[4/9] pohmelfs: directory operations.

[4/9] pohmelfs: directory operations.

[Evgeniy Polyakov]

This patch implementes all supported directory operations
like directory reading, object lookup, creation, removal
and so on.

Signed-off-by: Evgeniy Polyakov <zbr@ioremap.net>

diff --git a/fs/pohmelfs/dir.c b/fs/pohmelfs/dir.c
new file mode 100644
index 0000000..29eda12
--- /dev/null
+++ b/fs/pohmelfs/dir.c
@@ -0,0 +1,1139 @@
+/*
+ * 2007+ Copyright (c) Evgeniy Polyakov <zbr@ioremap.net>
+ * All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/kernel.h>
+#include <linux/fs.h>
+#include <linux/jhash.h>
+#include <linux/namei.h>
+#include <linux/pagemap.h>
+
+#include "netfs.h"
+
+static int pohmelfs_cmp_hash(struct pohmelfs_name *n, u32 hash)
+{
+	if (n->hash > hash)
+		return -1;
+	if (n->hash < hash)
+		return 1;
+
+	return 0;
+}
+
+static struct pohmelfs_name *pohmelfs_search_hash_unprecise(struct pohmelfs_inode *pi, u32 hash)
+{
+	struct rb_node *n = pi->hash_root.rb_node;
+	struct pohmelfs_name *tmp = NULL;
+	int cmp;
+
+	while (n) {
+		tmp = rb_entry(n, struct pohmelfs_name, hash_node);
+
+		cmp = pohmelfs_cmp_hash(tmp, hash);
+		if (cmp < 0)
+			n = n->rb_left;
+		else if (cmp > 0)
+			n = n->rb_right;
+		else
+			break;
+
+	}
+
+	return tmp;
+}
+
+struct pohmelfs_name *pohmelfs_search_hash(struct pohmelfs_inode *pi, u32 hash)
+{
+	struct pohmelfs_name *tmp;
+
+	tmp = pohmelfs_search_hash_unprecise(pi, hash);
+	if (tmp && (tmp->hash == hash))
+		return tmp;
+
+	return NULL;
+}
+
+static void __pohmelfs_name_del(struct pohmelfs_inode *parent, struct pohmelfs_name *node)
+{
+	rb_erase(&node->hash_node, &parent->hash_root);
+}
+
+/*
+ * Remove name cache entry from its caches and free it.
+ */
+static void pohmelfs_name_free(struct pohmelfs_inode *parent, struct pohmelfs_name *node)
+{
+	__pohmelfs_name_del(parent, node);
+	list_del(&node->sync_del_entry);
+	list_del(&node->sync_create_entry);
+	kfree(node);
+}
+
+static struct pohmelfs_name *pohmelfs_insert_hash(struct pohmelfs_inode *pi,
+		struct pohmelfs_name *new)
+{
+	struct rb_node **n = &pi->hash_root.rb_node, *parent = NULL;
+	struct pohmelfs_name *ret = NULL, *tmp;
+	int cmp;
+
+	while (*n) {
+		parent = *n;
+
+		tmp = rb_entry(parent, struct pohmelfs_name, hash_node);
+
+		cmp = pohmelfs_cmp_hash(tmp, new->hash);
+		if (cmp < 0)
+			n = &parent->rb_left;
+		else if (cmp > 0)
+			n = &parent->rb_right;
+		else {
+			ret = tmp;
+			break;
+		}
+	}
+
+	if (ret) {
+		printk("%s: exist: parent: %llu, ino: %llu, hash: %x, len: %u, data: '%s', "
+				           "new: ino: %llu, hash: %x, len: %u, data: '%s'.\n",
+				__func__, pi->ino,
+				ret->ino, ret->hash, ret->len, ret->data,
+				new->ino, new->hash, new->len, new->data);
+		ret->ino = new->ino;
+		return ret;
+	}
+
+	rb_link_node(&new->hash_node, parent, n);
+	rb_insert_color(&new->hash_node, &pi->hash_root);
+
+	return NULL;
+}
+
+/*
+ * Free name cache for given inode.
+ */
+void pohmelfs_free_names(struct pohmelfs_inode *parent)
+{
+	struct rb_node *rb_node;
+	struct pohmelfs_name *n;
+
+	for (rb_node = rb_first(&parent->hash_root); rb_node;) {
+		n = rb_entry(rb_node, struct pohmelfs_name, hash_node);
+		rb_node = rb_next(rb_node);
+
+		pohmelfs_name_free(parent, n);
+	}
+}
+
+static void pohmelfs_fix_offset(struct pohmelfs_inode *parent, struct pohmelfs_name *node)
+{
+	parent->total_len -= node->len;
+}
+
+/*
+ * Free name cache entry helper.
+ */
+void pohmelfs_name_del(struct pohmelfs_inode *parent, struct pohmelfs_name *node)
+{
+	pohmelfs_fix_offset(parent, node);
+	pohmelfs_name_free(parent, node);
+}
+
+/*
+ * Insert new name cache entry into all hash cache.
+ */
+static int pohmelfs_insert_name(struct pohmelfs_inode *parent, struct pohmelfs_name *n)
+{
+	struct pohmelfs_name *name;
+
+	name = pohmelfs_insert_hash(parent, n);
+	if (name)
+		return -EEXIST;
+
+	parent->total_len += n->len;
+	list_add_tail(&n->sync_create_entry, &parent->sync_create_list);
+
+	return 0;
+}
+
+/*
+ * Allocate new name cache entry.
+ */
+static struct pohmelfs_name *pohmelfs_name_alloc(unsigned int len)
+{
+	struct pohmelfs_name *n;
+
+	n = kzalloc(sizeof(struct pohmelfs_name) + len, GFP_KERNEL);
+	if (!n)
+		return NULL;
+
+	INIT_LIST_HEAD(&n->sync_create_entry);
+	INIT_LIST_HEAD(&n->sync_del_entry);
+
+	n->data = (char *)(n+1);
+
+	return n;
+}
+
+/*
+ * Add new name entry into directory's cache.
+ */
+static int pohmelfs_add_dir(struct pohmelfs_sb *psb, struct pohmelfs_inode *parent,
+		struct pohmelfs_inode *npi, struct qstr *str, unsigned int mode, int link)
+{
+	int err = -ENOMEM;
+	struct pohmelfs_name *n;
+	struct pohmelfs_path_entry *e = NULL;
+
+	n = pohmelfs_name_alloc(str->len + 1);
+	if (!n)
+		goto err_out_exit;
+
+	n->ino = npi->ino;
+	n->mode = mode;
+	n->len = str->len;
+	n->hash = str->hash;
+	sprintf(n->data, "%s", str->name);
+
+	if (!(str->len == 1 && str->name[0] == '.') &&
+			!(str->len == 2 && str->name[0] == '.' && str->name[1] == '.')) {
+		mutex_lock(&psb->path_lock);
+		e = pohmelfs_add_path_entry(psb, parent->ino, npi->ino, str, link, mode);
+		mutex_unlock(&psb->path_lock);
+		if (IS_ERR(e)) {
+			err = PTR_ERR(e);
+			goto err_out_free;
+		}
+	}
+
+	mutex_lock(&parent->offset_lock);
+	err = pohmelfs_insert_name(parent, n);
+	mutex_unlock(&parent->offset_lock);
+
+	if (err) {
+		if (err != -EEXIST)
+			goto err_out_remove;
+		kfree(n);
+	}
+
+	return 0;
+
+err_out_remove:
+	if (e) {
+		mutex_lock(&psb->path_lock);
+		pohmelfs_remove_path_entry(psb, e);
+		mutex_unlock(&psb->path_lock);
+	}
+err_out_free:
+	kfree(n);
+err_out_exit:
+	return err;
+}
+
+/*
+ * Create new inode for given parameters (name, inode info, parent).
+ * This does not create object on the server, it will be synced there during writeback.
+ */
+struct pohmelfs_inode *pohmelfs_new_inode(struct pohmelfs_sb *psb,
+		struct pohmelfs_inode *parent, struct qstr *str,
+		struct netfs_inode_info *info, int link)
+{
+	struct inode *new = NULL;
+	struct pohmelfs_inode *npi;
+	int err = -EEXIST;
+
+	dprintk("%s: creating inode: parent: %llu, ino: %llu, str: %p.\n",
+			__func__, (parent)?parent->ino:0, info->ino, str);
+
+	err = -ENOMEM;
+	new = iget_locked(psb->sb, info->ino);
+	if (!new)
+		goto err_out_exit;
+
+	npi = POHMELFS_I(new);
+	npi->ino = info->ino;
+	err = 0;
+
+	if (new->i_state & I_NEW) {
+		dprintk("%s: filling VFS inode: %lu/%llu.\n",
+				__func__, new->i_ino, info->ino);
+		pohmelfs_fill_inode(new, info);
+
+		if (S_ISDIR(info->mode)) {
+			struct qstr s;
+
+			s.name = ".";
+			s.len = 1;
+			s.hash = jhash(s.name, s.len, 0);
+
+			err = pohmelfs_add_dir(psb, npi, npi, &s, info->mode, 0);
+			if (err)
+				goto err_out_put;
+
+			s.name = "..";
+			s.len = 2;
+			s.hash = jhash(s.name, s.len, 0);
+
+			err = pohmelfs_add_dir(psb, npi, (parent)?parent:npi, &s,
+					(parent)?parent->vfs_inode.i_mode:npi->vfs_inode.i_mode, 0);
+			if (err)
+				goto err_out_put;
+		}
+	}
+
+	if (str) {
+		if (parent) {
+			err = pohmelfs_add_dir(psb, parent, npi, str, info->mode, link);
+
+			dprintk("%s: %s inserted name: '%s', new_offset: %llu, ino: %llu, parent: %llu.\n",
+					__func__, (err)?"unsuccessfully":"successfully",
+					str->name, parent->total_len, info->ino, parent->ino);
+
+			if (err && err != -EEXIST)
+				goto err_out_put;
+		} else {
+			mutex_lock(&psb->path_lock);
+			pohmelfs_add_path_entry(psb, npi->ino, npi->ino, str, link, info->mode);
+			mutex_unlock(&psb->path_lock);
+		}
+	}
+
+	if (new->i_state & I_NEW) {
+		if (parent)
+			mark_inode_dirty(&parent->vfs_inode);
+		mark_inode_dirty(new);
+	}
+	unlock_new_inode(new);
+
+	return npi;
+
+err_out_put:
+	printk("%s: putting inode: %p, npi: %p, error: %d.\n", __func__, new, npi, err);
+	iput(new);
+err_out_exit:
+	return ERR_PTR(err);
+}
+
+static int pohmelfs_remote_sync_complete(struct page **pages, unsigned int page_num,
+		void *private, int err)
+{
+	struct pohmelfs_inode *pi = private;
+	struct pohmelfs_sb *psb = POHMELFS_SB(pi->vfs_inode.i_sb);
+
+	dprintk("%s: ino: %llu, err: %d.\n", __func__, pi->ino, err);
+
+	if (err)
+		pi->error = err;
+	wake_up(&psb->wait);
+	pohmelfs_put_inode(pi);
+
+	return err;
+}
+
+/*
+ * Receive directory content from the server.
+ * This should be only done for objects, which were not created locally,
+ * and which were not synced previously.
+ */
+static int pohmelfs_sync_remote_dir(struct pohmelfs_inode *pi)
+{
+	struct inode *inode = &pi->vfs_inode;
+	struct pohmelfs_sb *psb = POHMELFS_SB(inode->i_sb);
+	long ret = msecs_to_jiffies(25000);
+	int err;
+
+	dprintk("%s: dir: %llu, state: %lx: created: %d, remote_synced: %d.\n",
+			__func__, pi->ino, pi->state, test_bit(NETFS_INODE_CREATED, &pi->state),
+			test_bit(NETFS_INODE_REMOTE_SYNCED, &pi->state));
+
+	if (!test_bit(NETFS_INODE_CREATED, &pi->state))
+		return 0;
+
+	if (test_bit(NETFS_INODE_REMOTE_SYNCED, &pi->state))
+		return 0;
+
+	if (!igrab(inode)) {
+		err = -ENOENT;
+		goto err_out_exit;
+	}
+
+	err = pohmelfs_meta_command(pi, NETFS_READDIR, NETFS_TRANS_SINGLE_DST,
+			pohmelfs_remote_sync_complete, pi, 0);
+	if (err)
+		goto err_out_exit;
+
+	pi->error = 0;
+	ret = wait_event_interruptible_timeout(psb->wait,
+			test_bit(NETFS_INODE_REMOTE_SYNCED, &pi->state) || pi->error, ret);
+	dprintk("%s: awake dir: %llu, ret: %ld, err: %d.\n", __func__, pi->ino, ret, pi->error);
+	if (ret <= 0) {
+		err = -ETIMEDOUT;
+		goto err_out_exit;
+	}
+
+	if (pi->error)
+		return pi->error;
+
+	return 0;
+
+err_out_exit:
+	clear_bit(NETFS_INODE_REMOTE_SYNCED, &pi->state);
+
+	return err;
+}
+
+static int pohmelfs_dir_open(struct inode *inode, struct file *file)
+{
+	file->private_data = NULL;
+	return 0;
+}
+
+/*
+ * VFS readdir callback. Syncs directory content from server if needed,
+ * and provides direntry info to the userspace.
+ */
+static int pohmelfs_readdir(struct file *file, void *dirent, filldir_t filldir)
+{
+	struct inode *inode = file->f_path.dentry->d_inode;
+	struct pohmelfs_inode *pi = POHMELFS_I(inode);
+	struct pohmelfs_name *n;
+	struct rb_node *rb_node;
+	int err = 0, mode;
+	u64 len;
+
+	dprintk("%s: parent: %llu, fpos: %llu, hash: %08lx.\n",
+			__func__, pi->ino, (u64)file->f_pos,
+			(unsigned long)file->private_data);
+
+	err = pohmelfs_data_lock(pi, 0, ~0, POHMELFS_READ_LOCK);
+	if (err)
+		return err;
+
+	err = pohmelfs_sync_remote_dir(pi);
+	if (err)
+		return err;
+
+	if (file->private_data && (file->private_data == (void *)(unsigned long)file->f_pos))
+		return 0;
+
+	mutex_lock(&pi->offset_lock);
+	n = pohmelfs_search_hash_unprecise(pi, (unsigned long)file->private_data);
+
+	while (n) {
+		mode = (n->mode >> 12) & 15;
+
+		dprintk("%s: offset: %llu, parent ino: %llu, name: '%s', len: %u, ino: %llu, "
+				"mode: %o/%o, fpos: %llu, hash: %08x.\n",
+				__func__, file->f_pos, pi->ino, n->data, n->len,
+				n->ino, n->mode, mode, file->f_pos, n->hash);
+
+		file->private_data = (void *)n->hash;
+
+		len = n->len;
+		err = filldir(dirent, n->data, n->len, file->f_pos, n->ino, mode);
+
+		if (err < 0) {
+			dprintk("%s: err: %d.\n", __func__, err);
+			err = 0;
+			break;
+		}
+
+		file->f_pos += len;
+
+		rb_node = rb_next(&n->hash_node);
+
+		if (!rb_node || (rb_node == &n->hash_node)) {
+			file->private_data = (void *)(unsigned long)file->f_pos;
+			break;
+		}
+
+		n = rb_entry(rb_node, struct pohmelfs_name, hash_node);
+	}
+	mutex_unlock(&pi->offset_lock);
+
+	return err;
+}
+
+static loff_t pohmelfs_dir_lseek(struct file *file, loff_t offset, int origin)
+{
+	file->f_pos = offset;
+	file->private_data = NULL;
+	return offset;
+}
+
+const struct file_operations pohmelfs_dir_fops = {
+	.open = pohmelfs_dir_open,
+	.read = generic_read_dir,
+	.llseek = pohmelfs_dir_lseek,
+	.readdir = pohmelfs_readdir,
+};
+
+/*
+ * Lookup single object on server.
+ */
+static int pohmelfs_lookup_single(struct pohmelfs_inode *parent,
+		struct qstr *str, u64 ino)
+{
+	struct pohmelfs_sb *psb = POHMELFS_SB(parent->vfs_inode.i_sb);
+	long ret = msecs_to_jiffies(5000);
+	int err;
+
+	set_bit(NETFS_COMMAND_PENDING, &parent->state);
+	err = pohmelfs_meta_command_data(parent, parent->ino, NETFS_LOOKUP,
+			(char *)str->name, NETFS_TRANS_SINGLE_DST, NULL, NULL, ino);
+	if (err)
+		goto err_out_exit;
+
+	err = 0;
+	ret = wait_event_interruptible_timeout(psb->wait,
+			!test_bit(NETFS_COMMAND_PENDING, &parent->state), ret);
+	if (ret == 0)
+		err = -ETIMEDOUT;
+	else if (signal_pending(current))
+		err = -EINTR;
+
+	if (err)
+		goto err_out_exit;
+
+	return 0;
+
+err_out_exit:
+	clear_bit(NETFS_COMMAND_PENDING, &parent->state);
+
+	printk("%s: failed: parent: %llu, ino: %llu, name: '%s', err: %d.\n",
+			__func__, parent->ino, ino, str->name, err);
+
+	return err;
+}
+
+/*
+ * VFS lookup callback.
+ * We first try to get inode number from local name cache, if we have one,
+ * then inode can be found in inode cache. If there is no inode or no object in
+ * local cache, try to lookup it on server. This only should be done for directories,
+ * which were not created locally, otherwise remote server does not know about dir at all,
+ * so no need to try to know that.
+ */
+struct dentry *pohmelfs_lookup(struct inode *dir, struct dentry *dentry, struct nameidata *nd)
+{
+	struct pohmelfs_inode *parent = POHMELFS_I(dir);
+	struct pohmelfs_name *n;
+	struct inode *inode = NULL;
+	unsigned long ino = 0;
+	int err, lock_type = POHMELFS_READ_LOCK;
+	struct qstr str = dentry->d_name;
+
+	if ((nd->intent.open.flags & O_ACCMODE) > 1)
+		lock_type = POHMELFS_WRITE_LOCK;
+
+	err = pohmelfs_data_lock(parent, 0, ~0, lock_type);
+	if (err)
+		goto out;
+
+	str.hash = jhash(dentry->d_name.name, dentry->d_name.len, 0);
+
+	mutex_lock(&parent->offset_lock);
+	n = pohmelfs_search_hash(parent, str.hash);
+	if (n)
+		ino = n->ino;
+	mutex_unlock(&parent->offset_lock);
+
+	dprintk("%s: 1 ino: %lu, inode: %p, name: '%s', hash: %x, parent_state: %lx.\n",
+			__func__, ino, inode, str.name, str.hash, parent->state);
+
+	if (ino) {
+		inode = ilookup(dir->i_sb, ino);
+		if (inode)
+			goto out;
+	}
+
+	dprintk("%s: dir: %p, dir_ino: %llu, name: '%s', len: %u, dir_state: %lx, ino: %lu.\n",
+			__func__, dir, parent->ino,
+			str.name, str.len, parent->state, ino);
+
+	if (!ino) {
+		if (!test_bit(NETFS_INODE_CREATED, &parent->state))
+			goto out;
+
+		if (test_bit(NETFS_INODE_REMOTE_SYNCED, &parent->state))
+			goto out;
+	}
+
+	err = pohmelfs_lookup_single(parent, &str, ino);
+	if (err)
+		goto out;
+
+	if (!ino) {
+		mutex_lock(&parent->offset_lock);
+		n = pohmelfs_search_hash(parent, str.hash);
+		if (n)
+			ino = n->ino;
+		mutex_unlock(&parent->offset_lock);
+	}
+
+	if (ino) {
+		inode = ilookup(dir->i_sb, ino);
+		printk("%s: second lookup ino: %lu, inode: %p, name: '%s', hash: %x.\n",
+				__func__, ino, inode, str.name, str.hash);
+		if (!inode) {
+			printk("%s: No inode for ino: %lu, name: '%s', hash: %x.\n",
+				__func__, ino, str.name, str.hash);
+			//return NULL;
+			return ERR_PTR(-EACCES);
+		}
+	} else {
+		printk("%s: No inode number : name: '%s', hash: %x.\n",
+			__func__, str.name, str.hash);
+	}
+out:
+	return d_splice_alias(inode, dentry);
+}
+
+/*
+ * Create new object in local cache. Object will be synced to server
+ * during writeback for given inode.
+ */
+struct pohmelfs_inode *pohmelfs_create_entry_local(struct pohmelfs_sb *psb,
+	struct pohmelfs_inode *parent, struct qstr *str, u64 start, int mode)
+{
+	struct pohmelfs_inode *npi;
+	int err = -ENOMEM;
+	struct netfs_inode_info info;
+
+	dprintk("%s: name: '%s', mode: %o, start: %llu.\n",
+			__func__, str->name, mode, start);
+
+	info.mode = mode;
+	info.ino = start;
+
+	if (!start)
+		info.ino = pohmelfs_new_ino(psb);
+
+	info.nlink = S_ISDIR(mode)?2:1;
+	info.uid = current->uid;
+	info.gid = current->gid;
+	info.size = 0;
+	info.blocksize = 512;
+	info.blocks = 0;
+	info.rdev = 0;
+	info.version = 0;
+
+	npi = pohmelfs_new_inode(psb, parent, str, &info, !!start);
+	if (IS_ERR(npi)) {
+		err = PTR_ERR(npi);
+		goto err_out_unlock;
+	}
+
+	set_bit(NETFS_INODE_REMOTE_SYNCED, &npi->state);
+
+	return npi;
+
+err_out_unlock:
+	dprintk("%s: err: %d.\n", __func__, err);
+	return ERR_PTR(err);
+}
+
+/*
+ * Create local object and bind it to dentry.
+ */
+static int pohmelfs_create_entry(struct inode *dir, struct dentry *dentry, u64 start, int mode)
+{
+	struct pohmelfs_sb *psb = POHMELFS_SB(dir->i_sb);
+	struct pohmelfs_inode *npi, *parent;
+	struct qstr str = dentry->d_name;
+	int err;
+
+	parent = POHMELFS_I(dir);
+
+	err = pohmelfs_data_lock(parent, 0, ~0, POHMELFS_WRITE_LOCK);
+	if (err)
+		return err;
+
+	str.hash = jhash(dentry->d_name.name, dentry->d_name.len, 0);
+
+	npi = pohmelfs_create_entry_local(psb, parent, &str, start, mode);
+	if (IS_ERR(npi))
+		return PTR_ERR(npi);
+
+	d_instantiate(dentry, &npi->vfs_inode);
+
+	dprintk("%s: parent: %llu, inode: %llu, name: '%s', parent_nlink: %d, nlink: %d.\n",
+			__func__, parent->ino, npi->ino, dentry->d_name.name,
+			(signed)dir->i_nlink, (signed)npi->vfs_inode.i_nlink);
+
+	return 0;
+}
+
+/*
+ * VFS create and mkdir callbacks.
+ */
+static int pohmelfs_create(struct inode *dir, struct dentry *dentry, int mode,
+		struct nameidata *nd)
+{
+	return pohmelfs_create_entry(dir, dentry, 0, mode);
+}
+
+static int pohmelfs_mkdir(struct inode *dir, struct dentry *dentry, int mode)
+{
+	int err;
+
+	inode_inc_link_count(dir);
+	err = pohmelfs_create_entry(dir, dentry, 0, mode | S_IFDIR);
+	if (err)
+		inode_dec_link_count(dir);
+
+	return err;
+}
+
+/*
+ * Remove entry from local cache.
+ * Object will not be removed from server, instead it will be queued into parent
+ * to-be-removed queue, which will be processed during parent writeback (parent
+ * also marked as dirty). Writeback will send remove request to server.
+ * Such approach allows to remove vey huge directories (like 2.6.24 kernel tree)
+ * with only single network command.
+ */
+static int pohmelfs_remove_entry(struct inode *dir, struct dentry *dentry)
+{
+	struct pohmelfs_sb *psb = POHMELFS_SB(dir->i_sb);
+	struct inode *inode = dentry->d_inode;
+	struct pohmelfs_inode *parent = POHMELFS_I(dir), *pi = POHMELFS_I(inode);
+	struct pohmelfs_name *n;
+	int err = -ENOENT;
+	struct qstr str = dentry->d_name;
+
+	err = pohmelfs_data_lock(parent, 0, ~0, POHMELFS_WRITE_LOCK);
+	if (err)
+		return err;
+
+	str.hash = jhash(dentry->d_name.name, dentry->d_name.len, 0);
+
+	dprintk("%s: dir_ino: %llu, inode: %llu, name: '%s', nlink: %d.\n",
+			__func__, parent->ino, pi->ino,
+			str.name, (signed)inode->i_nlink);
+
+	mutex_lock(&parent->offset_lock);
+	n = pohmelfs_search_hash(parent, str.hash);
+	if (n) {
+		pohmelfs_fix_offset(parent, n);
+		if (test_bit(NETFS_INODE_CREATED, &pi->state)) {
+			__pohmelfs_name_del(parent, n);
+			list_add_tail(&n->sync_del_entry, &parent->sync_del_list);
+		} else
+			pohmelfs_name_free(parent, n);
+		err = 0;
+	}
+	mutex_unlock(&parent->offset_lock);
+
+	if (!err) {
+		mutex_lock(&psb->path_lock);
+		psb->avail_size += inode->i_size;
+		pohmelfs_remove_path_entry_by_ino(psb, pi->ino);
+		mutex_unlock(&psb->path_lock);
+
+		pohmelfs_inode_del_inode(psb, pi);
+
+		mark_inode_dirty(dir);
+
+		inode->i_ctime = dir->i_ctime;
+		if (inode->i_nlink)
+			inode_dec_link_count(inode);
+	}
+	dprintk("%s: inode: %p, lock: %ld, unhashed: %d.\n",
+		__func__, pi, inode->i_state & I_LOCK, hlist_unhashed(&inode->i_hash));
+
+	return err;
+}
+
+/*
+ * Unlink and rmdir VFS callbacks.
+ */
+static int pohmelfs_unlink(struct inode *dir, struct dentry *dentry)
+{
+	return pohmelfs_remove_entry(dir, dentry);
+}
+
+static int pohmelfs_rmdir(struct inode *dir, struct dentry *dentry)
+{
+	int err;
+	struct inode *inode = dentry->d_inode;
+
+	dprintk("%s: parent: %llu, inode: %llu, name: '%s', parent_nlink: %d, nlink: %d.\n",
+			__func__, POHMELFS_I(dir)->ino, POHMELFS_I(inode)->ino,
+			dentry->d_name.name, (signed)dir->i_nlink, (signed)inode->i_nlink);
+
+	err = pohmelfs_remove_entry(dir, dentry);
+	if (!err) {
+		inode_dec_link_count(dir);
+		inode_dec_link_count(inode);
+	}
+
+	return err;
+}
+
+/*
+ * Link creation is synchronous.
+ * I'm lazy.
+ * Earth is somewhat round.
+ */
+static int pohmelfs_create_link(struct pohmelfs_inode *parent, struct qstr *obj,
+		struct pohmelfs_inode *target, struct qstr *tstr)
+{
+	struct super_block *sb = parent->vfs_inode.i_sb;
+	struct pohmelfs_sb *psb = POHMELFS_SB(sb);
+	struct netfs_cmd *cmd;
+	struct netfs_trans *t;
+	void *data;
+	int err, parent_len, target_len = 0, cur_len, path_size = 0;
+
+	err = pohmelfs_data_lock(parent, 0, ~0, POHMELFS_WRITE_LOCK);
+	if (err)
+		return err;
+
+	err = sb->s_op->write_inode(&parent->vfs_inode, 0);
+	if (err)
+		goto err_out_exit;
+
+	if (tstr)
+		target_len = tstr->len;
+
+	mutex_lock(&psb->path_lock);
+	parent_len = pohmelfs_path_length(parent);
+	if (target)
+		target_len += pohmelfs_path_length(target);
+	mutex_unlock(&psb->path_lock);
+
+	if (parent_len < 0) {
+		err = parent_len;
+		goto err_out_exit;
+	}
+	
+	if (target_len < 0) {
+		err = target_len;
+		goto err_out_exit;
+	}
+
+	t = netfs_trans_alloc(psb, parent_len + target_len + obj->len + 2, 0, 0);
+	if (!t) {
+		err = -ENOMEM;
+		goto err_out_exit;
+	}
+	cur_len = netfs_trans_cur_len(t);
+
+	cmd = netfs_trans_current(t);
+	if (IS_ERR(cmd)) {
+		err = PTR_ERR(cmd);
+		goto err_out_free;
+	}
+
+	data = (void *)(cmd + 1);
+	cur_len -= sizeof(struct netfs_cmd);
+
+	mutex_lock(&psb->path_lock);
+	err = pohmelfs_construct_path_string(parent, data, parent_len);
+	if (err > 0) {
+		path_size = err;
+		cur_len -= path_size;
+
+		err = snprintf(data + path_size, cur_len, "/%s|", obj->name);
+
+		path_size += err;
+		cur_len -= err;
+
+		cmd->ext = path_size - 1; /* No | symbol */
+
+		if (target) {
+			err = pohmelfs_construct_path_string(target, data + path_size, target_len);
+			if (err > 0) {
+				path_size += err + 1;
+				cur_len -= err + 1;
+			}
+		}
+	}
+	mutex_unlock(&psb->path_lock);
+
+	if (err < 0)
+		goto err_out_free;
+
+	cmd->start = 0;
+
+	if (!target && tstr) {
+		if (tstr->len > cur_len - 1) {
+			err = -ENAMETOOLONG;
+			goto err_out_free;
+		}
+
+		err = snprintf(data + path_size, cur_len, "%s", tstr->name) + 1 /* 0-byte */;
+		path_size += err;
+		cur_len -= err;
+		cmd->start = 1;
+	}
+
+	dprintk("%s: parent: %llu, obj: '%s', target_inode: %llu, target_str: '%s', full: '%s'.\n",
+			__func__, parent->ino, obj->name, (target)?target->ino:0, (tstr)?tstr->name:NULL,
+			(char *)data);
+
+	cmd->cmd = NETFS_LINK;
+	cmd->size = path_size;
+	cmd->id = parent->ino;
+
+	netfs_convert_cmd(cmd);
+	
+	netfs_trans_update(cmd, t, path_size);
+
+	err = netfs_trans_finish(t, psb);
+	if (err)
+		goto err_out_exit;
+
+	return 0;
+
+err_out_free:
+	t->result = err;
+	netfs_trans_put(t);
+err_out_exit:
+	return err;
+}
+
+/*
+ *  VFS hard and soft link callbacks.
+ */
+static int pohmelfs_link(struct dentry *old_dentry, struct inode *dir,
+	struct dentry *dentry)
+{
+	struct inode *inode = old_dentry->d_inode;
+	struct pohmelfs_inode *pi = POHMELFS_I(inode);
+	int err;
+	struct qstr str = dentry->d_name;
+
+	str.hash = jhash(dentry->d_name.name, dentry->d_name.len, 0);
+
+	err = inode->i_sb->s_op->write_inode(inode, 0);
+	if (err)
+		return err;
+
+	err = pohmelfs_create_link(POHMELFS_I(dir), &str, pi, NULL);
+	if (err)
+		return err;
+
+	return pohmelfs_create_entry(dir, dentry, pi->ino, inode->i_mode);
+}
+
+static int pohmelfs_symlink(struct inode *dir, struct dentry *dentry, const char *symname)
+{
+	struct qstr sym_str;
+	struct qstr str = dentry->d_name;
+	struct inode *inode;
+	int err;
+
+	str.hash = jhash(dentry->d_name.name, dentry->d_name.len, 0);
+
+	sym_str.name = symname;
+	sym_str.len = strlen(symname);
+
+	err = pohmelfs_create_link(POHMELFS_I(dir), &str, NULL, &sym_str);
+	if (err)
+		goto err_out_exit;
+
+	err = pohmelfs_create_entry(dir, dentry, 0, S_IFLNK | S_IRWXU | S_IRWXG | S_IRWXO);
+	if (err)
+		goto err_out_exit;
+
+	inode = dentry->d_inode;
+
+	err = page_symlink(inode, symname, sym_str.len + 1);
+	if (err)
+		goto err_out_put;
+
+	return 0;
+
+err_out_put:
+	iput(inode);
+err_out_exit:
+	return err;
+}
+
+static int pohmelfs_send_rename(struct pohmelfs_inode *pi, struct pohmelfs_inode *parent,
+		struct qstr *str)
+{
+	int path_len, err, total_len = 0, inode_len, parent_len;
+	char *path;
+	struct netfs_trans *t;
+	struct netfs_cmd *cmd;
+	struct pohmelfs_sb *psb = POHMELFS_SB(pi->vfs_inode.i_sb);
+
+	mutex_lock(&psb->path_lock);
+	parent_len = pohmelfs_path_length(parent);
+	inode_len = pohmelfs_path_length(pi);
+	mutex_unlock(&psb->path_lock);
+
+	if (parent_len < 0 || inode_len < 0)
+		return -EINVAL;
+
+	path_len = parent_len + inode_len + str->len + 3;
+
+	t = netfs_trans_alloc(psb, path_len, 0, 0);
+	if (!t)
+		return -ENOMEM;
+
+	cmd = netfs_trans_current(t);
+	path = (char *)(cmd + 1);
+
+	mutex_lock(&psb->path_lock);
+	err = pohmelfs_construct_path_string(pi, path, inode_len + 1);
+	if (err < 0)
+		goto err_out_unlock;
+
+	cmd->ext = err;
+
+	path += err;
+	total_len += err;
+	path_len -= err;
+
+	*path = '|';
+	path++;
+	total_len++;
+	path_len--;
+
+	err = pohmelfs_construct_path_string(parent, path, parent_len + 1);
+	if (err < 0)
+		goto err_out_unlock;
+	mutex_unlock(&psb->path_lock);
+
+	path += err;
+	total_len += err;
+	path_len -= err;
+
+	err = snprintf(path, path_len - 1, "/%s", str->name);
+
+	total_len += err + 1; /* 0 symbol */
+	path_len -= err + 1;
+
+	cmd->cmd = NETFS_RENAME;
+	cmd->id = pi->ino;
+	cmd->start = parent->ino;
+	cmd->size = total_len;
+
+	netfs_convert_cmd(cmd);
+	
+	netfs_trans_update(cmd, t, total_len);
+
+	return netfs_trans_finish(t, psb);
+
+err_out_unlock:
+	mutex_unlock(&psb->path_lock);
+	netfs_trans_free(t);
+	return err;
+}
+
+static int pohmelfs_rename(struct inode *old_dir, struct dentry *old_dentry,
+			struct inode *new_dir, struct dentry *new_dentry)
+{
+	struct pohmelfs_sb *psb = POHMELFS_SB(old_dir->i_sb);
+	struct inode *old_inode = old_dentry->d_inode;
+	struct pohmelfs_inode *old_parent, *old, *new_parent;
+	struct qstr str = new_dentry->d_name;
+	struct pohmelfs_name *n;
+	unsigned int old_hash;
+	int err = -ENOENT;
+	
+	if (new_dir) {
+		err = new_dir->i_sb->s_op->write_inode(new_dir, 0);
+		if (err)
+			return err;
+	}
+
+	err = old_inode->i_sb->s_op->write_inode(old_inode, 0);
+	if (err)
+		return err;
+
+	old_hash = jhash(old_dentry->d_name.name, old_dentry->d_name.len, 0);
+	str.hash = jhash(new_dentry->d_name.name, new_dentry->d_name.len, 0);
+
+	old = POHMELFS_I(old_inode);
+	old_parent = POHMELFS_I(old_dir);
+
+	str.len = new_dentry->d_name.len;
+	str.name = new_dentry->d_name.name;
+	str.hash = jhash(new_dentry->d_name.name, new_dentry->d_name.len, 0);
+
+	if (new_dir) {
+		new_parent = POHMELFS_I(new_dir);
+		err = -ENOTEMPTY;
+
+		if (S_ISDIR(old_inode->i_mode) &&
+				new_parent->total_len <= 3)
+			goto err_out_exit;
+	} else {
+		new_parent = old_parent;
+	}
+
+	dprintk("%s: ino: %llu, parent: %llu, name: '%s' -> parent: %llu, name: '%s'.\n",
+			__func__, old->ino, old_parent->ino, old_dentry->d_name.name,
+			new_parent->ino, new_dentry->d_name.name);
+
+	err = pohmelfs_send_rename(old, new_parent, &str);
+	if (err)
+		goto err_out_exit;
+
+	mutex_lock(&psb->path_lock);
+	err = pohmelfs_rename_path_entry(psb, old->ino, new_parent->ino, &str);
+	mutex_unlock(&psb->path_lock);
+	if (err)
+		goto err_out_exit;
+
+	n = pohmelfs_name_alloc(str.len + 1);
+	if (!n)
+		goto err_out_exit;
+
+	mutex_lock(&new_parent->offset_lock);
+	n->ino = old->ino;
+	n->mode = old_inode->i_mode;
+	n->len = str.len;
+	n->hash = str.hash;
+	sprintf(n->data, "%s", str.name);
+	
+	err = pohmelfs_insert_name(new_parent, n);
+	mutex_unlock(&new_parent->offset_lock);
+
+	if (err)
+		goto err_out_exit;
+
+	mutex_lock(&old_parent->offset_lock);
+	n = pohmelfs_search_hash(old_parent, old_hash);
+	if (n)
+		pohmelfs_name_del(old_parent, n);
+	mutex_unlock(&old_parent->offset_lock);
+
+	mark_inode_dirty(old_inode);
+	mark_inode_dirty(&new_parent->vfs_inode);
+
+	return 0;
+
+err_out_exit:
+	return err;
+}
+
+/*
+ * POHMELFS directory inode operations.
+ */
+const struct inode_operations pohmelfs_dir_inode_ops = {
+	.link		= pohmelfs_link,
+	.symlink	= pohmelfs_symlink,
+	.unlink		= pohmelfs_unlink,
+	.mkdir		= pohmelfs_mkdir,
+	.rmdir		= pohmelfs_rmdir,
+	.create		= pohmelfs_create,
+	.lookup 	= pohmelfs_lookup,
+	.setattr	= pohmelfs_setattr,
+	.rename		= pohmelfs_rename,
+};
diff --git a/fs/pohmelfs/path_entry.c b/fs/pohmelfs/path_entry.c
new file mode 100644
index 0000000..2dcbce7
--- /dev/null
+++ b/fs/pohmelfs/path_entry.c
@@ -0,0 +1,356 @@
+/*
+ * 2007+ Copyright (c) Evgeniy Polyakov <johnpol@2ka.mipt.ru>
+ * All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/module.h>
+#include <linux/slab.h>
+#include <linux/fs.h>
+#include <linux/ktime.h>
+#include <linux/fs.h>
+#include <linux/pagemap.h>
+#include <linux/writeback.h>
+#include <linux/mm.h>
+
+#include "netfs.h"
+
+/*
+ * Path cache.
+ *
+ * Used to create pathes to root, strings (or structures,
+ * containing name, mode, permissions and so on) used by userspace
+ * server to process data.
+ *
+ * Cache is local for client, and its inode numbers are never synced
+ * with anyone else, server operates on names and pathes, not some obscure ids.
+ */
+
+static void pohmelfs_free_path_entry(struct pohmelfs_path_entry *e)
+{
+	kfree(e);
+}
+
+static struct pohmelfs_path_entry *pohmelfs_alloc_path_entry(unsigned int len)
+{
+	struct pohmelfs_path_entry *e;
+
+	e = kzalloc(len + 1 + sizeof(struct pohmelfs_path_entry), GFP_KERNEL);
+	if (!e)
+		return NULL;
+
+	e->name = (char *)((struct pohmelfs_path_entry *)(e + 1));
+	e->len = len;
+	atomic_set(&e->refcnt, 1);
+
+	return e;
+}
+
+static inline int pohmelfs_cmp_path_entry(u64 path_ino, u64 new_ino)
+{
+	if (path_ino > new_ino)
+		return -1;
+	if (path_ino < new_ino)
+		return 1;
+	return 0;
+}
+
+static struct pohmelfs_path_entry *pohmelfs_search_path_entry(struct rb_root *root, u64 ino)
+{
+	struct rb_node *n = root->rb_node;
+	struct pohmelfs_path_entry *tmp;
+	int cmp;
+
+	while (n) {
+		tmp = rb_entry(n, struct pohmelfs_path_entry, path_entry);
+
+		cmp = pohmelfs_cmp_path_entry(tmp->ino, ino);
+		if (cmp < 0)
+			n = n->rb_left;
+		else if (cmp > 0)
+			n = n->rb_right;
+		else
+			return tmp;
+	}
+
+	dprintk("%s: Failed to find path entry for ino: %llu.\n", __func__, ino);
+	return NULL;
+}
+
+static struct pohmelfs_path_entry *pohmelfs_insert_path_entry(struct rb_root *root,
+		struct pohmelfs_path_entry *new)
+{
+	struct rb_node **n = &root->rb_node, *parent = NULL;
+	struct pohmelfs_path_entry *ret = NULL, *tmp;
+	int cmp;
+
+	while (*n) {
+		parent = *n;
+
+		tmp = rb_entry(parent, struct pohmelfs_path_entry, path_entry);
+
+		cmp = pohmelfs_cmp_path_entry(tmp->ino, new->ino);
+		if (cmp < 0)
+			n = &parent->rb_left;
+		else if (cmp > 0)
+			n = &parent->rb_right;
+		else {
+			ret = tmp;
+			break;
+		}
+	}
+
+	if (ret) {
+		printk("%s: exist: ino: %llu, data: '%s', new: ino: %llu, data: '%s'.\n",
+			__func__, ret->ino, ret->name, new->ino, new->name);
+		return ret;
+	}
+
+	rb_link_node(&new->path_entry, parent, n);
+	rb_insert_color(&new->path_entry, root);
+
+	dprintk("%s: inserted: ino: %llu, data: '%s', parent: ino: %llu, data: '%s'.\n",
+		__func__, new->ino, new->name, new->parent->ino, new->parent->name);
+
+	return new;
+}
+
+void pohmelfs_remove_path_entry(struct pohmelfs_sb *psb, struct pohmelfs_path_entry *e)
+{
+	if (atomic_dec_and_test(&e->refcnt)) {
+		rb_erase(&e->path_entry, &psb->path_root);
+
+		if (e->parent != e)
+			pohmelfs_remove_path_entry(psb, e->parent);
+		pohmelfs_free_path_entry(e);
+	}
+}
+
+void pohmelfs_remove_path_entry_by_ino(struct pohmelfs_sb *psb, u64 ino)
+{
+	struct pohmelfs_path_entry *e;
+
+	e = pohmelfs_search_path_entry(&psb->path_root, ino);
+	if (e)
+		pohmelfs_remove_path_entry(psb, e);
+}
+
+int pohmelfs_change_path_entry(struct pohmelfs_sb *psb, u64 ino, unsigned int mode)
+{
+	struct pohmelfs_path_entry *e;
+
+	e = pohmelfs_search_path_entry(&psb->path_root, ino);
+	if (!e)
+		return -ENOENT;
+
+	e->mode = mode;
+	return 0;
+}
+
+int pohmelfs_rename_path_entry(struct pohmelfs_sb *psb, u64 ino, u64 parent_ino, struct qstr *str)
+{
+	struct pohmelfs_path_entry *e;
+	unsigned int mode, link;
+
+	e = pohmelfs_search_path_entry(&psb->path_root, ino);
+	if (!e)
+		return -ENOENT;
+
+	if ((e->len >= str->len) && (parent_ino == e->parent->ino)) {
+		sprintf(e->name, "%s", str->name);
+		e->len = str->len;
+		e->hash = str->hash;
+
+		return 0;
+	}
+
+	mode = e->mode;
+	link = e->link;
+
+	pohmelfs_remove_path_entry(psb, e);
+
+	e = pohmelfs_add_path_entry(psb, parent_ino, ino, str, link, mode);
+	if (IS_ERR(e))
+		return PTR_ERR(e);
+
+	return 0;
+}
+
+struct pohmelfs_path_entry * pohmelfs_add_path_entry(struct pohmelfs_sb *psb,
+	u64 parent_ino, u64 ino, struct qstr *str, int link, unsigned int mode)
+{
+	struct pohmelfs_path_entry *e, *ret, *parent;
+
+	parent = pohmelfs_search_path_entry(&psb->path_root, parent_ino);
+
+	e = pohmelfs_alloc_path_entry(str->len);
+	if (!e)
+		return ERR_PTR(-ENOMEM);
+
+	e->parent = e;
+	if (parent) {
+		e->parent = parent;
+		atomic_inc(&parent->refcnt);
+	}
+
+	e->ino = ino;
+	e->hash = str->hash;
+	e->link = link;
+	e->mode = mode;
+
+	sprintf(e->name, "%s", str->name);
+
+	ret = pohmelfs_insert_path_entry(&psb->path_root, e);
+	if (ret != e) {
+		pohmelfs_free_path_entry(e);
+		e = ret;
+	}
+
+	dprintk("%s: parent: %llu, ino: %llu, name: '%s', len: %u.\n",
+			__func__, parent_ino, ino, e->name, e->len);
+
+	return e;
+}
+
+static int pohmelfs_prepare_path(struct pohmelfs_inode *pi, struct list_head *list, int len, int create)
+{
+	struct pohmelfs_path_entry *e;
+	struct pohmelfs_sb *psb = POHMELFS_SB(pi->vfs_inode.i_sb);
+
+	e = pohmelfs_search_path_entry(&psb->path_root, pi->ino);
+	if (!e)
+		return -ENOENT;
+
+	while (e && e->parent != e) {
+		if (len < e->len + create)
+			return -ETOOSMALL;
+
+		len -= e->len + create;
+
+		list_add(&e->entry, list);
+		e = e->parent;
+	}
+
+	return 0;
+}
+
+/*
+ * Create path from root for given inode.
+ * Path is formed as set of stuctures, containing name of the object
+ * and its inode data (mode, permissions and so on).
+ */
+int pohmelfs_construct_path(struct pohmelfs_inode *pi, void *data, int len)
+{
+	struct pohmelfs_path_entry *e;
+	struct netfs_path_entry *ne = data;
+	int used = 0, err;
+	LIST_HEAD(list);
+
+	err = pohmelfs_prepare_path(pi, &list, len, sizeof(struct netfs_path_entry));
+	if (err)
+		return err;
+
+	list_for_each_entry(e, &list, entry) {
+		ne = data;
+		ne->mode = e->mode;
+		ne->len = e->len;
+
+		used += sizeof(struct netfs_path_entry);
+		data += sizeof(struct netfs_path_entry);
+
+		if (ne->len <= sizeof(ne->unused)) {
+			memcpy(ne->unused, e->name, ne->len);
+		} else {
+			memcpy(data, e->name, ne->len);
+			data += ne->len;
+			used += ne->len;
+		}
+
+		dprintk("%s: ino: %llu, mode: %o, is_link: %d, name: '%s', used: %d, ne_len: %u.\n",
+				__func__, e->ino, ne->mode, e->link, e->name, used, ne->len);
+
+		netfs_convert_path_entry(ne);
+	}
+
+	return used;
+}
+
+/*
+ * Create path from root for given inode.
+ */
+int pohmelfs_construct_path_string(struct pohmelfs_inode *pi, void *data, int len)
+{
+	struct pohmelfs_path_entry *e;
+	int used = 0, err;
+	char *ptr = data;
+	LIST_HEAD(list);
+
+	err = pohmelfs_prepare_path(pi, &list, len, 0);
+	if (err)
+		return err;
+
+	if (list_empty(&list)) {
+		err = sprintf(ptr, "/");
+		ptr += err;
+		used += err;
+	} else {
+		list_for_each_entry(e, &list, entry) {
+			err = sprintf(ptr, "/%s", e->name);
+
+			BUG_ON(!e->name);
+
+			ptr += err;
+			used += err;
+		}
+	}
+
+	dprintk("%s: inode: %llu, full path: '%s', used: %d.\n",
+			__func__, pi->ino, (char *)data, used);
+
+	return used;
+}
+
+static int pohmelfs_get_path_length(struct pohmelfs_inode *pi, int create)
+{
+	struct pohmelfs_path_entry *e;
+	struct pohmelfs_sb *psb = POHMELFS_SB(pi->vfs_inode.i_sb);
+	int len = 1 + create;
+
+	e = pohmelfs_search_path_entry(&psb->path_root, pi->ino);
+
+	/*
+	 * This should never happen actually.
+	 */
+	if (!e)
+		return -ENOENT;
+
+	while (e && e->parent != e) {
+		len += e->len + create + 1;
+		e = e->parent;
+	}
+
+	return len;
+}
+
+int pohmelfs_path_length(struct pohmelfs_inode *pi)
+{
+	int len = pohmelfs_get_path_length(pi, 0);
+
+	if (len < 0)
+		return len;
+	return len;
+}
+
+int pohmelfs_path_length_create(struct pohmelfs_inode *pi)
+{
+	return pohmelfs_get_path_length(pi, sizeof(struct netfs_path_entry));
+}

Re: [4/9] pohmelfs: directory operations.

[Yehuda Sadeh Weinraub]

Hi Evgeniy,
  I've played a bit with pohmelfs and it looks very nice. Specifically
I was very happy with the easy configuration on both the server and
the client.

I understand that you manage the metadata operations asynchronously.
How do you handle multiple clients in pohmelfs? Any specific locking?

I did have some problems. I was trying to run dbench and it failed.
Also I understand that you haven't implemented all vfs interface yet
(chmod, chown, etc.). I was looking at the following code (more of at
the comments), and it explained some strange stuff that I had seen:

> +/*
> + * Create new inode for given parameters (name, inode info, parent).
> + * This does not create object on the server, it will be synced there
> during writeback.
> + */
> +struct pohmelfs_inode *pohmelfs_new_inode(struct pohmelfs_sb *psb,
> +               struct pohmelfs_inode *parent, struct qstr *str,
> +               struct netfs_inode_info *info, int link)

...
> +/*
> + * Create new object in local cache. Object will be synced to server
> + * during writeback for given inode.
> + */
> +struct pohmelfs_inode *pohmelfs_create_entry_local(struct pohmelfs_sb *psb,
> +       struct pohmelfs_inode *parent, struct qstr *str, u64 start, int mode)
> +{
...
> +/*
> + * Create local object and bind it to dentry.
> + */
> +static int pohmelfs_create_entry(struct inode *dir, struct dentry
> *dentry, u64 start, int mode)
> +{
> +       struct pohmelfs_sb *psb = POHMELFS_SB(dir->i_sb);
> +       struct pohmelfs_inode *npi, *parent;
> +       struct qstr str = dentry->d_name;
> +       int err;
> +
> +       parent = POHMELFS_I(dir);
> +
> +       err = pohmelfs_data_lock(parent, 0, ~0, POHMELFS_WRITE_LOCK);
> +       if (err)
> +               return err;
> +
> +       str.hash = jhash(dentry->d_name.name, dentry->d_name.len, 0);
> +
> +       npi = pohmelfs_create_entry_local(psb, parent, &str, start, mode);
> +       if (IS_ERR(npi))
> +               return PTR_ERR(npi);
> +
> +       d_instantiate(dentry, &npi->vfs_inode);
> +
> +       dprintk("%s: parent: %llu, inode: %llu, name: '%s', parent_nlink:
> %d, nlink: %d.\n",
> +                       __func__, parent->ino, npi->ino, dentry->d_name.name,
> +                       (signed)dir->i_nlink, (signed)npi->vfs_inode.i_nlink);
> +
> +       return 0;
> +}
> +
> +/*
> + * VFS create and mkdir callbacks.
> + */
> +static int pohmelfs_create(struct inode *dir, struct dentry *dentry, int mode,
> +               struct nameidata *nd)
> +{
> +       return pohmelfs_create_entry(dir, dentry, 0, mode);
> +}
> +
> +static int pohmelfs_mkdir(struct inode *dir, struct dentry *dentry, int mode)
> +{
> +       int err;
> +
> +       inode_inc_link_count(dir);
> +       err = pohmelfs_create_entry(dir, dentry, 0, mode | S_IFDIR);
> +       if (err)
> +               inode_dec_link_count(dir);
> +
> +       return err;
> +}

I understand that you handle metadata operations locally, and for
example when doing mkdir or creating a file, the file/directory is
created locally, the inode is set dirty and when a writeback occurs
everything is syncronized to the server. Isn't there some problem in
cases where there are more than one client operating on the same
directory with the same filename then? Moreover, it is possible that
client A creates a file and client B creates a directory with the same
name. Am I missing something, or is that the intended design?

Thanks,
Yehuda

Re: [4/9] pohmelfs: directory operations.

[Evgeniy Polyakov]

Hi.

On Tue, Jan 13, 2009 at 04:43:30PM -0800, Yehuda Sadeh Weinraub (yehudasa@gmail.com) wrote:
>   I've played a bit with pohmelfs and it looks very nice. Specifically
> I was very happy with the easy configuration on both the server and
> the client.

Thanks a lot for giving it a try!

> I understand that you manage the metadata operations asynchronously.
> How do you handle multiple clients in pohmelfs? Any specific locking?

There is a cache coherency protocol which is managed by the server, so
effectively every access requests a lock and if another client
influences the object, server will ask the first one to flush its
changes to the server which will then be sent to the second client after
it gets the lock. Locks are cached and not requested again when new
operation happens, it will be requested only if server requested to drop
the lock.

> I did have some problems. I was trying to run dbench and it failed.
> Also I understand that you haven't implemented all vfs interface yet
> (chmod, chown, etc.). I was looking at the following code (more of at
> the comments), and it explained some strange stuff that I had seen:

They should be implemented already in the change attribute code.
Attribute changes are flushed at writeback time or on server's demand.

I will try to reproduce the problem locally, thanks for the report.
Can you provide the output of the command, its parameters and the dmesg?

> I understand that you handle metadata operations locally, and for
> example when doing mkdir or creating a file, the file/directory is
> created locally, the inode is set dirty and when a writeback occurs
> everything is syncronized to the server. Isn't there some problem in
> cases where there are more than one client operating on the same
> directory with the same filename then? Moreover, it is possible that
> client A creates a file and client B creates a directory with the same
> name. Am I missing something, or is that the intended design?

There are distributed locks used in the POHMELFS, and appropriate
protocol maintains coherent cache on all the clients.

In the example you showed, the first client will grab the parent
directory lock, create a file there and then flush it to the server when
the second client starts an operation. It will then detect that there is
a file with given name and file the processing (or will overwrite its
content, depending on the application of course).

-- 
	Evgeniy Polyakov

Re: [4/9] pohmelfs: directory operations.

[Yehuda Sadeh Weinraub]

>> I did have some problems. I was trying to run dbench and it failed.
>> Also I understand that you haven't implemented all vfs interface yet
>> (chmod, chown, etc.). I was looking at the following code (more of at
>> the comments), and it explained some strange stuff that I had seen:
>
> They should be implemented already in the change attribute code.
> Attribute changes are flushed at writeback time or on server's demand.
>
ok, might be that the following is the problem:

int pohmelfs_setattr(struct dentry *dentry, struct iattr *attr)
{
...
        if (!(psb->state_flags & POHMELFS_FLAGS_XATTR))
                return -EOPNOTSUPP;


Probably this test should be only on setxattr?

Re: [4/9] pohmelfs: directory operations.

[Evgeniy Polyakov]

On Tue, Jan 13, 2009 at 05:28:29PM -0800, Yehuda Sadeh Weinraub (yehudasa@gmail.com) wrote:
> > They should be implemented already in the change attribute code.
> > Attribute changes are flushed at writeback time or on server's demand.
> >
> ok, might be that the following is the problem:
> 
> int pohmelfs_setattr(struct dentry *dentry, struct iattr *attr)
> {
> ...
>         if (!(psb->state_flags & POHMELFS_FLAGS_XATTR))
>                 return -EOPNOTSUPP;
> 
> 
> Probably this test should be only on setxattr?

Definitely :)
It should be moved there. Thanks for spotting this.

-- 
	Evgeniy Polyakov

[4/9] pohmelfs: directory operations.

[Evgeniy Polyakov]

This patch implementes all supported directory operations
like directory reading, object lookup, creation, removal
and so on.

Currently object removal is not optimized at all.

Signed-off-by: Evgeniy Polyakov <zbr@ioremap.net>

diff --git a/drivers/staging/pohmelfs/dir.c b/drivers/staging/pohmelfs/dir.c
new file mode 100644
index 0000000..3b795c9
--- /dev/null
+++ b/drivers/staging/pohmelfs/dir.c
@@ -0,0 +1,1093 @@
+/*
+ * 2007+ Copyright (c) Evgeniy Polyakov <zbr@ioremap.net>
+ * All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/kernel.h>
+#include <linux/fs.h>
+#include <linux/jhash.h>
+#include <linux/namei.h>
+#include <linux/pagemap.h>
+
+#include "netfs.h"
+
+static int pohmelfs_cmp_hash(struct pohmelfs_name *n, u32 hash)
+{
+	if (n->hash > hash)
+		return -1;
+	if (n->hash < hash)
+		return 1;
+
+	return 0;
+}
+
+static struct pohmelfs_name *pohmelfs_search_hash_unprecise(struct pohmelfs_inode *pi, u32 hash)
+{
+	struct rb_node *n = pi->hash_root.rb_node;
+	struct pohmelfs_name *tmp = NULL;
+	int cmp;
+
+	while (n) {
+		tmp = rb_entry(n, struct pohmelfs_name, hash_node);
+
+		cmp = pohmelfs_cmp_hash(tmp, hash);
+		if (cmp < 0)
+			n = n->rb_left;
+		else if (cmp > 0)
+			n = n->rb_right;
+		else
+			break;
+
+	}
+
+	return tmp;
+}
+
+struct pohmelfs_name *pohmelfs_search_hash(struct pohmelfs_inode *pi, u32 hash)
+{
+	struct pohmelfs_name *tmp;
+
+	tmp = pohmelfs_search_hash_unprecise(pi, hash);
+	if (tmp && (tmp->hash == hash))
+		return tmp;
+
+	return NULL;
+}
+
+static void __pohmelfs_name_del(struct pohmelfs_inode *parent, struct pohmelfs_name *node)
+{
+	rb_erase(&node->hash_node, &parent->hash_root);
+}
+
+/*
+ * Remove name cache entry from its caches and free it.
+ */
+static void pohmelfs_name_free(struct pohmelfs_inode *parent, struct pohmelfs_name *node)
+{
+	__pohmelfs_name_del(parent, node);
+	list_del(&node->sync_create_entry);
+	kfree(node);
+}
+
+static struct pohmelfs_name *pohmelfs_insert_hash(struct pohmelfs_inode *pi,
+		struct pohmelfs_name *new)
+{
+	struct rb_node **n = &pi->hash_root.rb_node, *parent = NULL;
+	struct pohmelfs_name *ret = NULL, *tmp;
+	int cmp;
+
+	while (*n) {
+		parent = *n;
+
+		tmp = rb_entry(parent, struct pohmelfs_name, hash_node);
+
+		cmp = pohmelfs_cmp_hash(tmp, new->hash);
+		if (cmp < 0)
+			n = &parent->rb_left;
+		else if (cmp > 0)
+			n = &parent->rb_right;
+		else {
+			ret = tmp;
+			break;
+		}
+	}
+
+	if (ret) {
+		printk("%s: exist: parent: %llu, ino: %llu, hash: %x, len: %u, data: '%s', "
+				           "new: ino: %llu, hash: %x, len: %u, data: '%s'.\n",
+				__func__, pi->ino,
+				ret->ino, ret->hash, ret->len, ret->data,
+				new->ino, new->hash, new->len, new->data);
+		ret->ino = new->ino;
+		return ret;
+	}
+
+	rb_link_node(&new->hash_node, parent, n);
+	rb_insert_color(&new->hash_node, &pi->hash_root);
+
+	return NULL;
+}
+
+/*
+ * Free name cache for given inode.
+ */
+void pohmelfs_free_names(struct pohmelfs_inode *parent)
+{
+	struct rb_node *rb_node;
+	struct pohmelfs_name *n;
+
+	for (rb_node = rb_first(&parent->hash_root); rb_node;) {
+		n = rb_entry(rb_node, struct pohmelfs_name, hash_node);
+		rb_node = rb_next(rb_node);
+
+		pohmelfs_name_free(parent, n);
+	}
+}
+
+static void pohmelfs_fix_offset(struct pohmelfs_inode *parent, struct pohmelfs_name *node)
+{
+	parent->total_len -= node->len;
+}
+
+/*
+ * Free name cache entry helper.
+ */
+void pohmelfs_name_del(struct pohmelfs_inode *parent, struct pohmelfs_name *node)
+{
+	pohmelfs_fix_offset(parent, node);
+	pohmelfs_name_free(parent, node);
+}
+
+/*
+ * Insert new name cache entry into all hash cache.
+ */
+static int pohmelfs_insert_name(struct pohmelfs_inode *parent, struct pohmelfs_name *n)
+{
+	struct pohmelfs_name *name;
+
+	name = pohmelfs_insert_hash(parent, n);
+	if (name)
+		return -EEXIST;
+
+	parent->total_len += n->len;
+	list_add_tail(&n->sync_create_entry, &parent->sync_create_list);
+
+	return 0;
+}
+
+/*
+ * Allocate new name cache entry.
+ */
+static struct pohmelfs_name *pohmelfs_name_alloc(unsigned int len)
+{
+	struct pohmelfs_name *n;
+
+	n = kzalloc(sizeof(struct pohmelfs_name) + len, GFP_KERNEL);
+	if (!n)
+		return NULL;
+
+	INIT_LIST_HEAD(&n->sync_create_entry);
+
+	n->data = (char *)(n+1);
+
+	return n;
+}
+
+/*
+ * Add new name entry into directory's cache.
+ */
+static int pohmelfs_add_dir(struct pohmelfs_sb *psb, struct pohmelfs_inode *parent,
+		struct pohmelfs_inode *npi, struct qstr *str, unsigned int mode, int link)
+{
+	int err = -ENOMEM;
+	struct pohmelfs_name *n;
+
+	n = pohmelfs_name_alloc(str->len + 1);
+	if (!n)
+		goto err_out_exit;
+
+	n->ino = npi->ino;
+	n->mode = mode;
+	n->len = str->len;
+	n->hash = str->hash;
+	sprintf(n->data, "%s", str->name);
+
+	mutex_lock(&parent->offset_lock);
+	err = pohmelfs_insert_name(parent, n);
+	mutex_unlock(&parent->offset_lock);
+
+	if (err) {
+		if (err != -EEXIST)
+			goto err_out_free;
+		kfree(n);
+	}
+
+	return 0;
+
+err_out_free:
+	kfree(n);
+err_out_exit:
+	return err;
+}
+
+/*
+ * Create new inode for given parameters (name, inode info, parent).
+ * This does not create object on the server, it will be synced there during writeback.
+ */
+struct pohmelfs_inode *pohmelfs_new_inode(struct pohmelfs_sb *psb,
+		struct pohmelfs_inode *parent, struct qstr *str,
+		struct netfs_inode_info *info, int link)
+{
+	struct inode *new = NULL;
+	struct pohmelfs_inode *npi;
+	int err = -EEXIST;
+
+	dprintk("%s: creating inode: parent: %llu, ino: %llu, str: %p.\n",
+			__func__, (parent)?parent->ino:0, info->ino, str);
+
+	err = -ENOMEM;
+	new = iget_locked(psb->sb, info->ino);
+	if (!new)
+		goto err_out_exit;
+
+	npi = POHMELFS_I(new);
+	npi->ino = info->ino;
+	err = 0;
+
+	if (new->i_state & I_NEW) {
+		dprintk("%s: filling VFS inode: %lu/%llu.\n",
+				__func__, new->i_ino, info->ino);
+		pohmelfs_fill_inode(new, info);
+
+		if (S_ISDIR(info->mode)) {
+			struct qstr s;
+
+			s.name = ".";
+			s.len = 1;
+			s.hash = jhash(s.name, s.len, 0);
+
+			err = pohmelfs_add_dir(psb, npi, npi, &s, info->mode, 0);
+			if (err)
+				goto err_out_put;
+
+			s.name = "..";
+			s.len = 2;
+			s.hash = jhash(s.name, s.len, 0);
+
+			err = pohmelfs_add_dir(psb, npi, (parent)?parent:npi, &s,
+					(parent)?parent->vfs_inode.i_mode:npi->vfs_inode.i_mode, 0);
+			if (err)
+				goto err_out_put;
+		}
+	}
+
+	if (str) {
+		if (parent) {
+			err = pohmelfs_add_dir(psb, parent, npi, str, info->mode, link);
+
+			dprintk("%s: %s inserted name: '%s', new_offset: %llu, ino: %llu, parent: %llu.\n",
+					__func__, (err)?"unsuccessfully":"successfully",
+					str->name, parent->total_len, info->ino, parent->ino);
+
+			if (err && err != -EEXIST)
+				goto err_out_put;
+		}
+	}
+
+	if (new->i_state & I_NEW) {
+		if (parent)
+			mark_inode_dirty(&parent->vfs_inode);
+		mark_inode_dirty(new);
+	}
+
+	set_bit(NETFS_INODE_OWNED, &npi->state);
+	npi->lock_type = POHMELFS_WRITE_LOCK;
+	unlock_new_inode(new);
+
+	return npi;
+
+err_out_put:
+	printk("%s: putting inode: %p, npi: %p, error: %d.\n", __func__, new, npi, err);
+	iput(new);
+err_out_exit:
+	return ERR_PTR(err);
+}
+
+static int pohmelfs_remote_sync_complete(struct page **pages, unsigned int page_num,
+		void *private, int err)
+{
+	struct pohmelfs_inode *pi = private;
+	struct pohmelfs_sb *psb = POHMELFS_SB(pi->vfs_inode.i_sb);
+
+	dprintk("%s: ino: %llu, err: %d.\n", __func__, pi->ino, err);
+
+	if (err)
+		pi->error = err;
+	wake_up(&psb->wait);
+	pohmelfs_put_inode(pi);
+
+	return err;
+}
+
+/*
+ * Receive directory content from the server.
+ * This should be only done for objects, which were not created locally,
+ * and which were not synced previously.
+ */
+static int pohmelfs_sync_remote_dir(struct pohmelfs_inode *pi)
+{
+	struct inode *inode = &pi->vfs_inode;
+	struct pohmelfs_sb *psb = POHMELFS_SB(inode->i_sb);
+	long ret = msecs_to_jiffies(25000);
+	int err;
+
+	dprintk("%s: dir: %llu, state: %lx: remote_synced: %d.\n",
+		__func__, pi->ino, pi->state, test_bit(NETFS_INODE_REMOTE_SYNCED, &pi->state));
+
+	if (test_bit(NETFS_INODE_REMOTE_DIR_SYNCED, &pi->state))
+		return 0;
+	
+	if (!igrab(inode)) {
+		err = -ENOENT;
+		goto err_out_exit;
+	}
+
+	err = pohmelfs_meta_command(pi, NETFS_READDIR, NETFS_TRANS_SINGLE_DST,
+			pohmelfs_remote_sync_complete, pi, 0);
+	if (err)
+		goto err_out_exit;
+
+	pi->error = 0;
+	ret = wait_event_interruptible_timeout(psb->wait,
+			test_bit(NETFS_INODE_REMOTE_DIR_SYNCED, &pi->state) || pi->error, ret);
+	dprintk("%s: awake dir: %llu, ret: %ld, err: %d.\n", __func__, pi->ino, ret, pi->error);
+	if (ret <= 0) {
+		err = -ETIMEDOUT;
+		goto err_out_exit;
+	}
+
+	if (pi->error)
+		return pi->error;
+
+	return 0;
+
+err_out_exit:
+	clear_bit(NETFS_INODE_REMOTE_SYNCED, &pi->state);
+
+	return err;
+}
+
+static int pohmelfs_dir_open(struct inode *inode, struct file *file)
+{
+	file->private_data = NULL;
+	return 0;
+}
+
+/*
+ * VFS readdir callback. Syncs directory content from server if needed,
+ * and provides direntry info to the userspace.
+ */
+static int pohmelfs_readdir(struct file *file, void *dirent, filldir_t filldir)
+{
+	struct inode *inode = file->f_path.dentry->d_inode;
+	struct pohmelfs_inode *pi = POHMELFS_I(inode);
+	struct pohmelfs_name *n;
+	struct rb_node *rb_node;
+	int err = 0, mode;
+	u64 len;
+
+	dprintk("%s: parent: %llu, fpos: %llu, hash: %08lx.\n",
+			__func__, pi->ino, (u64)file->f_pos,
+			(unsigned long)file->private_data);
+
+	err = pohmelfs_data_lock(pi, 0, ~0, POHMELFS_READ_LOCK);
+	if (err)
+		return err;
+
+	err = pohmelfs_sync_remote_dir(pi);
+	if (err)
+		return err;
+
+	if (file->private_data && (file->private_data == (void *)(unsigned long)file->f_pos))
+		return 0;
+
+	mutex_lock(&pi->offset_lock);
+	n = pohmelfs_search_hash_unprecise(pi, (unsigned long)file->private_data);
+
+	while (n) {
+		mode = (n->mode >> 12) & 15;
+
+		dprintk("%s: offset: %llu, parent ino: %llu, name: '%s', len: %u, ino: %llu, "
+				"mode: %o/%o, fpos: %llu, hash: %08x.\n",
+				__func__, file->f_pos, pi->ino, n->data, n->len,
+				n->ino, n->mode, mode, file->f_pos, n->hash);
+
+		file->private_data = (void *)n->hash;
+
+		len = n->len;
+		err = filldir(dirent, n->data, n->len, file->f_pos, n->ino, mode);
+
+		if (err < 0) {
+			dprintk("%s: err: %d.\n", __func__, err);
+			err = 0;
+			break;
+		}
+
+		file->f_pos += len;
+
+		rb_node = rb_next(&n->hash_node);
+
+		if (!rb_node || (rb_node == &n->hash_node)) {
+			file->private_data = (void *)(unsigned long)file->f_pos;
+			break;
+		}
+
+		n = rb_entry(rb_node, struct pohmelfs_name, hash_node);
+	}
+	mutex_unlock(&pi->offset_lock);
+
+	return err;
+}
+
+static loff_t pohmelfs_dir_lseek(struct file *file, loff_t offset, int origin)
+{
+	file->f_pos = offset;
+	file->private_data = NULL;
+	return offset;
+}
+
+const struct file_operations pohmelfs_dir_fops = {
+	.open = pohmelfs_dir_open,
+	.read = generic_read_dir,
+	.llseek = pohmelfs_dir_lseek,
+	.readdir = pohmelfs_readdir,
+};
+
+/*
+ * Lookup single object on server.
+ */
+static int pohmelfs_lookup_single(struct pohmelfs_inode *parent,
+		struct qstr *str, u64 ino)
+{
+	struct pohmelfs_sb *psb = POHMELFS_SB(parent->vfs_inode.i_sb);
+	long ret = msecs_to_jiffies(5000);
+	int err;
+
+	set_bit(NETFS_COMMAND_PENDING, &parent->state);
+	err = pohmelfs_meta_command_data(parent, parent->ino, NETFS_LOOKUP,
+			(char *)str->name, NETFS_TRANS_SINGLE_DST, NULL, NULL, ino);
+	if (err)
+		goto err_out_exit;
+
+	err = 0;
+	ret = wait_event_interruptible_timeout(psb->wait,
+			!test_bit(NETFS_COMMAND_PENDING, &parent->state), ret);
+	if (ret == 0)
+		err = -ETIMEDOUT;
+	else if (signal_pending(current))
+		err = -EINTR;
+
+	if (err)
+		goto err_out_exit;
+
+	return 0;
+
+err_out_exit:
+	clear_bit(NETFS_COMMAND_PENDING, &parent->state);
+
+	printk("%s: failed: parent: %llu, ino: %llu, name: '%s', err: %d.\n",
+			__func__, parent->ino, ino, str->name, err);
+
+	return err;
+}
+
+/*
+ * VFS lookup callback.
+ * We first try to get inode number from local name cache, if we have one,
+ * then inode can be found in inode cache. If there is no inode or no object in
+ * local cache, try to lookup it on server. This only should be done for directories,
+ * which were not created locally, otherwise remote server does not know about dir at all,
+ * so no need to try to know that.
+ */
+struct dentry *pohmelfs_lookup(struct inode *dir, struct dentry *dentry, struct nameidata *nd)
+{
+	struct pohmelfs_inode *parent = POHMELFS_I(dir);
+	struct pohmelfs_name *n;
+	struct inode *inode = NULL;
+	unsigned long ino = 0;
+	int err, lock_type = POHMELFS_READ_LOCK, need_lock;
+	struct qstr str = dentry->d_name;
+
+	if ((nd->intent.open.flags & O_ACCMODE) > 1)
+		lock_type = POHMELFS_WRITE_LOCK;
+
+	need_lock = pohmelfs_need_lock(parent, lock_type);
+
+	err = pohmelfs_data_lock(parent, 0, ~0, lock_type);
+	if (err)
+		goto out;
+
+	str.hash = jhash(dentry->d_name.name, dentry->d_name.len, 0);
+
+	mutex_lock(&parent->offset_lock);
+	n = pohmelfs_search_hash(parent, str.hash);
+	if (n)
+		ino = n->ino;
+	mutex_unlock(&parent->offset_lock);
+
+	dprintk("%s: 1 ino: %lu, inode: %p, name: '%s', hash: %x, parent_state: %lx.\n",
+			__func__, ino, inode, str.name, str.hash, parent->state);
+
+	if (ino) {
+		inode = ilookup(dir->i_sb, ino);
+		if (inode)
+			goto out;
+	}
+
+	dprintk("%s: dir: %p, dir_ino: %llu, name: '%s', len: %u, dir_state: %lx, ino: %lu.\n",
+			__func__, dir, parent->ino,
+			str.name, str.len, parent->state, ino);
+
+	if (!ino) {
+		if (!need_lock)
+			goto out;
+	}
+
+	err = pohmelfs_lookup_single(parent, &str, ino);
+	if (err)
+		goto out;
+
+	if (!ino) {
+		mutex_lock(&parent->offset_lock);
+		n = pohmelfs_search_hash(parent, str.hash);
+		if (n)
+			ino = n->ino;
+		mutex_unlock(&parent->offset_lock);
+	}
+
+	if (ino) {
+		inode = ilookup(dir->i_sb, ino);
+		printk("%s: second lookup ino: %lu, inode: %p, name: '%s', hash: %x.\n",
+				__func__, ino, inode, str.name, str.hash);
+		if (!inode) {
+			printk("%s: No inode for ino: %lu, name: '%s', hash: %x.\n",
+				__func__, ino, str.name, str.hash);
+			//return NULL;
+			return ERR_PTR(-EACCES);
+		}
+	} else {
+		printk("%s: No inode number : name: '%s', hash: %x.\n",
+			__func__, str.name, str.hash);
+	}
+out:
+	return d_splice_alias(inode, dentry);
+}
+
+/*
+ * Create new object in local cache. Object will be synced to server
+ * during writeback for given inode.
+ */
+struct pohmelfs_inode *pohmelfs_create_entry_local(struct pohmelfs_sb *psb,
+	struct pohmelfs_inode *parent, struct qstr *str, u64 start, int mode)
+{
+	struct pohmelfs_inode *npi;
+	int err = -ENOMEM;
+	struct netfs_inode_info info;
+
+	dprintk("%s: name: '%s', mode: %o, start: %llu.\n",
+			__func__, str->name, mode, start);
+
+	info.mode = mode;
+	info.ino = start;
+
+	if (!start)
+		info.ino = pohmelfs_new_ino(psb);
+
+	info.nlink = S_ISDIR(mode)?2:1;
+	info.uid = current_fsuid();
+	info.gid = current_fsgid();
+	info.size = 0;
+	info.blocksize = 512;
+	info.blocks = 0;
+	info.rdev = 0;
+	info.version = 0;
+
+	npi = pohmelfs_new_inode(psb, parent, str, &info, !!start);
+	if (IS_ERR(npi)) {
+		err = PTR_ERR(npi);
+		goto err_out_unlock;
+	}
+
+	return npi;
+
+err_out_unlock:
+	dprintk("%s: err: %d.\n", __func__, err);
+	return ERR_PTR(err);
+}
+
+/*
+ * Create local object and bind it to dentry.
+ */
+static int pohmelfs_create_entry(struct inode *dir, struct dentry *dentry, u64 start, int mode)
+{
+	struct pohmelfs_sb *psb = POHMELFS_SB(dir->i_sb);
+	struct pohmelfs_inode *npi, *parent;
+	struct qstr str = dentry->d_name;
+	int err;
+
+	parent = POHMELFS_I(dir);
+
+	err = pohmelfs_data_lock(parent, 0, ~0, POHMELFS_WRITE_LOCK);
+	if (err)
+		return err;
+
+	str.hash = jhash(dentry->d_name.name, dentry->d_name.len, 0);
+
+	npi = pohmelfs_create_entry_local(psb, parent, &str, start, mode);
+	if (IS_ERR(npi))
+		return PTR_ERR(npi);
+
+	d_instantiate(dentry, &npi->vfs_inode);
+
+	dprintk("%s: parent: %llu, inode: %llu, name: '%s', parent_nlink: %d, nlink: %d.\n",
+			__func__, parent->ino, npi->ino, dentry->d_name.name,
+			(signed)dir->i_nlink, (signed)npi->vfs_inode.i_nlink);
+
+	return 0;
+}
+
+/*
+ * VFS create and mkdir callbacks.
+ */
+static int pohmelfs_create(struct inode *dir, struct dentry *dentry, int mode,
+		struct nameidata *nd)
+{
+	return pohmelfs_create_entry(dir, dentry, 0, mode);
+}
+
+static int pohmelfs_mkdir(struct inode *dir, struct dentry *dentry, int mode)
+{
+	int err;
+
+	inode_inc_link_count(dir);
+	err = pohmelfs_create_entry(dir, dentry, 0, mode | S_IFDIR);
+	if (err)
+		inode_dec_link_count(dir);
+
+	return err;
+}
+
+static int pohmelfs_remove_entry(struct inode *dir, struct dentry *dentry)
+{
+	struct pohmelfs_sb *psb = POHMELFS_SB(dir->i_sb);
+	struct inode *inode = dentry->d_inode;
+	struct pohmelfs_inode *parent = POHMELFS_I(dir), *pi = POHMELFS_I(inode);
+	struct pohmelfs_name *n;
+	int err = -ENOENT;
+	struct qstr str = dentry->d_name;
+
+	err = pohmelfs_data_lock(parent, 0, ~0, POHMELFS_WRITE_LOCK);
+	if (err)
+		return err;
+
+	str.hash = jhash(dentry->d_name.name, dentry->d_name.len, 0);
+
+	dprintk("%s: dir_ino: %llu, inode: %llu, name: '%s', nlink: %d.\n",
+			__func__, parent->ino, pi->ino,
+			str.name, (signed)inode->i_nlink);
+
+	BUG_ON(!inode);
+
+	mutex_lock(&parent->offset_lock);
+	n = pohmelfs_search_hash(parent, str.hash);
+	if (n) {
+		pohmelfs_fix_offset(parent, n);
+		if (test_bit(NETFS_INODE_REMOTE_SYNCED, &pi->state)) {
+			pohmelfs_remove_child(pi, n);
+		}
+		pohmelfs_name_free(parent, n);
+		err = 0;
+	}
+	mutex_unlock(&parent->offset_lock);
+
+	if (!err) {
+		psb->avail_size += inode->i_size;
+
+		pohmelfs_inode_del_inode(psb, pi);
+
+		mark_inode_dirty(dir);
+
+		inode->i_ctime = dir->i_ctime;
+		if (inode->i_nlink)
+			inode_dec_link_count(inode);
+	}
+	dprintk("%s: inode: %p, lock: %ld, unhashed: %d.\n",
+		__func__, pi, inode->i_state & I_LOCK, hlist_unhashed(&inode->i_hash));
+
+	return err;
+}
+
+/*
+ * Unlink and rmdir VFS callbacks.
+ */
+static int pohmelfs_unlink(struct inode *dir, struct dentry *dentry)
+{
+	return pohmelfs_remove_entry(dir, dentry);
+}
+
+static int pohmelfs_rmdir(struct inode *dir, struct dentry *dentry)
+{
+	int err;
+	struct inode *inode = dentry->d_inode;
+
+	dprintk("%s: parent: %llu, inode: %llu, name: '%s', parent_nlink: %d, nlink: %d.\n",
+			__func__, POHMELFS_I(dir)->ino, POHMELFS_I(inode)->ino,
+			dentry->d_name.name, (signed)dir->i_nlink, (signed)inode->i_nlink);
+
+	err = pohmelfs_remove_entry(dir, dentry);
+	if (!err) {
+		inode_dec_link_count(dir);
+		inode_dec_link_count(inode);
+	}
+
+	return err;
+}
+
+/*
+ * Link creation is synchronous.
+ * I'm lazy.
+ * Earth is somewhat round.
+ */
+static int pohmelfs_create_link(struct pohmelfs_inode *parent, struct qstr *obj,
+		struct pohmelfs_inode *target, struct qstr *tstr)
+{
+	struct super_block *sb = parent->vfs_inode.i_sb;
+	struct pohmelfs_sb *psb = POHMELFS_SB(sb);
+	struct netfs_cmd *cmd;
+	struct netfs_trans *t;
+	void *data;
+	int err, parent_len, target_len = 0, cur_len, path_size = 0;
+
+	err = pohmelfs_data_lock(parent, 0, ~0, POHMELFS_WRITE_LOCK);
+	if (err)
+		return err;
+
+	err = sb->s_op->write_inode(&parent->vfs_inode, 0);
+	if (err)
+		goto err_out_exit;
+
+	if (tstr)
+		target_len = tstr->len;
+
+	parent_len = pohmelfs_path_length(parent);
+	if (target)
+		target_len += pohmelfs_path_length(target);
+
+	if (parent_len < 0) {
+		err = parent_len;
+		goto err_out_exit;
+	}
+	
+	if (target_len < 0) {
+		err = target_len;
+		goto err_out_exit;
+	}
+
+	t = netfs_trans_alloc(psb, parent_len + target_len + obj->len + 2, 0, 0);
+	if (!t) {
+		err = -ENOMEM;
+		goto err_out_exit;
+	}
+	cur_len = netfs_trans_cur_len(t);
+
+	cmd = netfs_trans_current(t);
+	if (IS_ERR(cmd)) {
+		err = PTR_ERR(cmd);
+		goto err_out_free;
+	}
+
+	data = (void *)(cmd + 1);
+	cur_len -= sizeof(struct netfs_cmd);
+
+	err = pohmelfs_construct_path_string(parent, data, parent_len);
+	if (err > 0) {
+		/* Do not place null-byte before the slash */
+		path_size = err - 1;
+		cur_len -= path_size;
+
+		err = snprintf(data + path_size, cur_len, "/%s|", obj->name);
+
+		path_size += err;
+		cur_len -= err;
+
+		cmd->ext = path_size - 1; /* No | symbol */
+
+		if (target) {
+			err = pohmelfs_construct_path_string(target, data + path_size, target_len);
+			if (err > 0) {
+				path_size += err;
+				cur_len -= err;
+			}
+		}
+	}
+
+	if (err < 0)
+		goto err_out_free;
+
+	cmd->start = 0;
+
+	if (!target && tstr) {
+		if (tstr->len > cur_len - 1) {
+			err = -ENAMETOOLONG;
+			goto err_out_free;
+		}
+
+		err = snprintf(data + path_size, cur_len, "%s", tstr->name) + 1; /* 0-byte */
+		path_size += err;
+		cur_len -= err;
+		cmd->start = 1;
+	}
+
+	dprintk("%s: parent: %llu, obj: '%s', target_inode: %llu, target_str: '%s', full: '%s'.\n",
+			__func__, parent->ino, obj->name, (target)?target->ino:0, (tstr)?tstr->name:NULL,
+			(char *)data);
+
+	cmd->cmd = NETFS_LINK;
+	cmd->size = path_size;
+	cmd->id = parent->ino;
+
+	netfs_convert_cmd(cmd);
+	
+	netfs_trans_update(cmd, t, path_size);
+
+	err = netfs_trans_finish(t, psb);
+	if (err)
+		goto err_out_exit;
+
+	return 0;
+
+err_out_free:
+	t->result = err;
+	netfs_trans_put(t);
+err_out_exit:
+	return err;
+}
+
+/*
+ *  VFS hard and soft link callbacks.
+ */
+static int pohmelfs_link(struct dentry *old_dentry, struct inode *dir,
+	struct dentry *dentry)
+{
+	struct inode *inode = old_dentry->d_inode;
+	struct pohmelfs_inode *pi = POHMELFS_I(inode);
+	int err;
+	struct qstr str = dentry->d_name;
+
+	str.hash = jhash(dentry->d_name.name, dentry->d_name.len, 0);
+
+	err = inode->i_sb->s_op->write_inode(inode, 0);
+	if (err)
+		return err;
+
+	err = pohmelfs_create_link(POHMELFS_I(dir), &str, pi, NULL);
+	if (err)
+		return err;
+
+	return pohmelfs_create_entry(dir, dentry, pi->ino, inode->i_mode);
+}
+
+static int pohmelfs_symlink(struct inode *dir, struct dentry *dentry, const char *symname)
+{
+	struct qstr sym_str;
+	struct qstr str = dentry->d_name;
+	struct inode *inode;
+	int err;
+
+	str.hash = jhash(dentry->d_name.name, dentry->d_name.len, 0);
+
+	sym_str.name = symname;
+	sym_str.len = strlen(symname);
+
+	err = pohmelfs_create_link(POHMELFS_I(dir), &str, NULL, &sym_str);
+	if (err)
+		goto err_out_exit;
+
+	err = pohmelfs_create_entry(dir, dentry, 0, S_IFLNK | S_IRWXU | S_IRWXG | S_IRWXO);
+	if (err)
+		goto err_out_exit;
+
+	inode = dentry->d_inode;
+
+	err = page_symlink(inode, symname, sym_str.len + 1);
+	if (err)
+		goto err_out_put;
+
+	return 0;
+
+err_out_put:
+	iput(inode);
+err_out_exit:
+	return err;
+}
+
+static int pohmelfs_send_rename(struct pohmelfs_inode *pi, struct pohmelfs_inode *parent,
+		struct qstr *str)
+{
+	int path_len, err, total_len = 0, inode_len, parent_len;
+	char *path;
+	struct netfs_trans *t;
+	struct netfs_cmd *cmd;
+	struct pohmelfs_sb *psb = POHMELFS_SB(pi->vfs_inode.i_sb);
+
+	parent_len = pohmelfs_path_length(parent);
+	inode_len = pohmelfs_path_length(pi);
+
+	if (parent_len < 0 || inode_len < 0)
+		return -EINVAL;
+
+	path_len = parent_len + inode_len + str->len + 3;
+
+	t = netfs_trans_alloc(psb, path_len, 0, 0);
+	if (!t)
+		return -ENOMEM;
+
+	cmd = netfs_trans_current(t);
+	path = (char *)(cmd + 1);
+
+	err = pohmelfs_construct_path_string(pi, path, inode_len);
+	if (err < 0)
+		goto err_out_unlock;
+
+	cmd->ext = err;
+
+	path += err;
+	total_len += err;
+	path_len -= err;
+
+	*path = '|';
+	path++;
+	total_len++;
+	path_len--;
+
+	err = pohmelfs_construct_path_string(parent, path, parent_len);
+	if (err < 0)
+		goto err_out_unlock;
+
+	/*
+	 * Do not place a null-byte before the final slash and the name.
+	 */
+	err--;
+	path += err;
+	total_len += err;
+	path_len -= err;
+
+	err = snprintf(path, path_len - 1, "/%s", str->name);
+	
+	total_len += err + 1; /* 0 symbol */
+	path_len -= err + 1;
+
+	cmd->cmd = NETFS_RENAME;
+	cmd->id = pi->ino;
+	cmd->start = parent->ino;
+	cmd->size = total_len;
+
+	netfs_convert_cmd(cmd);
+	
+	netfs_trans_update(cmd, t, total_len);
+
+	return netfs_trans_finish(t, psb);
+
+err_out_unlock:
+	netfs_trans_free(t);
+	return err;
+}
+
+static int pohmelfs_rename(struct inode *old_dir, struct dentry *old_dentry,
+			struct inode *new_dir, struct dentry *new_dentry)
+{
+	struct inode *inode = old_dentry->d_inode;
+	struct pohmelfs_inode *old_parent, *pi, *new_parent;
+	struct qstr str = new_dentry->d_name;
+	struct pohmelfs_name *n;
+	unsigned int old_hash;
+	int err = -ENOENT;
+
+	pi = POHMELFS_I(inode);
+	old_parent = POHMELFS_I(old_dir);
+
+	if (new_dir) {
+		new_dir->i_sb->s_op->write_inode(new_dir, 0);
+	}
+
+	old_hash = jhash(old_dentry->d_name.name, old_dentry->d_name.len, 0);
+	str.hash = jhash(new_dentry->d_name.name, new_dentry->d_name.len, 0);
+
+	str.len = new_dentry->d_name.len;
+	str.name = new_dentry->d_name.name;
+	str.hash = jhash(new_dentry->d_name.name, new_dentry->d_name.len, 0);
+
+	if (new_dir) {
+		new_parent = POHMELFS_I(new_dir);
+		err = -ENOTEMPTY;
+
+		if (S_ISDIR(inode->i_mode) &&
+				new_parent->total_len <= 3)
+			goto err_out_exit;
+	} else {
+		new_parent = old_parent;
+	}
+
+	dprintk("%s: ino: %llu, parent: %llu, name: '%s' -> parent: %llu, name: '%s', i_size: %llu.\n",
+			__func__, pi->ino, old_parent->ino, old_dentry->d_name.name,
+			new_parent->ino, new_dentry->d_name.name, inode->i_size);
+
+	if (test_bit(NETFS_INODE_REMOTE_SYNCED, &pi->state) &&
+			test_bit(NETFS_INODE_OWNED, &pi->state)) {
+		err = pohmelfs_send_rename(pi, new_parent, &str);
+		if (err)
+			goto err_out_exit;
+	}
+
+	n = pohmelfs_name_alloc(str.len + 1);
+	if (!n)
+		goto err_out_exit;
+
+	mutex_lock(&new_parent->offset_lock);
+	n->ino = pi->ino;
+	n->mode = inode->i_mode;
+	n->len = str.len;
+	n->hash = str.hash;
+	sprintf(n->data, "%s", str.name);
+
+	err = pohmelfs_insert_name(new_parent, n);
+	mutex_unlock(&new_parent->offset_lock);
+
+	if (err)
+		goto err_out_exit;
+
+	mutex_lock(&old_parent->offset_lock);
+	n = pohmelfs_search_hash(old_parent, old_hash);
+	if (n)
+		pohmelfs_name_del(old_parent, n);
+	mutex_unlock(&old_parent->offset_lock);
+
+	mark_inode_dirty(inode);
+	mark_inode_dirty(&new_parent->vfs_inode);
+
+	WARN_ON_ONCE(list_empty(&inode->i_dentry));
+
+	return 0;
+
+err_out_exit:
+
+	clear_bit(NETFS_INODE_REMOTE_SYNCED, &pi->state);
+
+	mutex_unlock(&inode->i_mutex);
+	return err;
+}
+
+/*
+ * POHMELFS directory inode operations.
+ */
+const struct inode_operations pohmelfs_dir_inode_ops = {
+	.link		= pohmelfs_link,
+	.symlink	= pohmelfs_symlink,
+	.unlink		= pohmelfs_unlink,
+	.mkdir		= pohmelfs_mkdir,
+	.rmdir		= pohmelfs_rmdir,
+	.create		= pohmelfs_create,
+	.lookup 	= pohmelfs_lookup,
+	.setattr	= pohmelfs_setattr,
+	.rename		= pohmelfs_rename,
+};
diff --git a/drivers/staging/pohmelfs/path_entry.c b/drivers/staging/pohmelfs/path_entry.c
new file mode 100644
index 0000000..b62374b
--- /dev/null
+++ b/drivers/staging/pohmelfs/path_entry.c
@@ -0,0 +1,114 @@
+/*
+ * 2007+ Copyright (c) Evgeniy Polyakov <zbr@ioremap.net>
+ * All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/module.h>
+#include <linux/slab.h>
+#include <linux/fs.h>
+#include <linux/ktime.h>
+#include <linux/fs.h>
+#include <linux/pagemap.h>
+#include <linux/writeback.h>
+#include <linux/mount.h>
+#include <linux/mm.h>
+
+#include "netfs.h"
+
+#define UNHASHED_OBSCURE_STRING_SIZE		sizeof(" (deleted)")
+
+/*
+ * Create path from root for given inode.
+ * Path is formed as set of stuctures, containing name of the object
+ * and its inode data (mode, permissions and so on).
+ */
+int pohmelfs_construct_path_string(struct pohmelfs_inode *pi, void *data, int len)
+{
+	struct path path;
+	struct dentry *d;
+	char *ptr;
+	int err = 0, strlen, reduce = 0;
+
+	d = d_find_alias(&pi->vfs_inode);
+	if (!d) {
+		printk("%s: no alias, list_empty: %d.\n", __func__, list_empty(&pi->vfs_inode.i_dentry));
+		return -ENOENT;
+	}
+
+	read_lock(&current->fs->lock);
+	path.mnt = mntget(current->fs->root.mnt);
+	read_unlock(&current->fs->lock);
+
+	path.dentry = d;
+	
+	if (!IS_ROOT(d) && d_unhashed(d))
+		reduce = 1;
+
+	ptr = d_path(&path, data, len);
+	if (IS_ERR(ptr)) {
+		err = PTR_ERR(ptr);
+		goto out;
+	}
+
+	if (reduce && len >= UNHASHED_OBSCURE_STRING_SIZE) {
+		char *end = data + len - UNHASHED_OBSCURE_STRING_SIZE;
+		*end = '\0';
+	}
+
+	strlen = len - (ptr - (char *)data);
+	memmove(data, ptr, strlen);
+	ptr = data;
+
+	err = strlen;
+
+	dprintk("%s: dname: '%s', len: %u, maxlen: %u, name: '%s', strlen: %d.\n",
+			__func__, d->d_name.name, d->d_name.len, len, ptr, strlen);
+
+out:
+	dput(d);
+	mntput(path.mnt);
+
+	return err;
+}
+
+int pohmelfs_path_length(struct pohmelfs_inode *pi)
+{
+	struct dentry *d, *root, *first;
+	int len = 1; /* Root slash */
+
+	first = d = d_find_alias(&pi->vfs_inode);
+	if (!d) {
+		dprintk("%s: ino: %llu, mode: %o.\n", __func__, pi->ino, pi->vfs_inode.i_mode);
+		return -ENOENT;
+	}
+
+	read_lock(&current->fs->lock);
+	root = dget(current->fs->root.dentry);
+	read_unlock(&current->fs->lock);
+
+	spin_lock(&dcache_lock);
+
+	if (!IS_ROOT(d) && d_unhashed(d))
+		len += UNHASHED_OBSCURE_STRING_SIZE; /* Obscure " (deleted)" string */
+
+	while (d && d != root && !IS_ROOT(d)) {
+		len += d->d_name.len + 1; /* Plus slash */
+		d = d->d_parent;
+	}
+	spin_unlock(&dcache_lock);
+
+	dput(root);
+	dput(first);
+
+	return len + 1; /* Including zero-byte */
+}