry,
 				   buffer */
 	int num = 0;
 	int nblocks, i, err;
-	struct inode *dir = dentry->d_parent->d_inode;
 	int namelen;
 
 	*res_dir = NULL;
 	sb = dir->i_sb;
-	namelen = dentry->d_name.len;
+	namelen = entry->len;
 	if (namelen > EXT3_NAME_LEN)
 		return NULL;
 	if (is_dx(dir)) {
-		bh = ext3_dx_find_entry(dentry, res_dir, &err);
+		bh = ext3_dx_find_entry(dir, entry, res_dir, &err);
 		/*
 		 * On success, or if the error was file not found,
 		 * return.  Otherwise, fall back to doing a search the
@@ -923,7 +922,7 @@ restart:
 			brelse(bh);
 			goto next;
 		}
-		i = search_dirblock(bh, dir, dentry,
+		i = search_dirblock(bh, dir, entry,
 			    block << EXT3_BLOCK_SIZE_BITS(sb), res_dir);
 		if (i == 1) {
 			EXT3_I(dir)->i_dir_start_lookup = block;
@@ -957,8 +956,9 @@ cleanup_and_exit:
 	return ret;
 }
 
-static struct buffer_head * ext3_dx_find_entry(struct dentry *dentry,
-		       struct ext3_dir_entry_2 **res_dir, int *err)
+static struct buffer_head * ext3_dx_find_entry(struct inode *dir,
+			struct qstr *entry, struct ext3_dir_entry_2 **res_dir,
+			int *err)
 {
 	struct super_block * sb;
 	struct dx_hash_info	hinfo;
@@ -968,14 +968,13 @@ static struct buffer_head * ext3_dx_find_entry(struct dentry *dentry,
 	struct buffer_head *bh;
 	unsigned long block;
 	int retval;
-	int namelen = dentry->d_name.len;
-	const u8 *name = dentry->d_name.name;
-	struct inode *dir = dentry->d_parent->d_inode;
+	int namelen = entry->len;
+	const u8 *name = entry->name;
 
 	sb = dir->i_sb;
 	/* NFS may look up ".." - look at dx_root directory block */
-	if (namelen > 2 || name[0] != '.'||(name[1] != '.' && name[1] != '\0')){
-		if (!(frame = dx_probe(dentry, NULL, &hinfo, frames, err)))
+	if (namelen > 2 || name[0] != '.'|| (namelen == 2 && name[1] != '.')) {
+		if (!(frame = dx_probe(entry, dir, &hinfo, frames, err)))
 			return NULL;
 	} else {
 		frame = frames;
@@ -1036,7 +1035,7 @@ static struct dentry *ext3_lookup(struct inode * dir, struct dentry *dentry, str
 	if (dentry->d_name.len > EXT3_NAME_LEN)
 		return ERR_PTR(-ENAMETOOLONG);
 
-	bh = ext3_find_entry(dentry, &de);
+	bh = ext3_find_entry(dir, &dentry->d_name, &de);
 	inode = NULL;
 	if (bh) {
 		unsigned long ino = le32_to_cpu(de->inode);
@@ -1057,18 +1056,11 @@ static struct dentry *ext3_lookup(struct inode * dir, struct dentry *dentry, str
 struct dentry *ext3_get_parent(struct dentry *child)
 {
 	unsigned long ino;
-	struct dentry *parent;
-	struct inode *inode;
-	struct dentry dotdot;
+	struct qstr dotdot = {.name = "..", .len = 2};
 	struct ext3_dir_entry_2 * de;
 	struct buffer_head *bh;
 
-	dotdot.d_name.name = "..";
-	dotdot.d_name.len = 2;
-	dotdot.d_parent = child; /* confusing, isn't it! */
-
-	bh = ext3_find_entry(&dotdot, &de);
-	inode = NULL;
+	bh = ext3_find_entry(child->d_inode, &dotdot, &de);
 	if (!bh)
 		return ERR_PTR(-ENOENT);
 	ino = le32_to_cpu(de->inode);
@@ -1080,16 +1072,7 @@ struct dentry *ext3_get_parent(struct dentry *child)
 		return ERR_PTR(-EIO);
 	}
 
-	inode = ext3_iget(child->d_inode->i_sb, ino);
-	if (IS_ERR(inode))
-		return ERR_CAST(inode);
-
-	parent = d_alloc_anon(inode);
-	if (!parent) {
-		iput(inode);
-		parent = ERR_PTR(-ENOMEM);
-	}
-	return parent;
+	return d_obtain_alias(ext3_iget(child->d_inode->i_sb, ino));
 }
 
 #define S_SHIFT 12
@@ -1503,7 +1486,7 @@ static int ext3_dx_add_entry(handle_t *handle, struct dentry *dentry,
 	struct ext3_dir_entry_2 *de;
 	int err;
 
-	frame = dx_probe(dentry, NULL, &hinfo, frames, &err);
+	frame = dx_probe(&dentry->d_name, dir, &hinfo, frames, &err);
 	if (!frame)
 		return err;
 	entries = frame->entries;
@@ -2056,7 +2039,7 @@ static int ext3_rmdir (struct inode * dir, struct dentry *dentry)
 		return PTR_ERR(handle);
 
 	retval = -ENOENT;
-	bh = ext3_find_entry (dentry, &de);
+	bh = ext3_find_entry(dir, &dentry->d_name, &de);
 	if (!bh)
 		goto end_rmdir;
 
@@ -2118,7 +2101,7 @@ static int ext3_unlink(struct inode * dir, struct dentry *dentry)
 		handle->h_sync = 1;
 
 	retval = -ENOENT;
-	bh = ext3_find_entry (dentry, &de);
+	bh = ext3_find_entry(dir, &dentry->d_name, &de);
 	if (!bh)
 		goto end_unlink;
 
@@ -2276,7 +2259,7 @@ static int ext3_rename (struct inode * old_dir, struct dentry *old_dentry,
 	if (IS_DIRSYNC(old_dir) || IS_DIRSYNC(new_dir))
 		handle->h_sync = 1;
 
-	old_bh = ext3_find_entry (old_dentry, &old_de);
+	old_bh = ext3_find_entry(old_dir, &old_dentry->d_name, &old_de);
 	/*
 	 *  Check for inode number is _not_ due to possible IO errors.
 	 *  We might rmdir the source, keep it as pwd of some process
@@ -2289,7 +2272,7 @@ static int ext3_rename (struct inode * old_dir, struct dentry *old_dentry,
 		goto end_rename;
 
 	new_inode = new_dentry->d_inode;
-	new_bh = ext3_find_entry (new_dentry, &new_de);
+	new_bh = ext3_find_entry(new_dir, &new_dentry->d_name, &new_de);
 	if (new_bh) {
 		if (!new_inode) {
 			brelse (new_bh);
@@ -2355,7 +2338,8 @@ static int ext3_rename (struct inode * old_dir, struct dentry *old_dentry,
 		struct buffer_head *old_bh2;
 		struct ext3_dir_entry_2 *old_de2;
 
-		old_bh2 = ext3_find_entry(old_dentry, &old_de2);
+		old_bh2 = ext3_find_entry(old_dir, &old_dentry->d_name,
+					  &old_de2);
 		if (old_bh2) {
 			retval = ext3_delete_entry(handle, old_dir,
 						   old_de2, old_bh2);

fs/ext3/super.c

@@ -2794,30 +2794,30 @@ static int ext3_quota_on_mount(struct super_block *sb, int type)
  * Standard function to be called on quota_on
  */
 static int ext3_quota_on(struct super_block *sb, int type, int format_id,
-			 char *path, int remount)
+			 char *name, int remount)
 {
 	int err;
-	struct nameidata nd;
+	struct path path;
 
 	if (!test_opt(sb, QUOTA))
 		return -EINVAL;
-	/* When remounting, no checks are needed and in fact, path is NULL */
+	/* When remounting, no checks are needed and in fact, name is NULL */
 	if (remount)
-		return vfs_quota_on(sb, type, format_id, path, remount);
+		return vfs_quota_on(sb, type, format_id, name, remount);
 
-	err = path_lookup(path, LOOKUP_FOLLOW, &nd);
+	err = kern_path(name, LOOKUP_FOLLOW, &path);
 	if (err)
 		return err;
 
 	/* Quotafile not on the same filesystem? */
-	if (nd.path.mnt->mnt_sb != sb) {
-		path_put(&nd.path);
+	if (path.mnt->mnt_sb != sb) {
+		path_put(&path);
 		return -EXDEV;
 	}
 	/* Journaling quota? */
 	if (EXT3_SB(sb)->s_qf_names[type]) {
 		/* Quotafile not of fs root? */
-		if (nd.path.dentry->d_parent->d_inode != sb->s_root->d_inode)
+		if (path.dentry->d_parent != sb->s_root)
 			printk(KERN_WARNING
 				"EXT3-fs: Quota file not on filesystem root. "
 				"Journaled quota will not work.\n");
@@ -2827,7 +2827,7 @@ static int ext3_quota_on(struct super_block *sb, int type, int format_id,
 	 * When we journal data on quota file, we have to flush journal to see
 	 * all updates to the file when we bypass pagecache...
 	 */
-	if (ext3_should_journal_data(nd.path.dentry->d_inode)) {
+	if (ext3_should_journal_data(path.dentry->d_inode)) {
 		/*
 		 * We don't need to lock updates but journal_flush() could
 		 * otherwise be livelocked...
@@ -2841,8 +2841,8 @@ static int ext3_quota_on(struct super_block *sb, int type, int format_id,
 		}
 	}
 
-	err = vfs_quota_on_path(sb, type, format_id, &nd.path);
-	path_put(&nd.path);
+	err = vfs_quota_on_path(sb, type, format_id, &path);
+	path_put(&path);
 	return err;
 }
 

fs/ext4/namei.c

@@ -1083,16 +1083,7 @@ struct dentry *ext4_get_parent(struct dentry *child)
 		return ERR_PTR(-EIO);
 	}
 
-	inode = ext4_iget(child->d_inode->i_sb, ino);
-	if (IS_ERR(inode))
-		return ERR_CAST(inode);
-
-	parent = d_alloc_anon(inode);
-	if (!parent) {
-		iput(inode);
-		parent = ERR_PTR(-ENOMEM);
-	}
-	return parent;
+	return d_obtain_alias(ext4_iget(child->d_inode->i_sb, ino));
 }
 
 #define S_SHIFT 12

fs/ext4/super.c

@@ -3328,30 +3328,30 @@ static int ext4_quota_on_mount(struct super_block *sb, int type)
  * Standard function to be called on quota_on
  */
 static int ext4_quota_on(struct super_block *sb, int type, int format_id,
-			 char *path, int remount)
+			 char *name, int remount)
 {
 	int err;
-	struct nameidata nd;
+	struct path path;
 
 	if (!test_opt(sb, QUOTA))
 		return -EINVAL;
-	/* When remounting, no checks are needed and in fact, path is NULL */
+	/* When remounting, no checks are needed and in fact, name is NULL */
 	if (remount)
-		return vfs_quota_on(sb, type, format_id, path, remount);
+		return vfs_quota_on(sb, type, format_id, name, remount);
 
-	err = path_lookup(path, LOOKUP_FOLLOW, &nd);
+	err = kern_path(name, LOOKUP_FOLLOW, &path);
 	if (err)
 		return err;
 
 	/* Quotafile not on the same filesystem? */
-	if (nd.path.mnt->mnt_sb != sb) {
-		path_put(&nd.path);
+	if (path.mnt->mnt_sb != sb) {
+		path_put(&path);
 		return -EXDEV;
 	}
 	/* Journaling quota? */
 	if (EXT4_SB(sb)->s_qf_names[type]) {
 		/* Quotafile not in fs root? */
-		if (nd.path.dentry->d_parent->d_inode != sb->s_root->d_inode)
+		if (path.dentry->d_parent != sb->s_root)
 			printk(KERN_WARNING
 				"EXT4-fs: Quota file not on filesystem root. "
 				"Journaled quota will not work.\n");
@@ -3361,7 +3361,7 @@ static int ext4_quota_on(struct super_block *sb, int type, int format_id,
 	 * When we journal data on quota file, we have to flush journal to see
 	 * all updates to the file when we bypass pagecache...
 	 */
-	if (ext4_should_journal_data(nd.path.dentry->d_inode)) {
+	if (ext4_should_journal_data(path.dentry->d_inode)) {
 		/*
 		 * We don't need to lock updates but journal_flush() could
 		 * otherwise be livelocked...
@@ -3370,13 +3370,13 @@ static int ext4_quota_on(struct super_block *sb, int type, int format_id,
 		err = jbd2_journal_flush(EXT4_SB(sb)->s_journal);
 		jbd2_journal_unlock_updates(EXT4_SB(sb)->s_journal);
 		if (err) {
-			path_put(&nd.path);
+			path_put(&path);
 			return err;
 		}
 	}
 
-	err = vfs_quota_on_path(sb, type, format_id, &nd.path);
-	path_put(&nd.path);
+	err = vfs_quota_on_path(sb, type, format_id, &path);
+	path_put(&path);
 	return err;
 }
 

fs/fat/dir.c

@@ -839,6 +839,7 @@ const struct file_operations fat_dir_operations = {
 	.compat_ioctl	= fat_compat_dir_ioctl,
 #endif
 	.fsync		= file_fsync,
+	.llseek		= generic_file_llseek,
 };
 
 static int fat_get_short_entry(struct inode *dir, loff_t *pos,

fs/fat/inode.c

@@ -681,33 +681,24 @@ static struct dentry *fat_fh_to_dentry(struct super_block *sb,
 			inode = NULL;
 		}
 	}
-	if (!inode) {
-		/* For now, do nothing
-		 * What we could do is:
-		 * follow the file starting at fh[4], and record
-		 * the ".." entry, and the name of the fh[2] entry.
-		 * The follow the ".." file finding the next step up.
-		 * This way we build a path to the root of
-		 * the tree. If this works, we lookup the path and so
-		 * get this inode into the cache.
-		 * Finally try the fat_iget lookup again
-		 * If that fails, then weare totally out of luck
-		 * But all that is for another day
-		 */
-	}
-	if (!inode)
-		return ERR_PTR(-ESTALE);
-
 
-	/* now to find a dentry.
-	 * If possible, get a well-connected one
+	/*
+	 * For now, do nothing if the inode is not found.
+	 *
+	 * What we could do is:
+	 *
+	 *	- follow the file starting at fh[4], and record the ".." entry,
+	 *	  and the name of the fh[2] entry.
+	 *	- then follow the ".." file finding the next step up.
+	 *
+	 * This way we build a path to the root of the tree. If this works, we
+	 * lookup the path and so get this inode into the cache.  Finally try
+	 * the fat_iget lookup again.  If that fails, then we are totally out
+	 * of luck.  But all that is for another day
 	 */
-	result = d_alloc_anon(inode);
-	if (result == NULL) {
-		iput(inode);
-		return ERR_PTR(-ENOMEM);
-	}
-	result->d_op = sb->s_root->d_op;
+	result = d_obtain_alias(inode);
+	if (!IS_ERR(result))
+		result->d_op = sb->s_root->d_op;
 	return result;
 }
 
@@ -754,15 +745,8 @@ static struct dentry *fat_get_parent(struct dentry *child)
 	}
 	inode = fat_build_inode(sb, de, i_pos);
 	brelse(bh);
-	if (IS_ERR(inode)) {
-		parent = ERR_CAST(inode);
-		goto out;
-	}
-	parent = d_alloc_anon(inode);
-	if (!parent) {
-		iput(inode);
-		parent = ERR_PTR(-ENOMEM);
-	}
+
+	parent = d_obtain_alias(inode);
 out:
 	unlock_super(sb);
 

fs/fuse/inode.c

@@ -596,12 +596,8 @@ static struct dentry *fuse_get_dentry(struct super_block *sb,
 	if (inode->i_generation != handle->generation)
 		goto out_iput;
 
-	entry = d_alloc_anon(inode);
-	err = -ENOMEM;
-	if (!entry)
-		goto out_iput;
-
-	if (get_node_id(inode) != FUSE_ROOT_ID) {
+	entry = d_obtain_alias(inode);
+	if (!IS_ERR(entry) && get_node_id(inode) != FUSE_ROOT_ID) {
 		entry->d_op = &fuse_dentry_operations;
 		fuse_invalidate_entry_cache(entry);
 	}
@@ -696,17 +692,14 @@ static struct dentry *fuse_get_parent(struct dentry *child)
 	name.name = "..";
 	err = fuse_lookup_name(child_inode->i_sb, get_node_id(child_inode),
 			       &name, &outarg, &inode);
-	if (err && err != -ENOENT)
+	if (err) {
+		if (err == -ENOENT)
+			return ERR_PTR(-ESTALE);
 		return ERR_PTR(err);
-	if (err || !inode)
-		return ERR_PTR(-ESTALE);
-
-	parent = d_alloc_anon(inode);
-	if (!parent) {
-		iput(inode);
-		return ERR_PTR(-ENOMEM);
 	}
-	if (get_node_id(inode) != FUSE_ROOT_ID) {
+
+	parent = d_obtain_alias(inode);
+	if (!IS_ERR(parent) && get_node_id(inode) != FUSE_ROOT_ID) {
 		parent->d_op = &fuse_dentry_operations;
 		fuse_invalidate_entry_cache(parent);
 	}

fs/gfs2/ops_export.c

@@ -130,28 +130,17 @@ static int gfs2_get_name(struct dentry *parent, char *name,
 static struct dentry *gfs2_get_parent(struct dentry *child)
 {
 	struct qstr dotdot;
-	struct inode *inode;
 	struct dentry *dentry;
 
-	gfs2_str2qstr(&dotdot, "..");
-	inode = gfs2_lookupi(child->d_inode, &dotdot, 1);
-
-	if (!inode)
-		return ERR_PTR(-ENOENT);
 	/*
-	 * In case of an error, @inode carries the error value, and we
-	 * have to return that as a(n invalid) pointer to dentry.
+	 * XXX(hch): it would be a good idea to keep this around as a
+	 *	     static variable.
 	 */
-	if (IS_ERR(inode))
-		return ERR_CAST(inode);
-
-	dentry = d_alloc_anon(inode);
-	if (!dentry) {
-		iput(inode);
-		return ERR_PTR(-ENOMEM);
-	}
+	gfs2_str2qstr(&dotdot, "..");
 
-	dentry->d_op = &gfs2_dops;
+	dentry = d_obtain_alias(gfs2_lookupi(child->d_inode, &dotdot, 1));
+	if (!IS_ERR(dentry))
+		dentry->d_op = &gfs2_dops;
 	return dentry;
 }
 
@@ -233,13 +222,9 @@ static struct dentry *gfs2_get_dentry(struct super_block *sb,
 	gfs2_glock_dq_uninit(&i_gh);
 
 out_inode:
-	dentry = d_alloc_anon(inode);
-	if (!dentry) {
-		iput(inode);
-		return ERR_PTR(-ENOMEM);
-	}
-
-	dentry->d_op = &gfs2_dops;
+	dentry = d_obtain_alias(inode);
+	if (!IS_ERR(dentry))
+		dentry->d_op = &gfs2_dops;
 	return dentry;
 
 fail_rgd:

fs/gfs2/ops_inode.c

@@ -69,7 +69,7 @@ static int gfs2_create(struct inode *dir, struct dentry *dentry,
 			mark_inode_dirty(inode);
 			break;
 		} else if (PTR_ERR(inode) != -EEXIST ||
-			   (nd && (nd->intent.open.flags & O_EXCL))) {
+			   (nd && nd->flags & LOOKUP_EXCL)) {
 			gfs2_holder_uninit(ghs);
 			return PTR_ERR(inode);
 		}

fs/hfs/inode.c

@@ -511,13 +511,6 @@ void hfs_clear_inode(struct inode *inode)
 	}
 }
 
-static int hfs_permission(struct inode *inode, int mask)
-{
-	if (S_ISREG(inode->i_mode) && mask & MAY_EXEC)
-		return 0;
-	return generic_permission(inode, mask, NULL);
-}
-
 static int hfs_file_open(struct inode *inode, struct file *file)
 {
 	if (HFS_IS_RSRC(inode))
@@ -616,7 +609,6 @@ static const struct inode_operations hfs_file_inode_operations = {
 	.lookup		= hfs_file_lookup,
 	.truncate	= hfs_file_truncate,
 	.setattr	= hfs_inode_setattr,
-	.permission	= hfs_permission,
 	.setxattr	= hfs_setxattr,
 	.getxattr	= hfs_getxattr,
 	.listxattr	= hfs_listxattr,

fs/hfsplus/inode.c

@@ -238,18 +238,6 @@ static void hfsplus_set_perms(struct inode *inode, struct hfsplus_perm *perms)
 	perms->dev = cpu_to_be32(HFSPLUS_I(inode).dev);
 }
 
-static int hfsplus_permission(struct inode *inode, int mask)
-{
-	/* MAY_EXEC is also used for lookup, if no x bit is set allow lookup,
-	 * open_exec has the same test, so it's still not executable, if a x bit
-	 * is set fall back to standard permission check.
-	 */
-	if (S_ISREG(inode->i_mode) && mask & MAY_EXEC && !(inode->i_mode & 0111))
-		return 0;
-	return generic_permission(inode, mask, NULL);
-}
-
-
 static int hfsplus_file_open(struct inode *inode, struct file *file)
 {
 	if (HFSPLUS_IS_RSRC(inode))
@@ -281,7 +269,6 @@ static int hfsplus_file_release(struct inode *inode, struct file *file)
 static const struct inode_operations hfsplus_file_inode_operations = {
 	.lookup		= hfsplus_file_lookup,
 	.truncate	= hfsplus_file_truncate,
-	.permission	= hfsplus_permission,
 	.setxattr	= hfsplus_setxattr,
 	.getxattr	= hfsplus_getxattr,
 	.listxattr	= hfsplus_listxattr,

fs/hpfs/file.c

@@ -143,5 +143,5 @@ const struct file_operations hpfs_file_ops =
 const struct inode_operations hpfs_file_iops =
 {
 	.truncate	= hpfs_truncate,
-	.setattr	= hpfs_notify_change,
+	.setattr	= hpfs_setattr,
 };

fs/hpfs/hpfs_fn.h

@@ -275,7 +275,7 @@ void hpfs_init_inode(struct inode *);
 void hpfs_read_inode(struct inode *);
 void hpfs_write_inode(struct inode *);
 void hpfs_write_inode_nolock(struct inode *);
-int hpfs_notify_change(struct dentry *, struct iattr *);
+int hpfs_setattr(struct dentry *, struct iattr *);
 void hpfs_write_if_changed(struct inode *);
 void hpfs_delete_inode(struct inode *);
 

fs/hpfs/inode.c

@@ -260,19 +260,28 @@ void hpfs_write_inode_nolock(struct inode *i)
 	brelse(bh);
 }
 
-int hpfs_notify_change(struct dentry *dentry, struct iattr *attr)
+int hpfs_setattr(struct dentry *dentry, struct iattr *attr)
 {
 	struct inode *inode = dentry->d_inode;
-	int error=0;
+	int error = -EINVAL;
+
 	lock_kernel();
-	if ( ((attr->ia_valid & ATTR_SIZE) && attr->ia_size > inode->i_size) ||
-	     (hpfs_sb(inode->i_sb)->sb_root == inode->i_ino) ) {
-		error = -EINVAL;
-	} else if ((error = inode_change_ok(inode, attr))) {
-	} else if ((error = inode_setattr(inode, attr))) {
-	} else {
-		hpfs_write_inode(inode);
-	}
+	if (inode->i_ino == hpfs_sb(inode->i_sb)->sb_root)
+		goto out_unlock;
+	if ((attr->ia_valid & ATTR_SIZE) && attr->ia_size > inode->i_size)
+		goto out_unlock;
+
+	error = inode_change_ok(inode, attr);
+	if (error)
+		goto out_unlock;
+
+	error = inode_setattr(inode, attr);
+	if (error)
+		goto out_unlock;
+
+	hpfs_write_inode(inode);
+
+ out_unlock:
 	unlock_kernel();
 	return error;
 }

fs/hpfs/namei.c

@@ -669,5 +669,5 @@ const struct inode_operations hpfs_dir_iops =
 	.rmdir		= hpfs_rmdir,
 	.mknod		= hpfs_mknod,
 	.rename		= hpfs_rename,
-	.setattr	= hpfs_notify_change,
+	.setattr	= hpfs_setattr,
 };

fs/isofs/export.c

@@ -22,7 +22,7 @@ isofs_export_iget(struct super_block *sb,
 		  __u32 generation)
 {
 	struct inode *inode;
-	struct dentry *result;
+
 	if (block == 0)
 		return ERR_PTR(-ESTALE);
 	inode = isofs_iget(sb, block, offset);
@@ -32,12 +32,7 @@ isofs_export_iget(struct super_block *sb,
 		iput(inode);
 		return ERR_PTR(-ESTALE);
 	}
-	result = d_alloc_anon(inode);
-	if (!result) {
-		iput(inode);
-		return ERR_PTR(-ENOMEM);
-	}
-	return result;
+	return d_obtain_alias(inode);
 }
 
 /* This function is surprisingly simple.  The trick is understanding
@@ -51,7 +46,6 @@ static struct dentry *isofs_export_get_parent(struct dentry *child)
 	unsigned long parent_offset = 0;
 	struct inode *child_inode = child->d_inode;
 	struct iso_inode_info *e_child_inode = ISOFS_I(child_inode);
-	struct inode *parent_inode = NULL;
 	struct iso_directory_record *de = NULL;
 	struct buffer_head * bh = NULL;
 	struct dentry *rv = NULL;
@@ -104,28 +98,11 @@ static struct dentry *isofs_export_get_parent(struct dentry *child)
 	/* Normalize */
 	isofs_normalize_block_and_offset(de, &parent_block, &parent_offset);
 
-	/* Get the inode. */
-	parent_inode = isofs_iget(child_inode->i_sb,
-				  parent_block,
-				  parent_offset);
-	if (IS_ERR(parent_inode)) {
-		rv = ERR_CAST(parent_inode);
-		if (rv != ERR_PTR(-ENOMEM))
-			rv = ERR_PTR(-EACCES);
-		goto out;
-	}
-
-	/* Allocate the dentry. */
-	rv = d_alloc_anon(parent_inode);
-	if (rv == NULL) {
-		rv = ERR_PTR(-ENOMEM);
-		goto out;
-	}
-
+	rv = d_obtain_alias(isofs_iget(child_inode->i_sb, parent_block,
+				     parent_offset));
  out:
-	if (bh) {
+	if (bh)
 		brelse(bh);
-	}
 	return rv;
 }
 

fs/jffs2/dir.c

@@ -39,7 +39,8 @@ const struct file_operations jffs2_dir_operations =
 	.read =		generic_read_dir,
 	.readdir =	jffs2_readdir,
 	.unlocked_ioctl=jffs2_ioctl,
-	.fsync =	jffs2_fsync
+	.fsync =	jffs2_fsync,
+	.llseek =	generic_file_llseek,
 };
 
 
@@ -108,9 +109,7 @@ static struct dentry *jffs2_lookup(struct inode *dir_i, struct dentry *target,
 		}
 	}
 
-	d_add(target, inode);
-
-	return NULL;
+	return d_splice_alias(inode, target);
 }
 
 /***********************************************************************/

fs/jffs2/super.c

@@ -22,6 +22,7 @@
 #include <linux/mtd/super.h>
 #include <linux/ctype.h>
 #include <linux/namei.h>
+#include <linux/exportfs.h>
 #include "compr.h"
 #include "nodelist.h"
 
@@ -62,6 +63,52 @@ static int jffs2_sync_fs(struct super_block *sb, int wait)
 	return 0;
 }
 
+static struct inode *jffs2_nfs_get_inode(struct super_block *sb, uint64_t ino,
+					 uint32_t generation)
+{
+	/* We don't care about i_generation. We'll destroy the flash
+	   before we start re-using inode numbers anyway. And even
+	   if that wasn't true, we'd have other problems...*/
+	return jffs2_iget(sb, ino);
+}
+
+static struct dentry *jffs2_fh_to_dentry(struct super_block *sb, struct fid *fid,
+					 int fh_len, int fh_type)
+{
+        return generic_fh_to_dentry(sb, fid, fh_len, fh_type,
+                                    jffs2_nfs_get_inode);
+}
+
+static struct dentry *jffs2_fh_to_parent(struct super_block *sb, struct fid *fid,
+					 int fh_len, int fh_type)
+{
+        return generic_fh_to_parent(sb, fid, fh_len, fh_type,
+                                    jffs2_nfs_get_inode);
+}
+
+static struct dentry *jffs2_get_parent(struct dentry *child)
+{
+	struct jffs2_inode_info *f;
+	uint32_t pino;
+
+	BUG_ON(!S_ISDIR(child->d_inode->i_mode));
+
+	f = JFFS2_INODE_INFO(child->d_inode);
+
+	pino = f->inocache->pino_nlink;
+
+	JFFS2_DEBUG("Parent of directory ino #%u is #%u\n",
+		    f->inocache->ino, pino);
+
+	return d_obtain_alias(jffs2_iget(child->d_inode->i_sb, pino));
+}
+
+static struct export_operations jffs2_export_ops = {
+	.get_parent = jffs2_get_parent,
+	.fh_to_dentry = jffs2_fh_to_dentry,
+	.fh_to_parent = jffs2_fh_to_parent,
+};
+
 static const struct super_operations jffs2_super_operations =
 {
 	.alloc_inode =	jffs2_alloc_inode,
@@ -104,6 +151,7 @@ static int jffs2_fill_super(struct super_block *sb, void *data, int silent)
 	spin_lock_init(&c->inocache_lock);
 
 	sb->s_op = &jffs2_super_operations;
+	sb->s_export_op = &jffs2_export_ops;
 	sb->s_flags = sb->s_flags | MS_NOATIME;
 	sb->s_xattr = jffs2_xattr_handlers;
 #ifdef CONFIG_JFFS2_FS_POSIX_ACL

fs/jfs/namei.c

@@ -1511,25 +1511,12 @@ struct dentry *jfs_fh_to_parent(struct super_block *sb, struct fid *fid,
 
 struct dentry *jfs_get_parent(struct dentry *dentry)
 {
-	struct super_block *sb = dentry->d_inode->i_sb;
-	struct dentry *parent = ERR_PTR(-ENOENT);
-	struct inode *inode;
 	unsigned long parent_ino;
 
 	parent_ino =
 		le32_to_cpu(JFS_IP(dentry->d_inode)->i_dtroot.header.idotdot);
-	inode = jfs_iget(sb, parent_ino);
-	if (IS_ERR(inode)) {
-		parent = ERR_CAST(inode);
-	} else {
-		parent = d_alloc_anon(inode);
-		if (!parent) {
-			parent = ERR_PTR(-ENOMEM);
-			iput(inode);
-		}
-	}
 
-	return parent;
+	return d_obtain_alias(jfs_iget(dentry->d_inode->i_sb, parent_ino));
 }
 
 const struct inode_operations jfs_dir_inode_operations = {
@@ -1560,6 +1547,7 @@ const struct file_operations jfs_dir_operations = {
 #ifdef CONFIG_COMPAT
 	.compat_ioctl	= jfs_compat_ioctl,
 #endif
+	.llseek		= generic_file_llseek,
 };
 
 static int jfs_ci_hash(struct dentry *dir, struct qstr *this)

fs/libfs.c

@@ -732,28 +732,6 @@ out:
 	return ret;
 }
 
-/*
- * This is what d_alloc_anon should have been.  Once the exportfs
- * argument transition has been finished I will update d_alloc_anon
- * to this prototype and this wrapper will go away.   --hch
- */
-static struct dentry *exportfs_d_alloc(struct inode *inode)
-{
-	struct dentry *dentry;
-
-	if (!inode)
-		return NULL;
-	if (IS_ERR(inode))
-		return ERR_PTR(PTR_ERR(inode));
-
-	dentry = d_alloc_anon(inode);
-	if (!dentry) {
-		iput(inode);
-		dentry = ERR_PTR(-ENOMEM);
-	}
-	return dentry;
-}
-
 /**
  * generic_fh_to_dentry - generic helper for the fh_to_dentry export operation
  * @sb:		filesystem to do the file handle conversion on
@@ -782,7 +760,7 @@ struct dentry *generic_fh_to_dentry(struct super_block *sb, struct fid *fid,
 		break;
 	}
 
-	return exportfs_d_alloc(inode);
+	return d_obtain_alias(inode);
 }
 EXPORT_SYMBOL_GPL(generic_fh_to_dentry);
 
@@ -815,7 +793,7 @@ struct dentry *generic_fh_to_parent(struct super_block *sb, struct fid *fid,
 		break;
 	}
 
-	return exportfs_d_alloc(inode);
+	return d_obtain_alias(inode);
 }
 EXPORT_SYMBOL_GPL(generic_fh_to_parent);
 

fs/namei.c

@@ -212,8 +212,7 @@ int generic_permission(struct inode *inode, int mask,
 	 * Read/write DACs are always overridable.
 	 * Executable DACs are overridable if at least one exec bit is set.
 	 */
-	if (!(mask & MAY_EXEC) ||
-	    (inode->i_mode & S_IXUGO) || S_ISDIR(inode->i_mode))
+	if (!(mask & MAY_EXEC) || execute_ok(inode))
 		if (capable(CAP_DAC_OVERRIDE))
 			return 0;
 
@@ -249,23 +248,11 @@ int inode_permission(struct inode *inode, int mask)
 	}
 
 	/* Ordinary permission routines do not understand MAY_APPEND. */
-	if (inode->i_op && inode->i_op->permission) {
+	if (inode->i_op && inode->i_op->permission)
 		retval = inode->i_op->permission(inode, mask);
-		if (!retval) {
-			/*
-			 * Exec permission on a regular file is denied if none
-			 * of the execute bits are set.
-			 *
-			 * This check should be done by the ->permission()
-			 * method.
-			 */
-			if ((mask & MAY_EXEC) && S_ISREG(inode->i_mode) &&
-			    !(inode->i_mode & S_IXUGO))
-				return -EACCES;
-		}
-	} else {
+	else
 		retval = generic_permission(inode, mask, NULL);
-	}
+
 	if (retval)
 		return retval;
 
@@ -1106,6 +1093,15 @@ int path_lookup(const char *name, unsigned int flags,
 	return do_path_lookup(AT_FDCWD, name, flags, nd);
 }
 
+int kern_path(const char *name, unsigned int flags, struct path *path)
+{
+	struct nameidata nd;
+	int res = do_path_lookup(AT_FDCWD, name, flags, &nd);
+	if (!res)
+		*path = nd.path;
+	return res;
+}
+
 /**
  * vfs_path_lookup - lookup a file path relative to a dentry-vfsmount pair
  * @dentry:  pointer to dentry of the base directory
@@ -1138,9 +1134,16 @@ int vfs_path_lookup(struct dentry *dentry, struct vfsmount *mnt,
 
 }
 
-static int __path_lookup_intent_open(int dfd, const char *name,
-		unsigned int lookup_flags, struct nameidata *nd,
-		int open_flags, int create_mode)
+/**
+ * path_lookup_open - lookup a file path with open intent
+ * @dfd: the directory to use as base, or AT_FDCWD
+ * @name: pointer to file name
+ * @lookup_flags: lookup intent flags
+ * @nd: pointer to nameidata
+ * @open_flags: open intent flags
+ */
+int path_lookup_open(int dfd, const char *name, unsigned int lookup_flags,
+		struct nameidata *nd, int open_flags)
 {
 	struct file *filp = get_empty_filp();
 	int err;
@@ -1149,7 +1152,7 @@ static int __path_lookup_intent_open(int dfd, const char *name,
 		return -ENFILE;
 	nd->intent.open.file = filp;
 	nd->intent.open.flags = open_flags;
-	nd->intent.open.create_mode = create_mode;
+	nd->intent.open.create_mode = 0;
 	err = do_path_lookup(dfd, name, lookup_flags|LOOKUP_OPEN, nd);
 	if (IS_ERR(nd->intent.open.file)) {
 		if (err == 0) {
@@ -1161,38 +1164,6 @@ static int __path_lookup_intent_open(int dfd, const char *name,
 	return err;
 }
 
-/**
- * path_lookup_open - lookup a file path with open intent
- * @dfd: the directory to use as base, or AT_FDCWD
- * @name: pointer to file name
- * @lookup_flags: lookup intent flags
- * @nd: pointer to nameidata
- * @open_flags: open intent flags
- */
-int path_lookup_open(int dfd, const char *name, unsigned int lookup_flags,
-		struct nameidata *nd, int open_flags)
-{
-	return __path_lookup_intent_open(dfd, name, lookup_flags, nd,
-			open_flags, 0);
-}
-
-/**
- * path_lookup_create - lookup a file path with open + create intent
- * @dfd: the directory to use as base, or AT_FDCWD
- * @name: pointer to file name
- * @lookup_flags: lookup intent flags
- * @nd: pointer to nameidata
- * @open_flags: open intent flags
- * @create_mode: create intent flags
- */
-static int path_lookup_create(int dfd, const char *name,
-			      unsigned int lookup_flags, struct nameidata *nd,
-			      int open_flags, int create_mode)
-{
-	return __path_lookup_intent_open(dfd, name, lookup_flags|LOOKUP_CREATE,
-			nd, open_flags, create_mode);
-}
-
 static struct dentry *__lookup_hash(struct qstr *name,
 		struct dentry *base, struct nameidata *nd)
 {
@@ -1470,20 +1441,18 @@ struct dentry *lock_rename(struct dentry *p1, struct dentry *p2)
 
 	mutex_lock(&p1->d_inode->i_sb->s_vfs_rename_mutex);
 
-	for (p = p1; p->d_parent != p; p = p->d_parent) {
-		if (p->d_parent == p2) {
-			mutex_lock_nested(&p2->d_inode->i_mutex, I_MUTEX_PARENT);
-			mutex_lock_nested(&p1->d_inode->i_mutex, I_MUTEX_CHILD);
-			return p;
-		}
+	p = d_ancestor(p2, p1);
+	if (p) {
+		mutex_lock_nested(&p2->d_inode->i_mutex, I_MUTEX_PARENT);
+		mutex_lock_nested(&p1->d_inode->i_mutex, I_MUTEX_CHILD);
+		return p;
 	}
 
-	for (p = p2; p->d_parent != p; p = p->d_parent) {
-		if (p->d_parent == p1) {
-			mutex_lock_nested(&p1->d_inode->i_mutex, I_MUTEX_PARENT);
-			mutex_lock_nested(&p2->d_inode->i_mutex, I_MUTEX_CHILD);
-			return p;
-		}
+	p = d_ancestor(p1, p2);
+	if (p) {
+		mutex_lock_nested(&p1->d_inode->i_mutex, I_MUTEX_PARENT);
+		mutex_lock_nested(&p2->d_inode->i_mutex, I_MUTEX_CHILD);
+		return p;
 	}
 
 	mutex_lock_nested(&p1->d_inode->i_mutex, I_MUTEX_PARENT);
@@ -1702,8 +1671,7 @@ struct file *do_filp_open(int dfd, const char *pathname,
 	/*
 	 * Create - we need to know the parent.
 	 */
-	error = path_lookup_create(dfd, pathname, LOOKUP_PARENT,
-				   &nd, flag, mode);
+	error = do_path_lookup(dfd, pathname, LOOKUP_PARENT, &nd);
 	if (error)
 		return ERR_PTR(error);
 
@@ -1714,10 +1682,20 @@ struct file *do_filp_open(int dfd, const char *pathname,
 	 */
 	error = -EISDIR;
 	if (nd.last_type != LAST_NORM || nd.last.name[nd.last.len])
-		goto exit;
+		goto exit_parent;
 
+	error = -ENFILE;
+	filp = get_empty_filp();
+	if (filp == NULL)
+		goto exit_parent;
+	nd.intent.open.file = filp;
+	nd.intent.open.flags = flag;
+	nd.intent.open.create_mode = mode;
 	dir = nd.path.dentry;
 	nd.flags &= ~LOOKUP_PARENT;
+	nd.flags |= LOOKUP_CREATE | LOOKUP_OPEN;
+	if (flag & O_EXCL)
+		nd.flags |= LOOKUP_EXCL;
 	mutex_lock(&dir->d_inode->i_mutex);
 	path.dentry = lookup_hash(&nd);
 	path.mnt = nd.path.mnt;
@@ -1822,6 +1800,7 @@ exit_dput:
 exit:
 	if (!IS_ERR(nd.intent.open.file))
 		release_open_intent(&nd);
+exit_parent:
 	path_put(&nd.path);
 	return ERR_PTR(error);
 
@@ -1914,7 +1893,7 @@ struct dentry *lookup_create(struct nameidata *nd, int is_dir)
 	if (nd->last_type != LAST_NORM)
 		goto fail;
 	nd->flags &= ~LOOKUP_PARENT;
-	nd->flags |= LOOKUP_CREATE;
+	nd->flags |= LOOKUP_CREATE | LOOKUP_EXCL;
 	nd->intent.open.flags = O_EXCL;
 
 	/*
@@ -2178,16 +2157,19 @@ static long do_rmdir(int dfd, const char __user *pathname)
 		return error;
 
 	switch(nd.last_type) {
-		case LAST_DOTDOT:
-			error = -ENOTEMPTY;
-			goto exit1;
-		case LAST_DOT:
-			error = -EINVAL;
-			goto exit1;
-		case LAST_ROOT:
-			error = -EBUSY;
-			goto exit1;
+	case LAST_DOTDOT:
+		error = -ENOTEMPTY;
+		goto exit1;
+	case LAST_DOT:
+		error = -EINVAL;
+		goto exit1;
+	case LAST_ROOT:
+		error = -EBUSY;
+		goto exit1;
 	}
+
+	nd.flags &= ~LOOKUP_PARENT;
+
 	mutex_lock_nested(&nd.path.dentry->d_inode->i_mutex, I_MUTEX_PARENT);
 	dentry = lookup_hash(&nd);
 	error = PTR_ERR(dentry);
@@ -2265,6 +2247,9 @@ static long do_unlinkat(int dfd, const char __user *pathname)
 	error = -EISDIR;
 	if (nd.last_type != LAST_NORM)
 		goto exit1;
+
+	nd.flags &= ~LOOKUP_PARENT;
+
 	mutex_lock_nested(&nd.path.dentry->d_inode->i_mutex, I_MUTEX_PARENT);
 	dentry = lookup_hash(&nd);
 	error = PTR_ERR(dentry);
@@ -2654,6 +2639,10 @@ asmlinkage long sys_renameat(int olddfd, const char __user *oldname,
 	if (newnd.last_type != LAST_NORM)
 		goto exit2;
 
+	oldnd.flags &= ~LOOKUP_PARENT;
+	newnd.flags &= ~LOOKUP_PARENT;
+	newnd.flags |= LOOKUP_RENAME_TARGET;
+
 	trap = lock_rename(new_dir, old_dir);
 
 	old_dentry = lookup_hash(&oldnd);
@@ -2855,6 +2844,7 @@ EXPORT_SYMBOL(__page_symlink);
 EXPORT_SYMBOL(page_symlink);
 EXPORT_SYMBOL(page_symlink_inode_operations);
 EXPORT_SYMBOL(path_lookup);
+EXPORT_SYMBOL(kern_path);
 EXPORT_SYMBOL(vfs_path_lookup);
 EXPORT_SYMBOL(inode_permission);
 EXPORT_SYMBOL(vfs_permission);

fs/namespace.c

@@ -1167,19 +1167,19 @@ asmlinkage long sys_oldumount(char __user * name)
 
 #endif
 
-static int mount_is_safe(struct nameidata *nd)
+static int mount_is_safe(struct path *path)
 {
 	if (capable(CAP_SYS_ADMIN))
 		return 0;
 	return -EPERM;
 #ifdef notyet
-	if (S_ISLNK(nd->path.dentry->d_inode->i_mode))
+	if (S_ISLNK(path->dentry->d_inode->i_mode))
 		return -EPERM;
-	if (nd->path.dentry->d_inode->i_mode & S_ISVTX) {
-		if (current->uid != nd->path.dentry->d_inode->i_uid)
+	if (path->dentry->d_inode->i_mode & S_ISVTX) {
+		if (current->uid != path->dentry->d_inode->i_uid)
 			return -EPERM;
 	}
-	if (vfs_permission(nd, MAY_WRITE))
+	if (inode_permission(path->dentry->d_inode, MAY_WRITE))
 		return -EPERM;
 	return 0;
 #endif
@@ -1425,11 +1425,10 @@ out_unlock:
 
 /*
  * recursively change the type of the mountpoint.
- * noinline this do_mount helper to save do_mount stack space.
  */
-static noinline int do_change_type(struct nameidata *nd, int flag)
+static int do_change_type(struct path *path, int flag)
 {
-	struct vfsmount *m, *mnt = nd->path.mnt;
+	struct vfsmount *m, *mnt = path->mnt;
 	int recurse = flag & MS_REC;
 	int type = flag & ~MS_REC;
 	int err = 0;
@@ -1437,7 +1436,7 @@ static noinline int do_change_type(struct nameidata *nd, int flag)
 	if (!capable(CAP_SYS_ADMIN))
 		return -EPERM;
 
-	if (nd->path.dentry != nd->path.mnt->mnt_root)
+	if (path->dentry != path->mnt->mnt_root)
 		return -EINVAL;
 
 	down_write(&namespace_sem);
@@ -1459,40 +1458,39 @@ static noinline int do_change_type(struct nameidata *nd, int flag)
 
 /*
  * do loopback mount.
- * noinline this do_mount helper to save do_mount stack space.
  */
-static noinline int do_loopback(struct nameidata *nd, char *old_name,
+static int do_loopback(struct path *path, char *old_name,
 				int recurse)
 {
-	struct nameidata old_nd;
+	struct path old_path;
 	struct vfsmount *mnt = NULL;
-	int err = mount_is_safe(nd);
+	int err = mount_is_safe(path);
 	if (err)
 		return err;
 	if (!old_name || !*old_name)
 		return -EINVAL;
-	err = path_lookup(old_name, LOOKUP_FOLLOW, &old_nd);
+	err = kern_path(old_name, LOOKUP_FOLLOW, &old_path);
 	if (err)
 		return err;
 
 	down_write(&namespace_sem);
 	err = -EINVAL;
-	if (IS_MNT_UNBINDABLE(old_nd.path.mnt))
+	if (IS_MNT_UNBINDABLE(old_path.mnt))
 		goto out;
 
-	if (!check_mnt(nd->path.mnt) || !check_mnt(old_nd.path.mnt))
+	if (!check_mnt(path->mnt) || !check_mnt(old_path.mnt))
 		goto out;
 
 	err = -ENOMEM;
 	if (recurse)
-		mnt = copy_tree(old_nd.path.mnt, old_nd.path.dentry, 0);
+		mnt = copy_tree(old_path.mnt, old_path.dentry, 0);
 	else
-		mnt = clone_mnt(old_nd.path.mnt, old_nd.path.dentry, 0);
+		mnt = clone_mnt(old_path.mnt, old_path.dentry, 0);
 
 	if (!mnt)
 		goto out;
 
-	err = graft_tree(mnt, &nd->path);
+	err = graft_tree(mnt, path);
 	if (err) {
 		LIST_HEAD(umount_list);
 		spin_lock(&vfsmount_lock);
@@ -1503,7 +1501,7 @@ static noinline int do_loopback(struct nameidata *nd, char *old_name,
 
 out:
 	up_write(&namespace_sem);
-	path_put(&old_nd.path);
+	path_put(&old_path);
 	return err;
 }
 
@@ -1528,33 +1526,37 @@ static int change_mount_flags(struct vfsmount *mnt, int ms_flags)
  * change filesystem flags. dir should be a physical root of filesystem.
  * If you've mounted a non-root directory somewhere and want to do remount
  * on it - tough luck.
- * noinline this do_mount helper to save do_mount stack space.
  */
-static noinline int do_remount(struct nameidata *nd, int flags, int mnt_flags,
+static int do_remount(struct path *path, int flags, int mnt_flags,
 		      void *data)
 {
 	int err;
-	struct super_block *sb = nd->path.mnt->mnt_sb;
+	struct super_block *sb = path->mnt->mnt_sb;
 
 	if (!capable(CAP_SYS_ADMIN))
 		return -EPERM;
 
-	if (!check_mnt(nd->path.mnt))
+	if (!check_mnt(path->mnt))
 		return -EINVAL;
 
-	if (nd->path.dentry != nd->path.mnt->mnt_root)
+	if (path->dentry != path->mnt->mnt_root)
 		return -EINVAL;
 
 	down_write(&sb->s_umount);
 	if (flags & MS_BIND)
-		err = change_mount_flags(nd->path.mnt, flags);
+		err = change_mount_flags(path->mnt, flags);
 	else
 		err = do_remount_sb(sb, flags, data, 0);
 	if (!err)
-		nd->path.mnt->mnt_flags = mnt_flags;
+		path->mnt->mnt_flags = mnt_flags;
 	up_write(&sb->s_umount);
-	if (!err)
-		security_sb_post_remount(nd->path.mnt, flags, data);
+	if (!err) {
+		security_sb_post_remount(path->mnt, flags, data);
+
+		spin_lock(&vfsmount_lock);
+		touch_mnt_namespace(path->mnt->mnt_ns);
+		spin_unlock(&vfsmount_lock);
+	}
 	return err;
 }
 
@@ -1568,90 +1570,85 @@ static inline int tree_contains_unbindable(struct vfsmount *mnt)
 	return 0;
 }
 
-/*
- * noinline this do_mount helper to save do_mount stack space.
- */
-static noinline int do_move_mount(struct nameidata *nd, char *old_name)
+static int do_move_mount(struct path *path, char *old_name)
 {
-	struct nameidata old_nd;
-	struct path parent_path;
+	struct path old_path, parent_path;
 	struct vfsmount *p;
 	int err = 0;
 	if (!capable(CAP_SYS_ADMIN))
 		return -EPERM;
 	if (!old_name || !*old_name)
 		return -EINVAL;
-	err = path_lookup(old_name, LOOKUP_FOLLOW, &old_nd);
+	err = kern_path(old_name, LOOKUP_FOLLOW, &old_path);
 	if (err)
 		return err;
 
 	down_write(&namespace_sem);
-	while (d_mountpoint(nd->path.dentry) &&
-	       follow_down(&nd->path.mnt, &nd->path.dentry))
+	while (d_mountpoint(path->dentry) &&
+	       follow_down(&path->mnt, &path->dentry))
 		;
 	err = -EINVAL;
-	if (!check_mnt(nd->path.mnt) || !check_mnt(old_nd.path.mnt))
+	if (!check_mnt(path->mnt) || !check_mnt(old_path.mnt))
 		goto out;
 
 	err = -ENOENT;
-	mutex_lock(&nd->path.dentry->d_inode->i_mutex);
-	if (IS_DEADDIR(nd->path.dentry->d_inode))
+	mutex_lock(&path->dentry->d_inode->i_mutex);
+	if (IS_DEADDIR(path->dentry->d_inode))
 		goto out1;
 
-	if (!IS_ROOT(nd->path.dentry) && d_unhashed(nd->path.dentry))
+	if (!IS_ROOT(path->dentry) && d_unhashed(path->dentry))
 		goto out1;
 
 	err = -EINVAL;
-	if (old_nd.path.dentry != old_nd.path.mnt->mnt_root)
+	if (old_path.dentry != old_path.mnt->mnt_root)
 		goto out1;
 
-	if (old_nd.path.mnt == old_nd.path.mnt->mnt_parent)
+	if (old_path.mnt == old_path.mnt->mnt_parent)
 		goto out1;
 
-	if (S_ISDIR(nd->path.dentry->d_inode->i_mode) !=
-	      S_ISDIR(old_nd.path.dentry->d_inode->i_mode))
+	if (S_ISDIR(path->dentry->d_inode->i_mode) !=
+	      S_ISDIR(old_path.dentry->d_inode->i_mode))
 		goto out1;
 	/*
 	 * Don't move a mount residing in a shared parent.
 	 */
-	if (old_nd.path.mnt->mnt_parent &&
-	    IS_MNT_SHARED(old_nd.path.mnt->mnt_parent))
+	if (old_path.mnt->mnt_parent &&
+	    IS_MNT_SHARED(old_path.mnt->mnt_parent))
 		goto out1;
 	/*
 	 * Don't move a mount tree containing unbindable mounts to a destination
 	 * mount which is shared.
 	 */
-	if (IS_MNT_SHARED(nd->path.mnt) &&
-	    tree_contains_unbindable(old_nd.path.mnt))
+	if (IS_MNT_SHARED(path->mnt) &&
+	    tree_contains_unbindable(old_path.mnt))
 		goto out1;
 	err = -ELOOP;
-	for (p = nd->path.mnt; p->mnt_parent != p; p = p->mnt_parent)
-		if (p == old_nd.path.mnt)
+	for (p = path->mnt; p->mnt_parent != p; p = p->mnt_parent)
+		if (p == old_path.mnt)
 			goto out1;
 
-	err = attach_recursive_mnt(old_nd.path.mnt, &nd->path, &parent_path);
+	err = attach_recursive_mnt(old_path.mnt, path, &parent_path);
 	if (err)
 		goto out1;
 
 	/* if the mount is moved, it should no longer be expire
 	 * automatically */
-	list_del_init(&old_nd.path.mnt->mnt_expire);
+	list_del_init(&old_path.mnt->mnt_expire);
 out1:
-	mutex_unlock(&nd->path.dentry->d_inode->i_mutex);
+	mutex_unlock(&path->dentry->d_inode->i_mutex);
 out:
 	up_write(&namespace_sem);
 	if (!err)
 		path_put(&parent_path);
-	path_put(&old_nd.path);
+	path_put(&old_path);
 	return err;
 }
 
 /*
  * create a new mount for userspace and request it to be added into the
  * namespace's tree
- * noinline this do_mount helper to save do_mount stack space.
  */
-static noinline int do_new_mount(struct nameidata *nd, char *type, int flags,
+static int do_new_mount(struct path *path, char *type, int flags,
 			int mnt_flags, char *name, void *data)
 {
 	struct vfsmount *mnt;
@@ -1667,7 +1664,7 @@ static noinline int do_new_mount(struct nameidata *nd, char *type, int flags,
 	if (IS_ERR(mnt))
 		return PTR_ERR(mnt);
 
-	return do_add_mount(mnt, &nd->path, mnt_flags, NULL);
+	return do_add_mount(mnt, path, mnt_flags, NULL);
 }
 
 /*
@@ -1902,7 +1899,7 @@ int copy_mount_options(const void __user * data, unsigned long *where)
 long do_mount(char *dev_name, char *dir_name, char *type_page,
 		  unsigned long flags, void *data_page)
 {
-	struct nameidata nd;
+	struct path path;
 	int retval = 0;
 	int mnt_flags = 0;
 
@@ -1940,29 +1937,29 @@ long do_mount(char *dev_name, char *dir_name, char *type_page,
 		   MS_NOATIME | MS_NODIRATIME | MS_RELATIME| MS_KERNMOUNT);
 
 	/* ... and get the mountpoint */
-	retval = path_lookup(dir_name, LOOKUP_FOLLOW, &nd);
+	retval = kern_path(dir_name, LOOKUP_FOLLOW, &path);
 	if (retval)
 		return retval;
 
-	retval = security_sb_mount(dev_name, &nd.path,
+	retval = security_sb_mount(dev_name, &path,
 				   type_page, flags, data_page);
 	if (retval)
 		goto dput_out;
 
 	if (flags & MS_REMOUNT)
-		retval = do_remount(&nd, flags & ~MS_REMOUNT, mnt_flags,
+		retval = do_remount(&path, flags & ~MS_REMOUNT, mnt_flags,
 				    data_page);
 	else if (flags & MS_BIND)
-		retval = do_loopback(&nd, dev_name, flags & MS_REC);
+		retval = do_loopback(&path, dev_name, flags & MS_REC);
 	else if (flags & (MS_SHARED | MS_PRIVATE | MS_SLAVE | MS_UNBINDABLE))
-		retval = do_change_type(&nd, flags);
+		retval = do_change_type(&path, flags);
 	else if (flags & MS_MOVE)
-		retval = do_move_mount(&nd, dev_name);
+		retval = do_move_mount(&path, dev_name);
 	else
-		retval = do_new_mount(&nd, type_page, flags, mnt_flags,
+		retval = do_new_mount(&path, type_page, flags, mnt_flags,
 				      dev_name, data_page);
 dput_out:
-	path_put(&nd.path);
+	path_put(&path);
 	return retval;
 }
 

fs/nfs/dir.c

@@ -707,9 +707,7 @@ static int nfs_is_exclusive_create(struct inode *dir, struct nameidata *nd)
 {
 	if (NFS_PROTO(dir)->version == 2)
 		return 0;
-	if (nd == NULL || nfs_lookup_check_intent(nd, LOOKUP_CREATE) == 0)
-		return 0;
-	return (nd->intent.open.flags & O_EXCL) != 0;
+	return nd && nfs_lookup_check_intent(nd, LOOKUP_EXCL);
 }
 
 /*
@@ -1009,7 +1007,7 @@ static struct dentry *nfs_atomic_lookup(struct inode *dir, struct dentry *dentry
 
 	/* Let vfs_create() deal with O_EXCL. Instantiate, but don't hash
 	 * the dentry. */
-	if (nd->intent.open.flags & O_EXCL) {
+	if (nd->flags & LOOKUP_EXCL) {
 		d_instantiate(dentry, NULL);
 		goto out;
 	}
@@ -1959,6 +1957,9 @@ force_lookup:
 	} else
 		res = PTR_ERR(cred);
 out:
+	if (!res && (mask & MAY_EXEC) && !execute_ok(inode))
+		res = -EACCES;
+
 	dfprintk(VFS, "NFS: permission(%s/%ld), mask=0x%x, res=%d\n",
 		inode->i_sb->s_id, inode->i_ino, mask, res);
 	return res;

fs/nfs/getroot.c

@@ -107,11 +107,10 @@ struct dentry *nfs_get_root(struct super_block *sb, struct nfs_fh *mntfh)
 	 * if the dentry tree reaches them; however if the dentry already
 	 * exists, we'll pick it up at this point and use it as the root
 	 */
-	mntroot = d_alloc_anon(inode);
-	if (!mntroot) {
-		iput(inode);
+	mntroot = d_obtain_alias(inode);
+	if (IS_ERR(mntroot)) {
 		dprintk("nfs_get_root: get root dentry failed\n");
-		return ERR_PTR(-ENOMEM);
+		return mntroot;
 	}
 
 	security_d_instantiate(mntroot, inode);
@@ -277,11 +276,10 @@ struct dentry *nfs4_get_root(struct super_block *sb, struct nfs_fh *mntfh)
 	 * if the dentry tree reaches them; however if the dentry already
 	 * exists, we'll pick it up at this point and use it as the root
 	 */
-	mntroot = d_alloc_anon(inode);
-	if (!mntroot) {
-		iput(inode);
+	mntroot = d_obtain_alias(inode);
+	if (IS_ERR(mntroot)) {
 		dprintk("nfs_get_root: get root dentry failed\n");
-		return ERR_PTR(-ENOMEM);
+		return mntroot;
 	}
 
 	security_d_instantiate(mntroot, inode);

fs/nfsd/export.c

@@ -163,18 +163,16 @@ static int expkey_parse(struct cache_detail *cd, char *mesg, int mlen)
 		if (!ek)
 			err = -ENOMEM;
 	} else {
-		struct nameidata nd;
-		err = path_lookup(buf, 0, &nd);
+		err = kern_path(buf, 0, &key.ek_path);
 		if (err)
 			goto out;
 
 		dprintk("Found the path %s\n", buf);
-		key.ek_path = nd.path;
 
 		ek = svc_expkey_update(&key, ek);
 		if (!ek)
 			err = -ENOMEM;
-		path_put(&nd.path);
+		path_put(&key.ek_path);
 	}
 	cache_flush();
  out:
@@ -501,35 +499,22 @@ static int svc_export_parse(struct cache_detail *cd, char *mesg, int mlen)
 	int len;
 	int err;
 	struct auth_domain *dom = NULL;
-	struct nameidata nd;
-	struct svc_export exp, *expp;
+	struct svc_export exp = {}, *expp;
 	int an_int;
 
-	nd.path.dentry = NULL;
-	exp.ex_pathname = NULL;
-
-	/* fs locations */
-	exp.ex_fslocs.locations = NULL;
-	exp.ex_fslocs.locations_count = 0;
-	exp.ex_fslocs.migrated = 0;
-
-	exp.ex_uuid = NULL;
-
-	/* secinfo */
-	exp.ex_nflavors = 0;
-
 	if (mesg[mlen-1] != '\n')
 		return -EINVAL;
 	mesg[mlen-1] = 0;
 
 	buf = kmalloc(PAGE_SIZE, GFP_KERNEL);
-	err = -ENOMEM;
-	if (!buf) goto out;
+	if (!buf)
+		return -ENOMEM;
 
 	/* client */
-	len = qword_get(&mesg, buf, PAGE_SIZE);
 	err = -EINVAL;
-	if (len <= 0) goto out;
+	len = qword_get(&mesg, buf, PAGE_SIZE);
+	if (len <= 0)
+		goto out;
 
 	err = -ENOENT;
 	dom = auth_domain_find(buf);
@@ -538,25 +523,25 @@ static int svc_export_parse(struct cache_detail *cd, char *mesg, int mlen)
 
 	/* path */
 	err = -EINVAL;
-	if ((len=qword_get(&mesg, buf, PAGE_SIZE)) <= 0)
-		goto out;
-	err = path_lookup(buf, 0, &nd);
-	if (err) goto out_no_path;
+	if ((len = qword_get(&mesg, buf, PAGE_SIZE)) <= 0)
+		goto out1;
+
+	err = kern_path(buf, 0, &exp.ex_path);
+	if (err)
+		goto out1;
 
-	exp.h.flags = 0;
 	exp.ex_client = dom;
-	exp.ex_path.mnt = nd.path.mnt;
-	exp.ex_path.dentry = nd.path.dentry;
-	exp.ex_pathname = kstrdup(buf, GFP_KERNEL);
+
 	err = -ENOMEM;
+	exp.ex_pathname = kstrdup(buf, GFP_KERNEL);
 	if (!exp.ex_pathname)
-		goto out;
+		goto out2;
 
 	/* expiry */
 	err = -EINVAL;
 	exp.h.expiry_time = get_expiry(&mesg);
 	if (exp.h.expiry_time == 0)
-		goto out;
+		goto out3;
 
 	/* flags */
 	err = get_int(&mesg, &an_int);
@@ -564,22 +549,26 @@ static int svc_export_parse(struct cache_detail *cd, char *mesg, int mlen)
 		err = 0;
 		set_bit(CACHE_NEGATIVE, &exp.h.flags);
 	} else {
-		if (err || an_int < 0) goto out;	
+		if (err || an_int < 0)
+			goto out3;
 		exp.ex_flags= an_int;
 	
 		/* anon uid */
 		err = get_int(&mesg, &an_int);
-		if (err) goto out;
+		if (err)
+			goto out3;
 		exp.ex_anon_uid= an_int;
 
 		/* anon gid */
 		err = get_int(&mesg, &an_int);
-		if (err) goto out;
+		if (err)
+			goto out3;
 		exp.ex_anon_gid= an_int;
 
 		/* fsid */
 		err = get_int(&mesg, &an_int);
-		if (err) goto out;
+		if (err)
+			goto out3;
 		exp.ex_fsid = an_int;
 
 		while ((len = qword_get(&mesg, buf, PAGE_SIZE)) > 0) {
@@ -605,12 +594,13 @@ static int svc_export_parse(struct cache_detail *cd, char *mesg, int mlen)
 				 */
 				break;
 			if (err)
-				goto out;
+				goto out4;
 		}
 
-		err = check_export(nd.path.dentry->d_inode, exp.ex_flags,
+		err = check_export(exp.ex_path.dentry->d_inode, exp.ex_flags,
 				   exp.ex_uuid);
-		if (err) goto out;
+		if (err)
+			goto out4;
 	}
 
 	expp = svc_export_lookup(&exp);
@@ -623,15 +613,16 @@ static int svc_export_parse(struct cache_detail *cd, char *mesg, int mlen)
 		err = -ENOMEM;
 	else
 		exp_put(expp);
- out:
+out4:
 	nfsd4_fslocs_free(&exp.ex_fslocs);
 	kfree(exp.ex_uuid);
+out3:
 	kfree(exp.ex_pathname);
-	if (nd.path.dentry)
-		path_put(&nd.path);
- out_no_path:
-	if (dom)
-		auth_domain_put(dom);
+out2:
+	path_put(&exp.ex_path);
+out1:
+	auth_domain_put(dom);
+out:
 	kfree(buf);
 	return err;
 }
@@ -999,7 +990,7 @@ exp_export(struct nfsctl_export *nxp)
 	struct svc_export	*exp = NULL;
 	struct svc_export	new;
 	struct svc_expkey	*fsid_key = NULL;
-	struct nameidata nd;
+	struct path path;
 	int		err;
 
 	/* Consistency check */
@@ -1022,12 +1013,12 @@ exp_export(struct nfsctl_export *nxp)
 
 
 	/* Look up the dentry */
-	err = path_lookup(nxp->ex_path, 0, &nd);
+	err = kern_path(nxp->ex_path, 0, &path);
 	if (err)
 		goto out_put_clp;
 	err = -EINVAL;
 
-	exp = exp_get_by_name(clp, nd.path.mnt, nd.path.dentry, NULL);
+	exp = exp_get_by_name(clp, path.mnt, path.dentry, NULL);
 
 	memset(&new, 0, sizeof(new));
 
@@ -1035,8 +1026,8 @@ exp_export(struct nfsctl_export *nxp)
 	if ((nxp->ex_flags & NFSEXP_FSID) &&
 	    (!IS_ERR(fsid_key = exp_get_fsid_key(clp, nxp->ex_dev))) &&
 	    fsid_key->ek_path.mnt &&
-	    (fsid_key->ek_path.mnt != nd.path.mnt ||
-	     fsid_key->ek_path.dentry != nd.path.dentry))
+	    (fsid_key->ek_path.mnt != path.mnt ||
+	     fsid_key->ek_path.dentry != path.dentry))
 		goto finish;
 
 	if (!IS_ERR(exp)) {
@@ -1052,7 +1043,7 @@ exp_export(struct nfsctl_export *nxp)
 		goto finish;
 	}
 
-	err = check_export(nd.path.dentry->d_inode, nxp->ex_flags, NULL);
+	err = check_export(path.dentry->d_inode, nxp->ex_flags, NULL);
 	if (err) goto finish;
 
 	err = -ENOMEM;
@@ -1065,7 +1056,7 @@ exp_export(struct nfsctl_export *nxp)
 	if (!new.ex_pathname)
 		goto finish;
 	new.ex_client = clp;
-	new.ex_path = nd.path;
+	new.ex_path = path;
 	new.ex_flags = nxp->ex_flags;
 	new.ex_anon_uid = nxp->ex_anon_uid;
 	new.ex_anon_gid = nxp->ex_anon_gid;
@@ -1091,7 +1082,7 @@ finish:
 		exp_put(exp);
 	if (fsid_key && !IS_ERR(fsid_key))
 		cache_put(&fsid_key->h, &svc_expkey_cache);
-	path_put(&nd.path);
+	path_put(&path);
 out_put_clp:
 	auth_domain_put(clp);
 out_unlock:
@@ -1122,7 +1113,7 @@ exp_unexport(struct nfsctl_export *nxp)
 {
 	struct auth_domain *dom;
 	svc_export *exp;
-	struct nameidata nd;
+	struct path path;
 	int		err;
 
 	/* Consistency check */
@@ -1139,13 +1130,13 @@ exp_unexport(struct nfsctl_export *nxp)
 		goto out_unlock;
 	}
 
-	err = path_lookup(nxp->ex_path, 0, &nd);
+	err = kern_path(nxp->ex_path, 0, &path);
 	if (err)
 		goto out_domain;
 
 	err = -EINVAL;
-	exp = exp_get_by_name(dom, nd.path.mnt, nd.path.dentry, NULL);
-	path_put(&nd.path);
+	exp = exp_get_by_name(dom, path.mnt, path.dentry, NULL);
+	path_put(&path);
 	if (IS_ERR(exp))
 		goto out_domain;
 
@@ -1167,26 +1158,26 @@ out_unlock:
  * since its harder to fool a kernel module than a user space program.
  */
 int
-exp_rootfh(svc_client *clp, char *path, struct knfsd_fh *f, int maxsize)
+exp_rootfh(svc_client *clp, char *name, struct knfsd_fh *f, int maxsize)
 {
 	struct svc_export	*exp;
-	struct nameidata	nd;
+	struct path		path;
 	struct inode		*inode;
 	struct svc_fh		fh;
 	int			err;
 
 	err = -EPERM;
 	/* NB: we probably ought to check that it's NUL-terminated */
-	if (path_lookup(path, 0, &nd)) {
-		printk("nfsd: exp_rootfh path not found %s", path);
+	if (kern_path(name, 0, &path)) {
+		printk("nfsd: exp_rootfh path not found %s", name);
 		return err;
 	}
-	inode = nd.path.dentry->d_inode;
+	inode = path.dentry->d_inode;
 
 	dprintk("nfsd: exp_rootfh(%s [%p] %s:%s/%ld)\n",
-		 path, nd.path.dentry, clp->name,
+		 name, path.dentry, clp->name,
 		 inode->i_sb->s_id, inode->i_ino);
-	exp = exp_parent(clp, nd.path.mnt, nd.path.dentry, NULL);
+	exp = exp_parent(clp, path.mnt, path.dentry, NULL);
 	if (IS_ERR(exp)) {
 		err = PTR_ERR(exp);
 		goto out;
@@ -1196,7 +1187,7 @@ exp_rootfh(svc_client *clp, char *path, struct knfsd_fh *f, int maxsize)
 	 * fh must be initialized before calling fh_compose
 	 */
 	fh_init(&fh, maxsize);
-	if (fh_compose(&fh, exp, nd.path.dentry, NULL))
+	if (fh_compose(&fh, exp, path.dentry, NULL))
 		err = -EINVAL;
 	else
 		err = 0;
@@ -1204,7 +1195,7 @@ exp_rootfh(svc_client *clp, char *path, struct knfsd_fh *f, int maxsize)
 	fh_put(&fh);
 	exp_put(exp);
 out:
-	path_put(&nd.path);
+	path_put(&path);
 	return err;
 }
 

fs/nfsd/nfs4recover.c

@@ -51,7 +51,7 @@
 #define NFSDDBG_FACILITY                NFSDDBG_PROC
 
 /* Globals */
-static struct nameidata rec_dir;
+static struct path rec_dir;
 static int rec_dir_init = 0;
 
 static void
@@ -121,9 +121,9 @@ out_no_tfm:
 static void
 nfsd4_sync_rec_dir(void)
 {
-	mutex_lock(&rec_dir.path.dentry->d_inode->i_mutex);
-	nfsd_sync_dir(rec_dir.path.dentry);
-	mutex_unlock(&rec_dir.path.dentry->d_inode->i_mutex);
+	mutex_lock(&rec_dir.dentry->d_inode->i_mutex);
+	nfsd_sync_dir(rec_dir.dentry);
+	mutex_unlock(&rec_dir.dentry->d_inode->i_mutex);
 }
 
 int
@@ -143,9 +143,9 @@ nfsd4_create_clid_dir(struct nfs4_client *clp)
 	nfs4_save_user(&uid, &gid);
 
 	/* lock the parent */
-	mutex_lock(&rec_dir.path.dentry->d_inode->i_mutex);
+	mutex_lock(&rec_dir.dentry->d_inode->i_mutex);
 
-	dentry = lookup_one_len(dname, rec_dir.path.dentry, HEXDIR_LEN-1);
+	dentry = lookup_one_len(dname, rec_dir.dentry, HEXDIR_LEN-1);
 	if (IS_ERR(dentry)) {
 		status = PTR_ERR(dentry);
 		goto out_unlock;
@@ -155,15 +155,15 @@ nfsd4_create_clid_dir(struct nfs4_client *clp)
 		dprintk("NFSD: nfsd4_create_clid_dir: DIRECTORY EXISTS\n");
 		goto out_put;
 	}
-	status = mnt_want_write(rec_dir.path.mnt);
+	status = mnt_want_write(rec_dir.mnt);
 	if (status)
 		goto out_put;
-	status = vfs_mkdir(rec_dir.path.dentry->d_inode, dentry, S_IRWXU);
-	mnt_drop_write(rec_dir.path.mnt);
+	status = vfs_mkdir(rec_dir.dentry->d_inode, dentry, S_IRWXU);
+	mnt_drop_write(rec_dir.mnt);
 out_put:
 	dput(dentry);
 out_unlock:
-	mutex_unlock(&rec_dir.path.dentry->d_inode->i_mutex);
+	mutex_unlock(&rec_dir.dentry->d_inode->i_mutex);
 	if (status == 0) {
 		clp->cl_firststate = 1;
 		nfsd4_sync_rec_dir();
@@ -226,7 +226,7 @@ nfsd4_list_rec_dir(struct dentry *dir, recdir_func *f)
 
 	nfs4_save_user(&uid, &gid);
 
-	filp = dentry_open(dget(dir), mntget(rec_dir.path.mnt), O_RDONLY);
+	filp = dentry_open(dget(dir), mntget(rec_dir.mnt), O_RDONLY);
 	status = PTR_ERR(filp);
 	if (IS_ERR(filp))
 		goto out;
@@ -291,9 +291,9 @@ nfsd4_unlink_clid_dir(char *name, int namlen)
 
 	dprintk("NFSD: nfsd4_unlink_clid_dir. name %.*s\n", namlen, name);
 
-	mutex_lock(&rec_dir.path.dentry->d_inode->i_mutex);
-	dentry = lookup_one_len(name, rec_dir.path.dentry, namlen);
-	mutex_unlock(&rec_dir.path.dentry->d_inode->i_mutex);
+	mutex_lock(&rec_dir.dentry->d_inode->i_mutex);
+	dentry = lookup_one_len(name, rec_dir.dentry, namlen);
+	mutex_unlock(&rec_dir.dentry->d_inode->i_mutex);
 	if (IS_ERR(dentry)) {
 		status = PTR_ERR(dentry);
 		return status;
@@ -302,7 +302,7 @@ nfsd4_unlink_clid_dir(char *name, int namlen)
 	if (!dentry->d_inode)
 		goto out;
 
-	status = nfsd4_clear_clid_dir(rec_dir.path.dentry, dentry);
+	status = nfsd4_clear_clid_dir(rec_dir.dentry, dentry);
 out:
 	dput(dentry);
 	return status;
@@ -318,7 +318,7 @@ nfsd4_remove_clid_dir(struct nfs4_client *clp)
 	if (!rec_dir_init || !clp->cl_firststate)
 		return;
 
-	status = mnt_want_write(rec_dir.path.mnt);
+	status = mnt_want_write(rec_dir.mnt);
 	if (status)
 		goto out;
 	clp->cl_firststate = 0;
@@ -327,7 +327,7 @@ nfsd4_remove_clid_dir(struct nfs4_client *clp)
 	nfs4_reset_user(uid, gid);
 	if (status == 0)
 		nfsd4_sync_rec_dir();
-	mnt_drop_write(rec_dir.path.mnt);
+	mnt_drop_write(rec_dir.mnt);
 out:
 	if (status)
 		printk("NFSD: Failed to remove expired client state directory"
@@ -357,17 +357,17 @@ nfsd4_recdir_purge_old(void) {
 
 	if (!rec_dir_init)
 		return;
-	status = mnt_want_write(rec_dir.path.mnt);
+	status = mnt_want_write(rec_dir.mnt);
 	if (status)
 		goto out;
-	status = nfsd4_list_rec_dir(rec_dir.path.dentry, purge_old);
+	status = nfsd4_list_rec_dir(rec_dir.dentry, purge_old);
 	if (status == 0)
 		nfsd4_sync_rec_dir();
-	mnt_drop_write(rec_dir.path.mnt);
+	mnt_drop_write(rec_dir.mnt);
 out:
 	if (status)
 		printk("nfsd4: failed to purge old clients from recovery"
-			" directory %s\n", rec_dir.path.dentry->d_name.name);
+			" directory %s\n", rec_dir.dentry->d_name.name);
 }
 
 static int
@@ -387,10 +387,10 @@ int
 nfsd4_recdir_load(void) {
 	int status;
 
-	status = nfsd4_list_rec_dir(rec_dir.path.dentry, load_recdir);
+	status = nfsd4_list_rec_dir(rec_dir.dentry, load_recdir);
 	if (status)
 		printk("nfsd4: failed loading clients from recovery"
-			" directory %s\n", rec_dir.path.dentry->d_name.name);
+			" directory %s\n", rec_dir.dentry->d_name.name);
 	return status;
 }
 
@@ -412,7 +412,7 @@ nfsd4_init_recdir(char *rec_dirname)
 
 	nfs4_save_user(&uid, &gid);
 
-	status = path_lookup(rec_dirname, LOOKUP_FOLLOW | LOOKUP_DIRECTORY,
+	status = kern_path(rec_dirname, LOOKUP_FOLLOW | LOOKUP_DIRECTORY,
 			&rec_dir);
 	if (status)
 		printk("NFSD: unable to find recovery directory %s\n",
@@ -429,5 +429,5 @@ nfsd4_shutdown_recdir(void)
 	if (!rec_dir_init)
 		return;
 	rec_dir_init = 0;
-	path_put(&rec_dir.path);
+	path_put(&rec_dir);
 }

fs/nfsd/nfs4state.c

@@ -3284,17 +3284,17 @@ int
 nfs4_reset_recoverydir(char *recdir)
 {
 	int status;
-	struct nameidata nd;
+	struct path path;
 
-	status = path_lookup(recdir, LOOKUP_FOLLOW, &nd);
+	status = kern_path(recdir, LOOKUP_FOLLOW, &path);
 	if (status)
 		return status;
 	status = -ENOTDIR;
-	if (S_ISDIR(nd.path.dentry->d_inode->i_mode)) {
+	if (S_ISDIR(path.dentry->d_inode->i_mode)) {
 		nfs4_set_recdir(recdir);
 		status = 0;
 	}
-	path_put(&nd.path);
+	path_put(&path);
 	return status;
 }
 

fs/nfsd/nfsctl.c

@@ -341,7 +341,7 @@ static ssize_t failover_unlock_ip(struct file *file, char *buf, size_t size)
 
 static ssize_t failover_unlock_fs(struct file *file, char *buf, size_t size)
 {
-	struct nameidata nd;
+	struct path path;
 	char *fo_path;
 	int error;
 
@@ -356,13 +356,13 @@ static ssize_t failover_unlock_fs(struct file *file, char *buf, size_t size)
 	if (qword_get(&buf, fo_path, size) < 0)
 		return -EINVAL;
 
-	error = path_lookup(fo_path, 0, &nd);
+	error = kern_path(fo_path, 0, &path);
 	if (error)
 		return error;
 
-	error = nlmsvc_unlock_all_by_sb(nd.path.mnt->mnt_sb);
+	error = nlmsvc_unlock_all_by_sb(path.mnt->mnt_sb);
 
-	path_put(&nd.path);
+	path_put(&path);
 	return error;
 }
 

fs/nfsd/vfs.c

@@ -1817,6 +1817,115 @@ out:
 	return err;
 }
 
+/*
+ * We do this buffering because we must not call back into the file
+ * system's ->lookup() method from the filldir callback. That may well
+ * deadlock a number of file systems.
+ *
+ * This is based heavily on the implementation of same in XFS.
+ */
+struct buffered_dirent {
+	u64		ino;
+	loff_t		offset;
+	int		namlen;
+	unsigned int	d_type;
+	char		name[];
+};
+
+struct readdir_data {
+	char		*dirent;
+	size_t		used;
+	int		full;
+};
+
+static int nfsd_buffered_filldir(void *__buf, const char *name, int namlen,
+				 loff_t offset, u64 ino, unsigned int d_type)
+{
+	struct readdir_data *buf = __buf;
+	struct buffered_dirent *de = (void *)(buf->dirent + buf->used);
+	unsigned int reclen;
+
+	reclen = ALIGN(sizeof(struct buffered_dirent) + namlen, sizeof(u64));
+	if (buf->used + reclen > PAGE_SIZE) {
+		buf->full = 1;
+		return -EINVAL;
+	}
+
+	de->namlen = namlen;
+	de->offset = offset;
+	de->ino = ino;
+	de->d_type = d_type;
+	memcpy(de->name, name, namlen);
+	buf->used += reclen;
+
+	return 0;
+}
+
+static int nfsd_buffered_readdir(struct file *file, filldir_t func,
+				 struct readdir_cd *cdp, loff_t *offsetp)
+{
+	struct readdir_data buf;
+	struct buffered_dirent *de;
+	int host_err;
+	int size;
+	loff_t offset;
+
+	buf.dirent = (void *)__get_free_page(GFP_KERNEL);
+	if (!buf.dirent)
+		return -ENOMEM;
+
+	offset = *offsetp;
+	cdp->err = nfserr_eof; /* will be cleared on successful read */
+
+	while (1) {
+		unsigned int reclen;
+
+		buf.used = 0;
+		buf.full = 0;
+
+		host_err = vfs_readdir(file, nfsd_buffered_filldir, &buf);
+		if (buf.full)
+			host_err = 0;
+
+		if (host_err < 0)
+			break;
+
+		size = buf.used;
+
+		if (!size)
+			break;
+
+		de = (struct buffered_dirent *)buf.dirent;
+		while (size > 0) {
+			offset = de->offset;
+
+			if (func(cdp, de->name, de->namlen, de->offset,
+				 de->ino, de->d_type))
+				goto done;
+
+			if (cdp->err != nfs_ok)
+				goto done;
+
+			reclen = ALIGN(sizeof(*de) + de->namlen,
+				       sizeof(u64));
+			size -= reclen;
+			de = (struct buffered_dirent *)((char *)de + reclen);
+		}
+		offset = vfs_llseek(file, 0, SEEK_CUR);
+		if (!buf.full)
+			break;
+	}
+
+ done:
+	free_page((unsigned long)(buf.dirent));
+
+	if (host_err)
+		return nfserrno(host_err);
+
+	*offsetp = offset;
+	return cdp->err;
+}
+
 /*
  * Read entries from a directory.
  * The  NFSv3/4 verifier we ignore for now.
@@ -1826,7 +1935,6 @@ nfsd_readdir(struct svc_rqst *rqstp, struct svc_fh *fhp, loff_t *offsetp,
 	     struct readdir_cd *cdp, filldir_t func)
 {
 	__be32		err;
-	int 		host_err;
 	struct file	*file;
 	loff_t		offset = *offsetp;
 
@@ -1840,21 +1948,7 @@ nfsd_readdir(struct svc_rqst *rqstp, struct svc_fh *fhp, loff_t *offsetp,
 		goto out_close;
 	}
 
-	/*
-	 * Read the directory entries. This silly loop is necessary because
-	 * readdir() is not guaranteed to fill up the entire buffer, but
-	 * may choose to do less.
-	 */
-
-	do {
-		cdp->err = nfserr_eof; /* will be cleared on successful read */
-		host_err = vfs_readdir(file, func, cdp);
-	} while (host_err >=0 && cdp->err == nfs_ok);
-	if (host_err)
-		err = nfserrno(host_err);
-	else
-		err = cdp->err;
-	*offsetp = vfs_llseek(file, 0, 1);
+	err = nfsd_buffered_readdir(file, func, cdp, offsetp);
 
 	if (err == nfserr_eof || err == nfserr_toosmall)
 		err = nfs_ok; /* can still be found in ->err */

fs/ntfs/namei.c

@@ -304,8 +304,6 @@ static struct dentry *ntfs_get_parent(struct dentry *child_dent)
 	ntfs_attr_search_ctx *ctx;
 	ATTR_RECORD *attr;
 	FILE_NAME_ATTR *fn;
-	struct inode *parent_vi;
-	struct dentry *parent_dent;
 	unsigned long parent_ino;
 	int err;
 
@@ -345,24 +343,8 @@ try_next:
 	/* Release the search context and the mft record of the child. */
 	ntfs_attr_put_search_ctx(ctx);
 	unmap_mft_record(ni);
-	/* Get the inode of the parent directory. */
-	parent_vi = ntfs_iget(vi->i_sb, parent_ino);
-	if (IS_ERR(parent_vi) || unlikely(is_bad_inode(parent_vi))) {
-		if (!IS_ERR(parent_vi))
-			iput(parent_vi);
-		ntfs_error(vi->i_sb, "Failed to get parent directory inode "
-				"0x%lx of child inode 0x%lx.", parent_ino,
-				vi->i_ino);
-		return ERR_PTR(-EACCES);
-	}
-	/* Finally get a dentry for the parent directory and return it. */
-	parent_dent = d_alloc_anon(parent_vi);
-	if (unlikely(!parent_dent)) {
-		iput(parent_vi);
-		return ERR_PTR(-ENOMEM);
-	}
-	ntfs_debug("Done for inode 0x%lx.", vi->i_ino);
-	return parent_dent;
+
+	return d_obtain_alias(ntfs_iget(vi->i_sb, parent_ino));
 }
 
 static struct inode *ntfs_nfs_get_inode(struct super_block *sb,

fs/ocfs2/export.c

@@ -68,14 +68,9 @@ static struct dentry *ocfs2_get_dentry(struct super_block *sb,
 		return ERR_PTR(-ESTALE);
 	}
 
-	result = d_alloc_anon(inode);
-
-	if (!result) {
-		iput(inode);
-		mlog_errno(-ENOMEM);
-		return ERR_PTR(-ENOMEM);
-	}
-	result->d_op = &ocfs2_dentry_ops;
+	result = d_obtain_alias(inode);
+	if (!IS_ERR(result))
+		result->d_op = &ocfs2_dentry_ops;
 
 	mlog_exit_ptr(result);
 	return result;
@@ -86,7 +81,6 @@ static struct dentry *ocfs2_get_parent(struct dentry *child)
 	int status;
 	u64 blkno;
 	struct dentry *parent;
-	struct inode *inode;
 	struct inode *dir = child->d_inode;
 
 	mlog_entry("(0x%p, '%.*s')\n", child,
@@ -109,21 +103,9 @@ static struct dentry *ocfs2_get_parent(struct dentry *child)
 		goto bail_unlock;
 	}
 
-	inode = ocfs2_iget(OCFS2_SB(dir->i_sb), blkno, 0, 0);
-	if (IS_ERR(inode)) {
-		mlog(ML_ERROR, "Unable to create inode %llu\n",
-		     (unsigned long long)blkno);
-		parent = ERR_PTR(-EACCES);
-		goto bail_unlock;
-	}
-
-	parent = d_alloc_anon(inode);
-	if (!parent) {
-		iput(inode);
-		parent = ERR_PTR(-ENOMEM);
-	}
-
-	parent->d_op = &ocfs2_dentry_ops;
+	parent = d_obtain_alias(ocfs2_iget(OCFS2_SB(dir->i_sb), blkno, 0, 0));
+	if (!IS_ERR(parent))
+		parent->d_op = &ocfs2_dentry_ops;
 
 bail_unlock:
 	ocfs2_inode_unlock(dir, 0);

fs/omfs/dir.c

@@ -501,4 +501,5 @@ struct inode_operations omfs_dir_inops = {
 struct file_operations omfs_dir_operations = {
 	.read = generic_read_dir,
 	.readdir = omfs_readdir,
+	.llseek = generic_file_llseek,
 };

fs/openpromfs/inode.c

@@ -167,6 +167,7 @@ static int openpromfs_readdir(struct file *, void *, filldir_t);
 static const struct file_operations openprom_operations = {
 	.read		= generic_read_dir,
 	.readdir	= openpromfs_readdir,
+	.llseek		= generic_file_llseek,
 };
 
 static struct dentry *openpromfs_lookup(struct inode *, struct dentry *, struct nameidata *);

fs/proc/proc_sysctl.c

@@ -298,13 +298,19 @@ static int proc_sys_permission(struct inode *inode, int mask)
 	 * sysctl entries that are not writeable,
 	 * are _NOT_ writeable, capabilities or not.
 	 */
-	struct ctl_table_header *head = grab_header(inode);
-	struct ctl_table *table = PROC_I(inode)->sysctl_entry;
+	struct ctl_table_header *head;
+	struct ctl_table *table;
 	int error;
 
+	/* Executable files are not allowed under /proc/sys/ */
+	if ((mask & MAY_EXEC) && S_ISREG(inode->i_mode))
+		return -EACCES;
+
+	head = grab_header(inode);
 	if (IS_ERR(head))
 		return PTR_ERR(head);
 
+	table = PROC_I(inode)->sysctl_entry;
 	if (!table) /* global root - r-xr-xr-x */
 		error = mask & MAY_WRITE ? -EACCES : 0;
 	else /* Use the permissions on the sysctl table entry */
@@ -353,6 +359,7 @@ static const struct file_operations proc_sys_file_operations = {
 
 static const struct file_operations proc_sys_dir_file_operations = {
 	.readdir	= proc_sys_readdir,
+	.llseek		= generic_file_llseek,
 };
 
 static const struct inode_operations proc_sys_inode_operations = {

fs/read_write.c

@@ -31,39 +31,61 @@ const struct file_operations generic_ro_fops = {
 
 EXPORT_SYMBOL(generic_ro_fops);
 
+/**
+ * generic_file_llseek_unlocked - lockless generic llseek implementation
+ * @file:	file structure to seek on
+ * @offset:	file offset to seek to
+ * @origin:	type of seek
+ *
+ * Updates the file offset to the value specified by @offset and @origin.
+ * Locking must be provided by the caller.
+ */
 loff_t
 generic_file_llseek_unlocked(struct file *file, loff_t offset, int origin)
 {
-	loff_t retval;
 	struct inode *inode = file->f_mapping->host;
 
 	switch (origin) {
-		case SEEK_END:
-			offset += inode->i_size;
-			break;
-		case SEEK_CUR:
-			offset += file->f_pos;
+	case SEEK_END:
+		offset += inode->i_size;
+		break;
+	case SEEK_CUR:
+		offset += file->f_pos;
+		break;
 	}
-	retval = -EINVAL;
-	if (offset>=0 && offset<=inode->i_sb->s_maxbytes) {
-		/* Special lock needed here? */
-		if (offset != file->f_pos) {
-			file->f_pos = offset;
-			file->f_version = 0;
-		}
-		retval = offset;
+
+	if (offset < 0 || offset > inode->i_sb->s_maxbytes)
+		return -EINVAL;
+
+	/* Special lock needed here? */
+	if (offset != file->f_pos) {
+		file->f_pos = offset;
+		file->f_version = 0;
 	}
-	return retval;
+
+	return offset;
 }
 EXPORT_SYMBOL(generic_file_llseek_unlocked);
 
+/**
+ * generic_file_llseek - generic llseek implementation for regular files
+ * @file:	file structure to seek on
+ * @offset:	file offset to seek to
+ * @origin:	type of seek
+ *
+ * This is a generic implemenation of ->llseek useable for all normal local
+ * filesystems.  It just updates the file offset to the value specified by
+ * @offset and @origin under i_mutex.
+ */
 loff_t generic_file_llseek(struct file *file, loff_t offset, int origin)
 {
-	loff_t n;
+	loff_t rval;
+
 	mutex_lock(&file->f_dentry->d_inode->i_mutex);
-	n = generic_file_llseek_unlocked(file, offset, origin);
+	rval = generic_file_llseek_unlocked(file, offset, origin);
 	mutex_unlock(&file->f_dentry->d_inode->i_mutex);
-	return n;
+
+	return rval;
 }
 EXPORT_SYMBOL(generic_file_llseek);
 

fs/readdir.c

@@ -117,7 +117,7 @@ asmlinkage long old_readdir(unsigned int fd, struct old_linux_dirent __user * di
 	buf.dirent = dirent;
 
 	error = vfs_readdir(file, fillonedir, &buf);
-	if (error >= 0)
+	if (buf.result)
 		error = buf.result;
 
 	fput(file);
@@ -209,9 +209,8 @@ asmlinkage long sys_getdents(unsigned int fd, struct linux_dirent __user * diren
 	buf.error = 0;
 
 	error = vfs_readdir(file, filldir, &buf);
-	if (error < 0)
-		goto out_putf;
-	error = buf.error;
+	if (error >= 0)
+		error = buf.error;
 	lastdirent = buf.previous;
 	if (lastdirent) {
 		if (put_user(file->f_pos, &lastdirent->d_off))
@@ -219,8 +218,6 @@ asmlinkage long sys_getdents(unsigned int fd, struct linux_dirent __user * diren
 		else
 			error = count - buf.count;
 	}
-
-out_putf:
 	fput(file);
 out:
 	return error;
@@ -293,19 +290,16 @@ asmlinkage long sys_getdents64(unsigned int fd, struct linux_dirent64 __user * d
 	buf.error = 0;
 
 	error = vfs_readdir(file, filldir64, &buf);
-	if (error < 0)
-		goto out_putf;
-	error = buf.error;
+	if (error >= 0)
+		error = buf.error;
 	lastdirent = buf.previous;
 	if (lastdirent) {
 		typeof(lastdirent->d_off) d_off = file->f_pos;
-		error = -EFAULT;
 		if (__put_user(d_off, &lastdirent->d_off))
-			goto out_putf;
-		error = count - buf.count;
+			error = -EFAULT;
+		else
+			error = count - buf.count;
 	}
-
-out_putf:
 	fput(file);
 out:
 	return error;

fs/reiserfs/file.c

@@ -296,6 +296,7 @@ const struct file_operations reiserfs_file_operations = {
 	.aio_write = generic_file_aio_write,
 	.splice_read = generic_file_splice_read,
 	.splice_write = generic_file_splice_write,
+	.llseek = generic_file_llseek,
 };
 
 const struct inode_operations reiserfs_file_inode_operations = {

fs/reiserfs/inode.c

@@ -1522,7 +1522,6 @@ static struct dentry *reiserfs_get_dentry(struct super_block *sb,
 
 {
 	struct cpu_key key;
-	struct dentry *result;
 	struct inode *inode;
 
 	key.on_disk_key.k_objectid = objectid;
@@ -1535,16 +1534,8 @@ static struct dentry *reiserfs_get_dentry(struct super_block *sb,
 		inode = NULL;
 	}
 	reiserfs_write_unlock(sb);
-	if (!inode)
-		inode = ERR_PTR(-ESTALE);
-	if (IS_ERR(inode))
-		return ERR_CAST(inode);
-	result = d_alloc_anon(inode);
-	if (!result) {
-		iput(inode);
-		return ERR_PTR(-ENOMEM);
-	}
-	return result;
+
+	return d_obtain_alias(inode);
 }
 
 struct dentry *reiserfs_fh_to_dentry(struct super_block *sb, struct fid *fid,

fs/reiserfs/namei.c

@@ -383,7 +383,6 @@ struct dentry *reiserfs_get_parent(struct dentry *child)
 	struct inode *inode = NULL;
 	struct reiserfs_dir_entry de;
 	INITIALIZE_PATH(path_to_entry);
-	struct dentry *parent;
 	struct inode *dir = child->d_inode;
 
 	if (dir->i_nlink == 0) {
@@ -401,15 +400,7 @@ struct dentry *reiserfs_get_parent(struct dentry *child)
 	inode = reiserfs_iget(dir->i_sb, (struct cpu_key *)&(de.de_dir_id));
 	reiserfs_write_unlock(dir->i_sb);
 
-	if (!inode || IS_ERR(inode)) {
-		return ERR_PTR(-EACCES);
-	}
-	parent = d_alloc_anon(inode);
-	if (!parent) {
-		iput(inode);
-		parent = ERR_PTR(-ENOMEM);
-	}
-	return parent;
+	return d_obtain_alias(inode);
 }
 
 /* add entry to the directory (entry can be hidden). 

fs/reiserfs/super.c

@@ -2058,10 +2058,10 @@ static int reiserfs_quota_on_mount(struct super_block *sb, int type)
  * Standard function to be called on quota_on
  */
 static int reiserfs_quota_on(struct super_block *sb, int type, int format_id,
-			     char *path, int remount)
+			     char *name, int remount)
 {
 	int err;
-	struct nameidata nd;
+	struct path path;
 	struct inode *inode;
 	struct reiserfs_transaction_handle th;
 
@@ -2069,16 +2069,16 @@ static int reiserfs_quota_on(struct super_block *sb, int type, int format_id,
 		return -EINVAL;
 	/* No more checks needed? Path and format_id are bogus anyway... */
 	if (remount)
-		return vfs_quota_on(sb, type, format_id, path, 1);
-	err = path_lookup(path, LOOKUP_FOLLOW, &nd);
+		return vfs_quota_on(sb, type, format_id, name, 1);
+	err = kern_path(name, LOOKUP_FOLLOW, &path);
 	if (err)
 		return err;
 	/* Quotafile not on the same filesystem? */
-	if (nd.path.mnt->mnt_sb != sb) {
+	if (path.mnt->mnt_sb != sb) {
 		err = -EXDEV;
 		goto out;
 	}
-	inode = nd.path.dentry->d_inode;
+	inode = path.dentry->d_inode;
 	/* We must not pack tails for quota files on reiserfs for quota IO to work */
 	if (!(REISERFS_I(inode)->i_flags & i_nopack_mask)) {
 		err = reiserfs_unpack(inode, NULL);
@@ -2094,7 +2094,7 @@ static int reiserfs_quota_on(struct super_block *sb, int type, int format_id,
 	/* Journaling quota? */
 	if (REISERFS_SB(sb)->s_qf_names[type]) {
 		/* Quotafile not of fs root? */
-		if (nd.path.dentry->d_parent->d_inode != sb->s_root->d_inode)
+		if (path.dentry->d_parent != sb->s_root)
 			reiserfs_warning(sb,
 				 "reiserfs: Quota file not on filesystem root. "
 				 "Journalled quota will not work.");
@@ -2113,9 +2113,9 @@ static int reiserfs_quota_on(struct super_block *sb, int type, int format_id,
 		if (err)
 			goto out;
 	}
-	err = vfs_quota_on_path(sb, type, format_id, &nd.path);
+	err = vfs_quota_on_path(sb, type, format_id, &path);
 out:
-	path_put(&nd.path);
+	path_put(&path);
 	return err;
 }
 

fs/super.c

@@ -682,7 +682,7 @@ void emergency_remount(void)
  * filesystems which don't use real block-devices.  -- jrs
  */
 
-static struct idr unnamed_dev_idr;
+static DEFINE_IDA(unnamed_dev_ida);
 static DEFINE_SPINLOCK(unnamed_dev_lock);/* protects the above */
 
 int set_anon_super(struct super_block *s, void *data)
@@ -691,10 +691,10 @@ int set_anon_super(struct super_block *s, void *data)
 	int error;
 
  retry:
-	if (idr_pre_get(&unnamed_dev_idr, GFP_ATOMIC) == 0)
+	if (ida_pre_get(&unnamed_dev_ida, GFP_ATOMIC) == 0)
 		return -ENOMEM;
 	spin_lock(&unnamed_dev_lock);
-	error = idr_get_new(&unnamed_dev_idr, NULL, &dev);
+	error = ida_get_new(&unnamed_dev_ida, &dev);
 	spin_unlock(&unnamed_dev_lock);
 	if (error == -EAGAIN)
 		/* We raced and lost with another CPU. */
@@ -704,7 +704,7 @@ int set_anon_super(struct super_block *s, void *data)
 
 	if ((dev & MAX_ID_MASK) == (1 << MINORBITS)) {
 		spin_lock(&unnamed_dev_lock);
-		idr_remove(&unnamed_dev_idr, dev);
+		ida_remove(&unnamed_dev_ida, dev);
 		spin_unlock(&unnamed_dev_lock);
 		return -EMFILE;
 	}
@@ -720,17 +720,12 @@ void kill_anon_super(struct super_block *sb)
 
 	generic_shutdown_super(sb);
 	spin_lock(&unnamed_dev_lock);
-	idr_remove(&unnamed_dev_idr, slot);
+	ida_remove(&unnamed_dev_ida, slot);
 	spin_unlock(&unnamed_dev_lock);
 }
 
 EXPORT_SYMBOL(kill_anon_super);
 
-void __init unnamed_dev_init(void)
-{
-	idr_init(&unnamed_dev_idr);
-}
-
 void kill_litter_super(struct super_block *sb)
 {
 	if (sb->s_root)

fs/sysfs/dir.c

@@ -983,4 +983,5 @@ static int sysfs_readdir(struct file * filp, void * dirent, filldir_t filldir)
 const struct file_operations sysfs_dir_operations = {
 	.read		= generic_read_dir,
 	.readdir	= sysfs_readdir,
+	.llseek		= generic_file_llseek,
 };

fs/udf/namei.c

@@ -142,7 +142,7 @@ int udf_write_fi(struct inode *inode, struct fileIdentDesc *cfi,
 }
 
 static struct fileIdentDesc *udf_find_entry(struct inode *dir,
-					    struct dentry *dentry,
+					    struct qstr *child,
 					    struct udf_fileident_bh *fibh,
 					    struct fileIdentDesc *cfi)
 {
@@ -159,8 +159,8 @@ static struct fileIdentDesc *udf_find_entry(struct inode *dir,
 	sector_t offset;
 	struct extent_position epos = {};
 	struct udf_inode_info *dinfo = UDF_I(dir);
-	int isdotdot = dentry->d_name.len == 2 &&
-		dentry->d_name.name[0] == '.' && dentry->d_name.name[1] == '.';
+	int isdotdot = child->len == 2 &&
+		child->name[0] == '.' && child->name[1] == '.';
 
 	size = udf_ext0_offset(dir) + dir->i_size;
 	f_pos = udf_ext0_offset(dir);
@@ -238,8 +238,7 @@ static struct fileIdentDesc *udf_find_entry(struct inode *dir,
 			continue;
 
 		flen = udf_get_filename(dir->i_sb, nameptr, fname, lfi);
-		if (flen && udf_match(flen, fname, dentry->d_name.len,
-				      dentry->d_name.name))
+		if (flen && udf_match(flen, fname, child->len, child->name))
 			goto out_ok;
 	}
 
@@ -283,7 +282,7 @@ static struct dentry *udf_lookup(struct inode *dir, struct dentry *dentry,
 	} else
 #endif /* UDF_RECOVERY */
 
-	if (udf_find_entry(dir, dentry, &fibh, &cfi)) {
+	if (udf_find_entry(dir, &dentry->d_name, &fibh, &cfi)) {
 		if (fibh.sbh != fibh.ebh)
 			brelse(fibh.ebh);
 		brelse(fibh.sbh);
@@ -783,7 +782,7 @@ static int udf_rmdir(struct inode *dir, struct dentry *dentry)
 
 	retval = -ENOENT;
 	lock_kernel();
-	fi = udf_find_entry(dir, dentry, &fibh, &cfi);
+	fi = udf_find_entry(dir, &dentry->d_name, &fibh, &cfi);
 	if (!fi)
 		goto out;
 
@@ -829,7 +828,7 @@ static int udf_unlink(struct inode *dir, struct dentry *dentry)
 
 	retval = -ENOENT;
 	lock_kernel();
-	fi = udf_find_entry(dir, dentry, &fibh, &cfi);
+	fi = udf_find_entry(dir, &dentry->d_name, &fibh, &cfi);
 	if (!fi)
 		goto out;
 
@@ -1113,7 +1112,7 @@ static int udf_rename(struct inode *old_dir, struct dentry *old_dentry,
 	struct udf_inode_info *old_iinfo = UDF_I(old_inode);
 
 	lock_kernel();
-	ofi = udf_find_entry(old_dir, old_dentry, &ofibh, &ocfi);
+	ofi = udf_find_entry(old_dir, &old_dentry->d_name, &ofibh, &ocfi);
 	if (ofi) {
 		if (ofibh.sbh != ofibh.ebh)
 			brelse(ofibh.ebh);
@@ -1124,7 +1123,7 @@ static int udf_rename(struct inode *old_dir, struct dentry *old_dentry,
 	    != old_inode->i_ino)
 		goto end_rename;
 
-	nfi = udf_find_entry(new_dir, new_dentry, &nfibh, &ncfi);
+	nfi = udf_find_entry(new_dir, &new_dentry->d_name, &nfibh, &ncfi);
 	if (nfi) {
 		if (!new_inode) {
 			if (nfibh.sbh != nfibh.ebh)
@@ -1192,7 +1191,7 @@ static int udf_rename(struct inode *old_dir, struct dentry *old_dentry,
 	udf_write_fi(new_dir, &ncfi, nfi, &nfibh, NULL, NULL);
 
 	/* The old fid may have moved - find it again */
-	ofi = udf_find_entry(old_dir, old_dentry, &ofibh, &ocfi);
+	ofi = udf_find_entry(old_dir, &old_dentry->d_name, &ofibh, &ocfi);
 	udf_delete_entry(old_dir, ofi, &ofibh, &ocfi);
 
 	if (new_inode) {
@@ -1243,15 +1242,11 @@ end_rename:
 
 static struct dentry *udf_get_parent(struct dentry *child)
 {
-	struct dentry *parent;
 	struct inode *inode = NULL;
-	struct dentry dotdot;
+	struct qstr dotdot = {.name = "..", .len = 2};
 	struct fileIdentDesc cfi;
 	struct udf_fileident_bh fibh;
 
-	dotdot.d_name.name = "..";
-	dotdot.d_name.len = 2;
-
 	lock_kernel();
 	if (!udf_find_entry(child->d_inode, &dotdot, &fibh, &cfi))
 		goto out_unlock;
@@ -1266,13 +1261,7 @@ static struct dentry *udf_get_parent(struct dentry *child)
 		goto out_unlock;
 	unlock_kernel();
 
-	parent = d_alloc_anon(inode);
-	if (!parent) {
-		iput(inode);
-		parent = ERR_PTR(-ENOMEM);
-	}
-
-	return parent;
+	return d_obtain_alias(inode);
 out_unlock:
 	unlock_kernel();
 	return ERR_PTR(-EACCES);
@@ -1283,7 +1272,6 @@ static struct dentry *udf_nfs_get_inode(struct super_block *sb, u32 block,
 					u16 partref, __u32 generation)
 {
 	struct inode *inode;
-	struct dentry *result;
 	kernel_lb_addr loc;
 
 	if (block == 0)
@@ -1300,12 +1288,7 @@ static struct dentry *udf_nfs_get_inode(struct super_block *sb, u32 block,
 		iput(inode);
 		return ERR_PTR(-ESTALE);
 	}
-	result = d_alloc_anon(inode);
-	if (!result) {
-		iput(inode);
-		return ERR_PTR(-ENOMEM);
-	}
-	return result;
+	return d_obtain_alias(inode);
 }
 
 static struct dentry *udf_fh_to_dentry(struct super_block *sb,

fs/ufs/dir.c

@@ -667,4 +667,5 @@ const struct file_operations ufs_dir_operations = {
 	.read		= generic_read_dir,
 	.readdir	= ufs_readdir,
 	.fsync		= file_fsync,
+	.llseek		= generic_file_llseek,
 };

fs/xfs/linux-2.6/xfs_export.c

@@ -148,7 +148,6 @@ xfs_fs_fh_to_dentry(struct super_block *sb, struct fid *fid,
 {
 	struct xfs_fid64	*fid64 = (struct xfs_fid64 *)fid;
 	struct inode		*inode = NULL;
-	struct dentry		*result;
 
 	if (fh_len < xfs_fileid_length(fileid_type))
 		return NULL;
@@ -164,16 +163,7 @@ xfs_fs_fh_to_dentry(struct super_block *sb, struct fid *fid,
 		break;
 	}
 
-	if (!inode)
-		return NULL;
-	if (IS_ERR(inode))
-		return ERR_CAST(inode);
-	result = d_alloc_anon(inode);
-	if (!result) {
-		iput(inode);
-		return ERR_PTR(-ENOMEM);
-	}
-	return result;
+	return d_obtain_alias(inode);
 }
 
 STATIC struct dentry *
@@ -182,7 +172,6 @@ xfs_fs_fh_to_parent(struct super_block *sb, struct fid *fid,
 {
 	struct xfs_fid64	*fid64 = (struct xfs_fid64 *)fid;
 	struct inode		*inode = NULL;
-	struct dentry		*result;
 
 	switch (fileid_type) {
 	case FILEID_INO32_GEN_PARENT:
@@ -195,16 +184,7 @@ xfs_fs_fh_to_parent(struct super_block *sb, struct fid *fid,
 		break;
 	}
 
-	if (!inode)
-		return NULL;
-	if (IS_ERR(inode))
-		return ERR_CAST(inode);
-	result = d_alloc_anon(inode);
-	if (!result) {
-		iput(inode);
-		return ERR_PTR(-ENOMEM);
-	}
-	return result;
+	return d_obtain_alias(inode);
 }
 
 STATIC struct dentry *
@@ -213,18 +193,12 @@ xfs_fs_get_parent(
 {
 	int			error;
 	struct xfs_inode	*cip;
-	struct dentry		*parent;
 
 	error = xfs_lookup(XFS_I(child->d_inode), &xfs_name_dotdot, &cip, NULL);
 	if (unlikely(error))
 		return ERR_PTR(-error);
 
-	parent = d_alloc_anon(VFS_I(cip));
-	if (unlikely(!parent)) {
-		iput(VFS_I(cip));
-		return ERR_PTR(-ENOMEM);
-	}
-	return parent;
+	return d_obtain_alias(VFS_I(cip));
 }
 
 const struct export_operations xfs_export_operations = {

fs/xfs/linux-2.6/xfs_file.c

@@ -204,15 +204,6 @@ xfs_file_fsync(
 	return -xfs_fsync(XFS_I(dentry->d_inode));
 }
 
-/*
- * Unfortunately we can't just use the clean and simple readdir implementation
- * below, because nfs might call back into ->lookup from the filldir callback
- * and that will deadlock the low-level btree code.
- *
- * Hopefully we'll find a better workaround that allows to use the optimal
- * version at least for local readdirs for 2.6.25.
- */
-#if 0
 STATIC int
 xfs_file_readdir(
 	struct file	*filp,
@@ -244,125 +235,6 @@ xfs_file_readdir(
 		return -error;
 	return 0;
 }
-#else
-
-struct hack_dirent {
-	u64		ino;
-	loff_t		offset;
-	int		namlen;
-	unsigned int	d_type;
-	char		name[];
-};
-
-struct hack_callback {
-	char		*dirent;
-	size_t		len;
-	size_t		used;
-};
-
-STATIC int
-xfs_hack_filldir(
-	void		*__buf,
-	const char	*name,
-	int		namlen,
-	loff_t		offset,
-	u64		ino,
-	unsigned int	d_type)
-{
-	struct hack_callback *buf = __buf;
-	struct hack_dirent *de = (struct hack_dirent *)(buf->dirent + buf->used);
-	unsigned int reclen;
-
-	reclen = ALIGN(sizeof(struct hack_dirent) + namlen, sizeof(u64));
-	if (buf->used + reclen > buf->len)
-		return -EINVAL;
-
-	de->namlen = namlen;
-	de->offset = offset;
-	de->ino = ino;
-	de->d_type = d_type;
-	memcpy(de->name, name, namlen);
-	buf->used += reclen;
-	return 0;
-}
-
-STATIC int
-xfs_file_readdir(
-	struct file	*filp,
-	void		*dirent,
-	filldir_t	filldir)
-{
-	struct inode	*inode = filp->f_path.dentry->d_inode;
-	xfs_inode_t	*ip = XFS_I(inode);
-	struct hack_callback buf;
-	struct hack_dirent *de;
-	int		error;
-	loff_t		size;
-	int		eof = 0;
-	xfs_off_t       start_offset, curr_offset, offset;
-
-	/*
-	 * Try fairly hard to get memory
-	 */
-	buf.len = PAGE_CACHE_SIZE;
-	do {
-		buf.dirent = kmalloc(buf.len, GFP_KERNEL);
-		if (buf.dirent)
-			break;
-		buf.len >>= 1;
-	} while (buf.len >= 1024);
-
-	if (!buf.dirent)
-		return -ENOMEM;
-
-	curr_offset = filp->f_pos;
-	if (curr_offset == 0x7fffffff)
-		offset = 0xffffffff;
-	else
-		offset = filp->f_pos;
-
-	while (!eof) {
-		unsigned int reclen;
-
-		start_offset = offset;
-
-		buf.used = 0;
-		error = -xfs_readdir(ip, &buf, buf.len, &offset,
-				     xfs_hack_filldir);
-		if (error || offset == start_offset) {
-			size = 0;
-			break;
-		}
-
-		size = buf.used;
-		de = (struct hack_dirent *)buf.dirent;
-		while (size > 0) {
-			curr_offset = de->offset /* & 0x7fffffff */;
-			if (filldir(dirent, de->name, de->namlen,
-					curr_offset & 0x7fffffff,
-					de->ino, de->d_type)) {
-				goto done;
-			}
-
-			reclen = ALIGN(sizeof(struct hack_dirent) + de->namlen,
-				       sizeof(u64));
-			size -= reclen;
-			de = (struct hack_dirent *)((char *)de + reclen);
-		}
-	}
-
- done:
-	if (!error) {
-		if (size == 0)
-			filp->f_pos = offset & 0x7fffffff;
-		else if (de)
-			filp->f_pos = curr_offset;
-	}
-
-	kfree(buf.dirent);
-	return error;
-}
-#endif
 
 STATIC int
 xfs_file_mmap(

fs/xfs/linux-2.6/xfs_ioctl.c

@@ -311,11 +311,10 @@ xfs_open_by_handle(
 		return new_fd;
 	}
 
-	dentry = d_alloc_anon(inode);
-	if (dentry == NULL) {
-		iput(inode);
+	dentry = d_obtain_alias(inode);
+	if (IS_ERR(dentry)) {
 		put_unused_fd(new_fd);
-		return -XFS_ERROR(ENOMEM);
+		return PTR_ERR(dentry);
 	}
 
 	/* Ensure umount returns EBUSY on umounts while this file is open. */

include/linux/dcache.h

@@ -228,9 +228,9 @@ extern void d_delete(struct dentry *);
 
 /* allocate/de-allocate */
 extern struct dentry * d_alloc(struct dentry *, const struct qstr *);
-extern struct dentry * d_alloc_anon(struct inode *);
 extern struct dentry * d_splice_alias(struct inode *, struct dentry *);
 extern struct dentry * d_add_ci(struct dentry *, struct inode *, struct qstr *);
+extern struct dentry * d_obtain_alias(struct inode *);
 extern void shrink_dcache_sb(struct super_block *);
 extern void shrink_dcache_parent(struct dentry *);
 extern void shrink_dcache_for_umount(struct super_block *);
@@ -287,6 +287,7 @@ static inline struct dentry *d_add_unique(struct dentry *entry, struct inode *in
 
 /* used for rename() and baskets */
 extern void d_move(struct dentry *, struct dentry *);
+extern struct dentry *d_ancestor(struct dentry *, struct dentry *);
 
 /* appendix may either be NULL or be used for transname suffixes */
 extern struct dentry * d_lookup(struct dentry *, struct qstr *);

include/linux/fs.h

@@ -136,7 +136,7 @@ extern int dir_notify_enable;
 /*
  * Superblock flags that can be altered by MS_REMOUNT
  */
-#define MS_RMT_MASK	(MS_RDONLY|MS_SYNCHRONOUS|MS_MANDLOCK)
+#define MS_RMT_MASK	(MS_RDONLY|MS_SYNCHRONOUS|MS_MANDLOCK|MS_I_VERSION)
 
 /*
  * Old magic mount flag and mask
@@ -1593,7 +1593,6 @@ extern int get_sb_pseudo(struct file_system_type *, char *,
 	struct vfsmount *mnt);
 extern int simple_set_mnt(struct vfsmount *mnt, struct super_block *sb);
 int __put_super_and_need_restart(struct super_block *sb);
-void unnamed_dev_init(void);
 
 /* Alas, no aliases. Too much hassle with bringing module.h everywhere */
 #define fops_get(fops) \
@@ -1852,6 +1851,11 @@ extern int inode_permission(struct inode *, int);
 extern int generic_permission(struct inode *, int,
 		int (*check_acl)(struct inode *, int));
 
+static inline bool execute_ok(struct inode *inode)
+{
+	return (inode->i_mode & S_IXUGO) || S_ISDIR(inode->i_mode);
+}
+
 extern int get_write_access(struct inode *);
 extern int deny_write_access(struct file *);
 static inline void put_write_access(struct inode * inode)

include/linux/namei.h

@@ -51,8 +51,10 @@ enum {LAST_NORM, LAST_ROOT, LAST_DOT, LAST_DOTDOT, LAST_BIND};
 /*
  * Intent data
  */
-#define LOOKUP_OPEN		(0x0100)
-#define LOOKUP_CREATE		(0x0200)
+#define LOOKUP_OPEN		0x0100
+#define LOOKUP_CREATE		0x0200
+#define LOOKUP_EXCL		0x0400
+#define LOOKUP_RENAME_TARGET	0x0800
 
 extern int user_path_at(int, const char __user *, unsigned, struct path *);
 
@@ -61,6 +63,8 @@ extern int user_path_at(int, const char __user *, unsigned, struct path *);
 #define user_path_dir(name, path) \
 	user_path_at(AT_FDCWD, name, LOOKUP_FOLLOW | LOOKUP_DIRECTORY, path)
 
+extern int kern_path(const char *, unsigned, struct path *);
+
 extern int path_lookup(const char *, unsigned, struct nameidata *);
 extern int vfs_path_lookup(struct dentry *, struct vfsmount *,
 			   const char *, unsigned int, struct nameidata *);

init/main.c

@@ -672,7 +672,6 @@ asmlinkage void __init start_kernel(void)
 	fork_init(num_physpages);
 	proc_caches_init();
 	buffer_init();
-	unnamed_dev_init();
 	key_init();
 	security_init();
 	vfs_caches_init(num_physpages);

kernel/audit_tree.c

@@ -532,7 +532,7 @@ void audit_trim_trees(void)
 	list_add(&cursor, &tree_list);
 	while (cursor.next != &tree_list) {
 		struct audit_tree *tree;
-		struct nameidata nd;
+		struct path path;
 		struct vfsmount *root_mnt;
 		struct node *node;
 		struct list_head list;
@@ -544,12 +544,12 @@ void audit_trim_trees(void)
 		list_add(&cursor, &tree->list);
 		mutex_unlock(&audit_filter_mutex);
 
-		err = path_lookup(tree->pathname, 0, &nd);
+		err = kern_path(tree->pathname, 0, &path);
 		if (err)
 			goto skip_it;
 
-		root_mnt = collect_mounts(nd.path.mnt, nd.path.dentry);
-		path_put(&nd.path);
+		root_mnt = collect_mounts(path.mnt, path.dentry);
+		path_put(&path);
 		if (!root_mnt)
 			goto skip_it;
 
@@ -580,19 +580,19 @@ skip_it:
 }
 
 static int is_under(struct vfsmount *mnt, struct dentry *dentry,
-		    struct nameidata *nd)
+		    struct path *path)
 {
-	if (mnt != nd->path.mnt) {
+	if (mnt != path->mnt) {
 		for (;;) {
 			if (mnt->mnt_parent == mnt)
 				return 0;
-			if (mnt->mnt_parent == nd->path.mnt)
+			if (mnt->mnt_parent == path->mnt)
 					break;
 			mnt = mnt->mnt_parent;
 		}
 		dentry = mnt->mnt_mountpoint;
 	}
-	return is_subdir(dentry, nd->path.dentry);
+	return is_subdir(dentry, path->dentry);
 }
 
 int audit_make_tree(struct audit_krule *rule, char *pathname, u32 op)
@@ -618,7 +618,7 @@ void audit_put_tree(struct audit_tree *tree)
 int audit_add_tree_rule(struct audit_krule *rule)
 {
 	struct audit_tree *seed = rule->tree, *tree;
-	struct nameidata nd;
+	struct path path;
 	struct vfsmount *mnt, *p;
 	struct list_head list;
 	int err;
@@ -637,11 +637,11 @@ int audit_add_tree_rule(struct audit_krule *rule)
 	/* do not set rule->tree yet */
 	mutex_unlock(&audit_filter_mutex);
 
-	err = path_lookup(tree->pathname, 0, &nd);
+	err = kern_path(tree->pathname, 0, &path);
 	if (err)
 		goto Err;
-	mnt = collect_mounts(nd.path.mnt, nd.path.dentry);
-	path_put(&nd.path);
+	mnt = collect_mounts(path.mnt, path.dentry);
+	path_put(&path);
 	if (!mnt) {
 		err = -ENOMEM;
 		goto Err;
@@ -690,29 +690,29 @@ int audit_tag_tree(char *old, char *new)
 {
 	struct list_head cursor, barrier;
 	int failed = 0;
-	struct nameidata nd;
+	struct path path;
 	struct vfsmount *tagged;
 	struct list_head list;
 	struct vfsmount *mnt;
 	struct dentry *dentry;
 	int err;
 
-	err = path_lookup(new, 0, &nd);
+	err = kern_path(new, 0, &path);
 	if (err)
 		return err;
-	tagged = collect_mounts(nd.path.mnt, nd.path.dentry);
-	path_put(&nd.path);
+	tagged = collect_mounts(path.mnt, path.dentry);
+	path_put(&path);
 	if (!tagged)
 		return -ENOMEM;
 
-	err = path_lookup(old, 0, &nd);
+	err = kern_path(old, 0, &path);
 	if (err) {
 		drop_collected_mounts(tagged);
 		return err;
 	}
-	mnt = mntget(nd.path.mnt);
-	dentry = dget(nd.path.dentry);
-	path_put(&nd.path);
+	mnt = mntget(path.mnt);
+	dentry = dget(path.dentry);
+	path_put(&path);
 
 	if (dentry == tagged->mnt_root && dentry == mnt->mnt_root)
 		follow_up(&mnt, &dentry);
@@ -733,7 +733,7 @@ int audit_tag_tree(char *old, char *new)
 		list_add(&cursor, &tree->list);
 		mutex_unlock(&audit_filter_mutex);
 
-		err = path_lookup(tree->pathname, 0, &nd);
+		err = kern_path(tree->pathname, 0, &path);
 		if (err) {
 			put_tree(tree);
 			mutex_lock(&audit_filter_mutex);
@@ -741,15 +741,15 @@ int audit_tag_tree(char *old, char *new)
 		}
 
 		spin_lock(&vfsmount_lock);
-		if (!is_under(mnt, dentry, &nd)) {
+		if (!is_under(mnt, dentry, &path)) {
 			spin_unlock(&vfsmount_lock);
-			path_put(&nd.path);
+			path_put(&path);
 			put_tree(tree);
 			mutex_lock(&audit_filter_mutex);
 			continue;
 		}
 		spin_unlock(&vfsmount_lock);
-		path_put(&nd.path);
+		path_put(&path);
 
 		list_for_each_entry(p, &list, mnt_list) {
 			failed = tag_chunk(p->mnt_root->d_inode, tree);

net/unix/af_unix.c

@@ -711,28 +711,30 @@ static struct sock *unix_find_other(struct net *net,
 				    int type, unsigned hash, int *error)
 {
 	struct sock *u;
-	struct nameidata nd;
+	struct path path;
 	int err = 0;
 
 	if (sunname->sun_path[0]) {
-		err = path_lookup(sunname->sun_path, LOOKUP_FOLLOW, &nd);
+		struct inode *inode;
+		err = kern_path(sunname->sun_path, LOOKUP_FOLLOW, &path);
 		if (err)
 			goto fail;
-		err = vfs_permission(&nd, MAY_WRITE);
+		inode = path.dentry->d_inode;
+		err = inode_permission(inode, MAY_WRITE);
 		if (err)
 			goto put_fail;
 
 		err = -ECONNREFUSED;
-		if (!S_ISSOCK(nd.path.dentry->d_inode->i_mode))
+		if (!S_ISSOCK(inode->i_mode))
 			goto put_fail;
-		u = unix_find_socket_byinode(net, nd.path.dentry->d_inode);
+		u = unix_find_socket_byinode(net, inode);
 		if (!u)
 			goto put_fail;
 
 		if (u->sk_type == type)
-			touch_atime(nd.path.mnt, nd.path.dentry);
+			touch_atime(path.mnt, path.dentry);
 
-		path_put(&nd.path);
+		path_put(&path);
 
 		err=-EPROTOTYPE;
 		if (u->sk_type != type) {
@@ -753,7 +755,7 @@ static struct sock *unix_find_other(struct net *net,
 	return u;
 
 put_fail:
-	path_put(&nd.path);
+	path_put(&path);
 fail:
 	*error=err;
 	return NULL;