Merge tag 'for-f2fs-3.17' of git://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk/f2fs

Pull f2fs updates from Jaegeuk Kim:
 "This series includes patches to:
   - add nobarrier mount option
   - support tmpfile and rename2
   - enhance the fdatasync behavior
   - fix the error path
   - fix the recovery routine
   - refactor a part of the checkpoint procedure
   - reduce some lock contentions"

* tag 'for-f2fs-3.17' of git://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk/f2fs: (40 commits)
  f2fs: use for_each_set_bit to simplify the code
  f2fs: add f2fs_balance_fs for expand_inode_data
  f2fs: invalidate xattr node page when evict inode
  f2fs: avoid skipping recover_inline_xattr after recover_inline_data
  f2fs: add tracepoint for f2fs_direct_IO
  f2fs: reduce competition among node page writes
  f2fs: fix coding style
  f2fs: remove redundant lines in allocate_data_block
  f2fs: add tracepoint for f2fs_issue_flush
  f2fs: avoid retrying wrong recovery routine when error was occurred
  f2fs: test before set/clear bits
  f2fs: fix wrong condition for unlikely
  f2fs: enable in-place-update for fdatasync
  f2fs: skip unnecessary data writes during fsync
  f2fs: add info of appended or updated data writes
  f2fs: use radix_tree for ino management
  f2fs: add infra for ino management
  f2fs: punch the core function for inode management
  f2fs: add nobarrier mount option
  f2fs: fix to put root inode in error path of fill_super
  ...

18 files changed, 791 insertions, 292 deletions

diff --git a/fs/f2fs/acl.c b/fs/f2fs/acl.c
index dbe2141d10ad..83b9b5a8d112 100644
--- a/fs/f2fs/acl.c
+++ b/fs/f2fs/acl.c
@@ -203,12 +203,6 @@ static int __f2fs_set_acl(struct inode *inode, int type,
 	size_t size = 0;
 	int error;
 
-	if (acl) {
-		error = posix_acl_valid(acl);
-		if (error < 0)
-			return error;
-	}
-
 	switch (type) {
 	case ACL_TYPE_ACCESS:
 		name_index = F2FS_XATTR_INDEX_POSIX_ACL_ACCESS;
diff --git a/fs/f2fs/checkpoint.c b/fs/f2fs/checkpoint.c
index 0b4710c1d370..6aeed5bada52 100644
--- a/fs/f2fs/checkpoint.c
+++ b/fs/f2fs/checkpoint.c
@@ -22,7 +22,7 @@
 #include "segment.h"
 #include <trace/events/f2fs.h>
 
-static struct kmem_cache *orphan_entry_slab;
+static struct kmem_cache *ino_entry_slab;
 static struct kmem_cache *inode_entry_slab;
 
 /*
@@ -282,72 +282,120 @@ const struct address_space_operations f2fs_meta_aops = {
 	.set_page_dirty	= f2fs_set_meta_page_dirty,
 };
 
+static void __add_ino_entry(struct f2fs_sb_info *sbi, nid_t ino, int type)
+{
+	struct ino_entry *e;
+retry:
+	spin_lock(&sbi->ino_lock[type]);
+
+	e = radix_tree_lookup(&sbi->ino_root[type], ino);
+	if (!e) {
+		e = kmem_cache_alloc(ino_entry_slab, GFP_ATOMIC);
+		if (!e) {
+			spin_unlock(&sbi->ino_lock[type]);
+			goto retry;
+		}
+		if (radix_tree_insert(&sbi->ino_root[type], ino, e)) {
+			spin_unlock(&sbi->ino_lock[type]);
+			kmem_cache_free(ino_entry_slab, e);
+			goto retry;
+		}
+		memset(e, 0, sizeof(struct ino_entry));
+		e->ino = ino;
+
+		list_add_tail(&e->list, &sbi->ino_list[type]);
+	}
+	spin_unlock(&sbi->ino_lock[type]);
+}
+
+static void __remove_ino_entry(struct f2fs_sb_info *sbi, nid_t ino, int type)
+{
+	struct ino_entry *e;
+
+	spin_lock(&sbi->ino_lock[type]);
+	e = radix_tree_lookup(&sbi->ino_root[type], ino);
+	if (e) {
+		list_del(&e->list);
+		radix_tree_delete(&sbi->ino_root[type], ino);
+		if (type == ORPHAN_INO)
+			sbi->n_orphans--;
+		spin_unlock(&sbi->ino_lock[type]);
+		kmem_cache_free(ino_entry_slab, e);
+		return;
+	}
+	spin_unlock(&sbi->ino_lock[type]);
+}
+
+void add_dirty_inode(struct f2fs_sb_info *sbi, nid_t ino, int type)
+{
+	/* add new dirty ino entry into list */
+	__add_ino_entry(sbi, ino, type);
+}
+
+void remove_dirty_inode(struct f2fs_sb_info *sbi, nid_t ino, int type)
+{
+	/* remove dirty ino entry from list */
+	__remove_ino_entry(sbi, ino, type);
+}
+
+/* mode should be APPEND_INO or UPDATE_INO */
+bool exist_written_data(struct f2fs_sb_info *sbi, nid_t ino, int mode)
+{
+	struct ino_entry *e;
+	spin_lock(&sbi->ino_lock[mode]);
+	e = radix_tree_lookup(&sbi->ino_root[mode], ino);
+	spin_unlock(&sbi->ino_lock[mode]);
+	return e ? true : false;
+}
+
+static void release_dirty_inode(struct f2fs_sb_info *sbi)
+{
+	struct ino_entry *e, *tmp;
+	int i;
+
+	for (i = APPEND_INO; i <= UPDATE_INO; i++) {
+		spin_lock(&sbi->ino_lock[i]);
+		list_for_each_entry_safe(e, tmp, &sbi->ino_list[i], list) {
+			list_del(&e->list);
+			radix_tree_delete(&sbi->ino_root[i], e->ino);
+			kmem_cache_free(ino_entry_slab, e);
+		}
+		spin_unlock(&sbi->ino_lock[i]);
+	}
+}
+
 int acquire_orphan_inode(struct f2fs_sb_info *sbi)
 {
 	int err = 0;
 
-	spin_lock(&sbi->orphan_inode_lock);
+	spin_lock(&sbi->ino_lock[ORPHAN_INO]);
 	if (unlikely(sbi->n_orphans >= sbi->max_orphans))
 		err = -ENOSPC;
 	else
 		sbi->n_orphans++;
-	spin_unlock(&sbi->orphan_inode_lock);
+	spin_unlock(&sbi->ino_lock[ORPHAN_INO]);
 
 	return err;
 }
 
 void release_orphan_inode(struct f2fs_sb_info *sbi)
 {
-	spin_lock(&sbi->orphan_inode_lock);
+	spin_lock(&sbi->ino_lock[ORPHAN_INO]);
 	f2fs_bug_on(sbi->n_orphans == 0);
 	sbi->n_orphans--;
-	spin_unlock(&sbi->orphan_inode_lock);
+	spin_unlock(&sbi->ino_lock[ORPHAN_INO]);
 }
 
 void add_orphan_inode(struct f2fs_sb_info *sbi, nid_t ino)
 {
-	struct list_head *head;
-	struct orphan_inode_entry *new, *orphan;
-
-	new = f2fs_kmem_cache_alloc(orphan_entry_slab, GFP_ATOMIC);
-	new->ino = ino;
-
-	spin_lock(&sbi->orphan_inode_lock);
-	head = &sbi->orphan_inode_list;
-	list_for_each_entry(orphan, head, list) {
-		if (orphan->ino == ino) {
-			spin_unlock(&sbi->orphan_inode_lock);
-			kmem_cache_free(orphan_entry_slab, new);
-			return;
-		}
-
-		if (orphan->ino > ino)
-			break;
-	}
-
-	/* add new orphan entry into list which is sorted by inode number */
-	list_add_tail(&new->list, &orphan->list);
-	spin_unlock(&sbi->orphan_inode_lock);
+	/* add new orphan ino entry into list */
+	__add_ino_entry(sbi, ino, ORPHAN_INO);
 }
 
 void remove_orphan_inode(struct f2fs_sb_info *sbi, nid_t ino)
 {
-	struct list_head *head;
-	struct orphan_inode_entry *orphan;
-
-	spin_lock(&sbi->orphan_inode_lock);
-	head = &sbi->orphan_inode_list;
-	list_for_each_entry(orphan, head, list) {
-		if (orphan->ino == ino) {
-			list_del(&orphan->list);
-			f2fs_bug_on(sbi->n_orphans == 0);
-			sbi->n_orphans--;
-			spin_unlock(&sbi->orphan_inode_lock);
-			kmem_cache_free(orphan_entry_slab, orphan);
-			return;
-		}
-	}
-	spin_unlock(&sbi->orphan_inode_lock);
+	/* remove orphan entry from orphan list */
+	__remove_ino_entry(sbi, ino, ORPHAN_INO);
 }
 
 static void recover_orphan_inode(struct f2fs_sb_info *sbi, nid_t ino)
@@ -401,14 +449,14 @@ static void write_orphan_inodes(struct f2fs_sb_info *sbi, block_t start_blk)
 	unsigned short orphan_blocks = (unsigned short)((sbi->n_orphans +
 		(F2FS_ORPHANS_PER_BLOCK - 1)) / F2FS_ORPHANS_PER_BLOCK);
 	struct page *page = NULL;
-	struct orphan_inode_entry *orphan = NULL;
+	struct ino_entry *orphan = NULL;
 
 	for (index = 0; index < orphan_blocks; index++)
 		grab_meta_page(sbi, start_blk + index);
 
 	index = 1;
-	spin_lock(&sbi->orphan_inode_lock);
-	head = &sbi->orphan_inode_list;
+	spin_lock(&sbi->ino_lock[ORPHAN_INO]);
+	head = &sbi->ino_list[ORPHAN_INO];
 
 	/* loop for each orphan inode entry and write them in Jornal block */
 	list_for_each_entry(orphan, head, list) {
@@ -448,7 +496,7 @@ static void write_orphan_inodes(struct f2fs_sb_info *sbi, block_t start_blk)
 		f2fs_put_page(page, 1);
 	}
 
-	spin_unlock(&sbi->orphan_inode_lock);
+	spin_unlock(&sbi->ino_lock[ORPHAN_INO]);
 }
 
 static struct page *validate_checkpoint(struct f2fs_sb_info *sbi,
@@ -714,10 +762,10 @@ retry_flush_dents:
 	 * until finishing nat/sit flush.
 	 */
 retry_flush_nodes:
-	mutex_lock(&sbi->node_write);
+	down_write(&sbi->node_write);
 
 	if (get_pages(sbi, F2FS_DIRTY_NODES)) {
-		mutex_unlock(&sbi->node_write);
+		up_write(&sbi->node_write);
 		sync_node_pages(sbi, 0, &wbc);
 		goto retry_flush_nodes;
 	}
@@ -726,7 +774,7 @@ retry_flush_nodes:
 
 static void unblock_operations(struct f2fs_sb_info *sbi)
 {
-	mutex_unlock(&sbi->node_write);
+	up_write(&sbi->node_write);
 	f2fs_unlock_all(sbi);
 }
 
@@ -748,6 +796,7 @@ static void wait_on_all_pages_writeback(struct f2fs_sb_info *sbi)
 static void do_checkpoint(struct f2fs_sb_info *sbi, bool is_umount)
 {
 	struct f2fs_checkpoint *ckpt = F2FS_CKPT(sbi);
+	struct curseg_info *curseg = CURSEG_I(sbi, CURSEG_WARM_NODE);
 	nid_t last_nid = 0;
 	block_t start_blk;
 	struct page *cp_page;
@@ -761,7 +810,7 @@ static void do_checkpoint(struct f2fs_sb_info *sbi, bool is_umount)
 	 * This avoids to conduct wrong roll-forward operations and uses
 	 * metapages, so should be called prior to sync_meta_pages below.
 	 */
-	discard_next_dnode(sbi);
+	discard_next_dnode(sbi, NEXT_FREE_BLKADDR(sbi, curseg));
 
 	/* Flush all the NAT/SIT pages */
 	while (get_pages(sbi, F2FS_DIRTY_META))
@@ -885,8 +934,9 @@ static void do_checkpoint(struct f2fs_sb_info *sbi, bool is_umount)
 	/* Here, we only have one bio having CP pack */
 	sync_meta_pages(sbi, META_FLUSH, LONG_MAX);
 
-	if (unlikely(!is_set_ckpt_flags(ckpt, CP_ERROR_FLAG))) {
+	if (!is_set_ckpt_flags(ckpt, CP_ERROR_FLAG)) {
 		clear_prefree_segments(sbi);
+		release_dirty_inode(sbi);
 		F2FS_RESET_SB_DIRT(sbi);
 	}
 }
@@ -932,31 +982,37 @@ void write_checkpoint(struct f2fs_sb_info *sbi, bool is_umount)
 	trace_f2fs_write_checkpoint(sbi->sb, is_umount, "finish checkpoint");
 }
 
-void init_orphan_info(struct f2fs_sb_info *sbi)
+void init_ino_entry_info(struct f2fs_sb_info *sbi)
 {
-	spin_lock_init(&sbi->orphan_inode_lock);
-	INIT_LIST_HEAD(&sbi->orphan_inode_list);
-	sbi->n_orphans = 0;
+	int i;
+
+	for (i = 0; i < MAX_INO_ENTRY; i++) {
+		INIT_RADIX_TREE(&sbi->ino_root[i], GFP_ATOMIC);
+		spin_lock_init(&sbi->ino_lock[i]);
+		INIT_LIST_HEAD(&sbi->ino_list[i]);
+	}
+
 	/*
 	 * considering 512 blocks in a segment 8 blocks are needed for cp
 	 * and log segment summaries. Remaining blocks are used to keep
 	 * orphan entries with the limitation one reserved segment
 	 * for cp pack we can have max 1020*504 orphan entries
 	 */
+	sbi->n_orphans = 0;
 	sbi->max_orphans = (sbi->blocks_per_seg - 2 - NR_CURSEG_TYPE)
 				* F2FS_ORPHANS_PER_BLOCK;
 }
 
 int __init create_checkpoint_caches(void)
 {
-	orphan_entry_slab = f2fs_kmem_cache_create("f2fs_orphan_entry",
-			sizeof(struct orphan_inode_entry));
-	if (!orphan_entry_slab)
+	ino_entry_slab = f2fs_kmem_cache_create("f2fs_ino_entry",
+			sizeof(struct ino_entry));
+	if (!ino_entry_slab)
 		return -ENOMEM;
 	inode_entry_slab = f2fs_kmem_cache_create("f2fs_dirty_dir_entry",
 			sizeof(struct dir_inode_entry));
 	if (!inode_entry_slab) {
-		kmem_cache_destroy(orphan_entry_slab);
+		kmem_cache_destroy(ino_entry_slab);
 		return -ENOMEM;
 	}
 	return 0;
@@ -964,6 +1020,6 @@ int __init create_checkpoint_caches(void)
 
 void destroy_checkpoint_caches(void)
 {
-	kmem_cache_destroy(orphan_entry_slab);
+	kmem_cache_destroy(ino_entry_slab);
 	kmem_cache_destroy(inode_entry_slab);
 }
diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c
index f8cf619edb5f..03313099c51c 100644
--- a/fs/f2fs/data.c
+++ b/fs/f2fs/data.c
@@ -139,7 +139,10 @@ void f2fs_submit_merged_bio(struct f2fs_sb_info *sbi,
 	/* change META to META_FLUSH in the checkpoint procedure */
 	if (type >= META_FLUSH) {
 		io->fio.type = META_FLUSH;
-		io->fio.rw = WRITE_FLUSH_FUA | REQ_META | REQ_PRIO;
+		if (test_opt(sbi, NOBARRIER))
+			io->fio.rw = WRITE_FLUSH | REQ_META | REQ_PRIO;
+		else
+			io->fio.rw = WRITE_FLUSH_FUA | REQ_META | REQ_PRIO;
 	}
 	__submit_merged_bio(io);
 	up_write(&io->io_rwsem);
@@ -626,8 +629,10 @@ static int __get_data_block(struct inode *inode, sector_t iblock,
 	if (check_extent_cache(inode, pgofs, bh_result))
 		goto out;
 
-	if (create)
+	if (create) {
+		f2fs_balance_fs(sbi);
 		f2fs_lock_op(sbi);
+	}
 
 	/* When reading holes, we need its node page */
 	set_new_dnode(&dn, inode, NULL, NULL, 0);
@@ -784,9 +789,11 @@ int do_write_data_page(struct page *page, struct f2fs_io_info *fio)
 			!is_cold_data(page) &&
 			need_inplace_update(inode))) {
 		rewrite_data_page(page, old_blkaddr, fio);
+		set_inode_flag(F2FS_I(inode), FI_UPDATE_WRITE);
 	} else {
 		write_data_page(page, &dn, &new_blkaddr, fio);
 		update_extent_cache(new_blkaddr, &dn);
+		set_inode_flag(F2FS_I(inode), FI_APPEND_WRITE);
 	}
 out_writepage:
 	f2fs_put_dnode(&dn);
@@ -914,6 +921,16 @@ skip_write:
 	return 0;
 }
 
+static void f2fs_write_failed(struct address_space *mapping, loff_t to)
+{
+	struct inode *inode = mapping->host;
+
+	if (to > inode->i_size) {
+		truncate_pagecache(inode, inode->i_size);
+		truncate_blocks(inode, inode->i_size);
+	}
+}
+
 static int f2fs_write_begin(struct file *file, struct address_space *mapping,
 		loff_t pos, unsigned len, unsigned flags,
 		struct page **pagep, void **fsdata)
@@ -931,11 +948,13 @@ static int f2fs_write_begin(struct file *file, struct address_space *mapping,
 repeat:
 	err = f2fs_convert_inline_data(inode, pos + len);
 	if (err)
-		return err;
+		goto fail;
 
 	page = grab_cache_page_write_begin(mapping, index, flags);
-	if (!page)
-		return -ENOMEM;
+	if (!page) {
+		err = -ENOMEM;
+		goto fail;
+	}
 
 	/* to avoid latency during memory pressure */
 	unlock_page(page);
@@ -949,10 +968,9 @@ repeat:
 	set_new_dnode(&dn, inode, NULL, NULL, 0);
 	err = f2fs_reserve_block(&dn, index);
 	f2fs_unlock_op(sbi);
-
 	if (err) {
 		f2fs_put_page(page, 0);
-		return err;
+		goto fail;
 	}
 inline_data:
 	lock_page(page);
@@ -982,19 +1000,20 @@ inline_data:
 			err = f2fs_read_inline_data(inode, page);
 			if (err) {
 				page_cache_release(page);
-				return err;
+				goto fail;
 			}
 		} else {
 			err = f2fs_submit_page_bio(sbi, page, dn.data_blkaddr,
 							READ_SYNC);
 			if (err)
-				return err;
+				goto fail;
 		}
 
 		lock_page(page);
 		if (unlikely(!PageUptodate(page))) {
 			f2fs_put_page(page, 1);
-			return -EIO;
+			err = -EIO;
+			goto fail;
 		}
 		if (unlikely(page->mapping != mapping)) {
 			f2fs_put_page(page, 1);
@@ -1005,6 +1024,9 @@ out:
 	SetPageUptodate(page);
 	clear_cold_data(page);
 	return 0;
+fail:
+	f2fs_write_failed(mapping, pos + len);
+	return err;
 }
 
 static int f2fs_write_end(struct file *file,
@@ -1016,7 +1038,6 @@ static int f2fs_write_end(struct file *file,
 
 	trace_f2fs_write_end(inode, pos, len, copied);
 
-	SetPageUptodate(page);
 	set_page_dirty(page);
 
 	if (pos + copied > i_size_read(inode)) {
@@ -1050,7 +1071,10 @@ static ssize_t f2fs_direct_IO(int rw, struct kiocb *iocb,
 		struct iov_iter *iter, loff_t offset)
 {
 	struct file *file = iocb->ki_filp;
-	struct inode *inode = file->f_mapping->host;
+	struct address_space *mapping = file->f_mapping;
+	struct inode *inode = mapping->host;
+	size_t count = iov_iter_count(iter);
+	int err;
 
 	/* Let buffer I/O handle the inline data case. */
 	if (f2fs_has_inline_data(inode))
@@ -1062,8 +1086,15 @@ static ssize_t f2fs_direct_IO(int rw, struct kiocb *iocb,
 	/* clear fsync mark to recover these blocks */
 	fsync_mark_clear(F2FS_SB(inode->i_sb), inode->i_ino);
 
-	return blockdev_direct_IO(rw, iocb, inode, iter, offset,
-				  get_data_block);
+	trace_f2fs_direct_IO_enter(inode, offset, count, rw);
+
+	err = blockdev_direct_IO(rw, iocb, inode, iter, offset, get_data_block);
+	if (err < 0 && (rw & WRITE))
+		f2fs_write_failed(mapping, offset + count);
+
+	trace_f2fs_direct_IO_exit(inode, offset, count, rw, err);
+
+	return err;
 }
 
 static void f2fs_invalidate_data_page(struct page *page, unsigned int offset,
diff --git a/fs/f2fs/debug.c b/fs/f2fs/debug.c
index b52c12cf5873..a441ba33be11 100644
--- a/fs/f2fs/debug.c
+++ b/fs/f2fs/debug.c
@@ -167,7 +167,7 @@ get_cache:
 	si->cache_mem += npages << PAGE_CACHE_SHIFT;
 	npages = META_MAPPING(sbi)->nrpages;
 	si->cache_mem += npages << PAGE_CACHE_SHIFT;
-	si->cache_mem += sbi->n_orphans * sizeof(struct orphan_inode_entry);
+	si->cache_mem += sbi->n_orphans * sizeof(struct ino_entry);
 	si->cache_mem += sbi->n_dirty_dirs * sizeof(struct dir_inode_entry);
 }
 
@@ -345,21 +345,14 @@ void __init f2fs_create_root_stats(void)
 
 	f2fs_debugfs_root = debugfs_create_dir("f2fs", NULL);
 	if (!f2fs_debugfs_root)
-		goto bail;
+		return;
 
 	file = debugfs_create_file("status", S_IRUGO, f2fs_debugfs_root,
 			NULL, &stat_fops);
-	if (!file)
-		goto free_debugfs_dir;
-
-	return;
-
-free_debugfs_dir:
-	debugfs_remove(f2fs_debugfs_root);
-
-bail:
-	f2fs_debugfs_root = NULL;
-	return;
+	if (!file) {
+		debugfs_remove(f2fs_debugfs_root);
+		f2fs_debugfs_root = NULL;
+	}
 }
 
 void f2fs_destroy_root_stats(void)
diff --git a/fs/f2fs/dir.c b/fs/f2fs/dir.c
index a4addd72ebbd..bcf893c3d903 100644
--- a/fs/f2fs/dir.c
+++ b/fs/f2fs/dir.c
@@ -77,8 +77,8 @@ static unsigned long dir_block_index(unsigned int level,
 	return bidx;
 }
 
-static bool early_match_name(const char *name, size_t namelen,
-			f2fs_hash_t namehash, struct f2fs_dir_entry *de)
+static bool early_match_name(size_t namelen, f2fs_hash_t namehash,
+				struct f2fs_dir_entry *de)
 {
 	if (le16_to_cpu(de->name_len) != namelen)
 		return false;
@@ -90,7 +90,7 @@ static bool early_match_name(const char *name, size_t namelen,
 }
 
 static struct f2fs_dir_entry *find_in_block(struct page *dentry_page,
-			const char *name, size_t namelen, int *max_slots,
+			struct qstr *name, int *max_slots,
 			f2fs_hash_t namehash, struct page **res_page)
 {
 	struct f2fs_dir_entry *de;
@@ -109,9 +109,10 @@ static struct f2fs_dir_entry *find_in_block(struct page *dentry_page,
 			continue;
 		}
 		de = &dentry_blk->dentry[bit_pos];
-		if (early_match_name(name, namelen, namehash, de)) {
+		if (early_match_name(name->len, namehash, de)) {
 			if (!memcmp(dentry_blk->filename[bit_pos],
-							name, namelen)) {
+							name->name,
+							name->len)) {
 				*res_page = dentry_page;
 				goto found;
 			}
@@ -120,6 +121,13 @@ static struct f2fs_dir_entry *find_in_block(struct page *dentry_page,
 			*max_slots = max_len;
 			max_len = 0;
 		}
+
+		/*
+		 * For the most part, it should be a bug when name_len is zero.
+		 * We stop here for figuring out where the bugs are occurred.
+		 */
+		f2fs_bug_on(!de->name_len);
+
 		bit_pos += GET_DENTRY_SLOTS(le16_to_cpu(de->name_len));
 	}
 
@@ -132,10 +140,10 @@ found:
 }
 
 static struct f2fs_dir_entry *find_in_level(struct inode *dir,
-		unsigned int level, const char *name, size_t namelen,
+			unsigned int level, struct qstr *name,
 			f2fs_hash_t namehash, struct page **res_page)
 {
-	int s = GET_DENTRY_SLOTS(namelen);
+	int s = GET_DENTRY_SLOTS(name->len);
 	unsigned int nbucket, nblock;
 	unsigned int bidx, end_block;
 	struct page *dentry_page;
@@ -160,8 +168,8 @@ static struct f2fs_dir_entry *find_in_level(struct inode *dir,
 			continue;
 		}
 
-		de = find_in_block(dentry_page, name, namelen,
-					&max_slots, namehash, res_page);
+		de = find_in_block(dentry_page, name, &max_slots,
+					namehash, res_page);
 		if (de)
 			break;
 
@@ -187,8 +195,6 @@ static struct f2fs_dir_entry *find_in_level(struct inode *dir,
 struct f2fs_dir_entry *f2fs_find_entry(struct inode *dir,
 			struct qstr *child, struct page **res_page)
 {
-	const char *name = child->name;
-	size_t namelen = child->len;
 	unsigned long npages = dir_blocks(dir);
 	struct f2fs_dir_entry *de = NULL;
 	f2fs_hash_t name_hash;
@@ -200,12 +206,11 @@ struct f2fs_dir_entry *f2fs_find_entry(struct inode *dir,
 
 	*res_page = NULL;
 
-	name_hash = f2fs_dentry_hash(name, namelen);
+	name_hash = f2fs_dentry_hash(child);
 	max_depth = F2FS_I(dir)->i_current_depth;
 
 	for (level = 0; level < max_depth; level++) {
-		de = find_in_level(dir, level, name,
-				namelen, name_hash, res_page);
+		de = find_in_level(dir, level, child, name_hash, res_page);
 		if (de)
 			break;
 	}
@@ -298,14 +303,13 @@ static int make_empty_dir(struct inode *inode,
 	struct page *dentry_page;
 	struct f2fs_dentry_block *dentry_blk;
 	struct f2fs_dir_entry *de;
-	void *kaddr;
 
 	dentry_page = get_new_data_page(inode, page, 0, true);
 	if (IS_ERR(dentry_page))
 		return PTR_ERR(dentry_page);
 
-	kaddr = kmap_atomic(dentry_page);
-	dentry_blk = (struct f2fs_dentry_block *)kaddr;
+
+	dentry_blk = kmap_atomic(dentry_page);
 
 	de = &dentry_blk->dentry[0];
 	de->name_len = cpu_to_le16(1);
@@ -323,7 +327,7 @@ static int make_empty_dir(struct inode *inode,
 
 	test_and_set_bit_le(0, &dentry_blk->dentry_bitmap);
 	test_and_set_bit_le(1, &dentry_blk->dentry_bitmap);
-	kunmap_atomic(kaddr);
+	kunmap_atomic(dentry_blk);
 
 	set_page_dirty(dentry_page);
 	f2fs_put_page(dentry_page, 1);
@@ -333,11 +337,12 @@ static int make_empty_dir(struct inode *inode,
 static struct page *init_inode_metadata(struct inode *inode,
 		struct inode *dir, const struct qstr *name)
 {
+	struct f2fs_sb_info *sbi = F2FS_SB(dir->i_sb);
 	struct page *page;
 	int err;
 
 	if (is_inode_flag_set(F2FS_I(inode), FI_NEW_INODE)) {
-		page = new_inode_page(inode, name);
+		page = new_inode_page(inode);
 		if (IS_ERR(page))
 			return page;
 
@@ -362,7 +367,8 @@ static struct page *init_inode_metadata(struct inode *inode,
 		set_cold_node(inode, page);
 	}
 
-	init_dent_inode(name, page);
+	if (name)
+		init_dent_inode(name, page);
 
 	/*
 	 * This file should be checkpointed during fsync.
@@ -370,6 +376,12 @@ static struct page *init_inode_metadata(struct inode *inode,
 	 */
 	if (is_inode_flag_set(F2FS_I(inode), FI_INC_LINK)) {
 		file_lost_pino(inode);
+		/*
+		 * If link the tmpfile to alias through linkat path,
+		 * we should remove this inode from orphan list.
+		 */
+		if (inode->i_nlink == 0)
+			remove_orphan_inode(sbi, inode->i_ino);
 		inc_nlink(inode);
 	}
 	return page;
@@ -453,7 +465,7 @@ int __f2fs_add_link(struct inode *dir, const struct qstr *name,
 	int err = 0;
 	int i;
 
-	dentry_hash = f2fs_dentry_hash(name->name, name->len);
+	dentry_hash = f2fs_dentry_hash(name);
 	level = 0;
 	current_depth = F2FS_I(dir)->i_current_depth;
 	if (F2FS_I(dir)->chash == dentry_hash) {
@@ -529,6 +541,27 @@ fail:
 	return err;
 }
 
+int f2fs_do_tmpfile(struct inode *inode, struct inode *dir)
+{
+	struct page *page;
+	int err = 0;
+
+	down_write(&F2FS_I(inode)->i_sem);
+	page = init_inode_metadata(inode, dir, NULL);
+	if (IS_ERR(page)) {
+		err = PTR_ERR(page);
+		goto fail;
+	}
+	/* we don't need to mark_inode_dirty now */
+	update_inode(inode, page);
+	f2fs_put_page(page, 1);
+
+	clear_inode_flag(F2FS_I(inode), FI_NEW_INODE);
+fail:
+	up_write(&F2FS_I(inode)->i_sem);
+	return err;
+}
+
 /*
  * It only removes the dentry from the dentry page,corresponding name
  * entry in name page does not need to be touched during deletion.
@@ -541,14 +574,13 @@ void f2fs_delete_entry(struct f2fs_dir_entry *dentry, struct page *page,
 	struct address_space *mapping = page->mapping;
 	struct inode *dir = mapping->host;
 	int slots = GET_DENTRY_SLOTS(le16_to_cpu(dentry->name_len));
-	void *kaddr = page_address(page);
 	int i;
 
 	lock_page(page);
 	f2fs_wait_on_page_writeback(page, DATA);
 
-	dentry_blk = (struct f2fs_dentry_block *)kaddr;
-	bit_pos = dentry - (struct f2fs_dir_entry *)dentry_blk->dentry;
+	dentry_blk = page_address(page);
+	bit_pos = dentry - dentry_blk->dentry;
 	for (i = 0; i < slots; i++)
 		test_and_clear_bit_le(bit_pos + i, &dentry_blk->dentry_bitmap);
 
@@ -603,7 +635,6 @@ bool f2fs_empty_dir(struct inode *dir)
 	unsigned long nblock = dir_blocks(dir);
 
 	for (bidx = 0; bidx < nblock; bidx++) {
-		void *kaddr;
 		dentry_page = get_lock_data_page(dir, bidx);
 		if (IS_ERR(dentry_page)) {
 			if (PTR_ERR(dentry_page) == -ENOENT)
@@ -612,8 +643,8 @@ bool f2fs_empty_dir(struct inode *dir)
 				return false;
 		}
 
-		kaddr = kmap_atomic(dentry_page);
-		dentry_blk = (struct f2fs_dentry_block *)kaddr;
+
+		dentry_blk = kmap_atomic(dentry_page);
 		if (bidx == 0)
 			bit_pos = 2;
 		else
@@ -621,7 +652,7 @@ bool f2fs_empty_dir(struct inode *dir)
 		bit_pos = find_next_bit_le(&dentry_blk->dentry_bitmap,
 						NR_DENTRY_IN_BLOCK,
 						bit_pos);
-		kunmap_atomic(kaddr);
+		kunmap_atomic(dentry_blk);
 
 		f2fs_put_page(dentry_page, 1);
 
diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h
index 58df97e174d0..4dab5338a97a 100644
--- a/fs/f2fs/f2fs.h
+++ b/fs/f2fs/f2fs.h
@@ -41,6 +41,7 @@
 #define F2FS_MOUNT_INLINE_XATTR		0x00000080
 #define F2FS_MOUNT_INLINE_DATA		0x00000100
 #define F2FS_MOUNT_FLUSH_MERGE		0x00000200
+#define F2FS_MOUNT_NOBARRIER		0x00000400
 
 #define clear_opt(sbi, option)	(sbi->mount_opt.opt &= ~F2FS_MOUNT_##option)
 #define set_opt(sbi, option)	(sbi->mount_opt.opt |= F2FS_MOUNT_##option)
@@ -99,8 +100,15 @@ enum {
 	META_SSA
 };
 
-/* for the list of orphan inodes */
-struct orphan_inode_entry {
+/* for the list of ino */
+enum {
+	ORPHAN_INO,		/* for orphan ino list */
+	APPEND_INO,		/* for append ino list */
+	UPDATE_INO,		/* for update ino list */
+	MAX_INO_ENTRY,		/* max. list */
+};
+
+struct ino_entry {
 	struct list_head list;	/* list head */
 	nid_t ino;		/* inode number */
 };
@@ -256,6 +264,8 @@ struct f2fs_nm_info {
 	unsigned int nat_cnt;		/* the # of cached nat entries */
 	struct list_head nat_entries;	/* cached nat entry list (clean) */
 	struct list_head dirty_nat_entries; /* cached nat entry list (dirty) */
+	struct list_head nat_entry_set;	/* nat entry set list */
+	unsigned int dirty_nat_cnt;	/* total num of nat entries in set */
 
 	/* free node ids management */
 	struct radix_tree_root free_nid_root;/* root of the free_nid cache */
@@ -442,14 +452,17 @@ struct f2fs_sb_info {
 	struct inode *meta_inode;		/* cache meta blocks */
 	struct mutex cp_mutex;			/* checkpoint procedure lock */
 	struct rw_semaphore cp_rwsem;		/* blocking FS operations */
-	struct mutex node_write;		/* locking node writes */
+	struct rw_semaphore node_write;		/* locking node writes */
 	struct mutex writepages;		/* mutex for writepages() */
 	bool por_doing;				/* recovery is doing or not */
 	wait_queue_head_t cp_wait;
 
-	/* for orphan inode management */
-	struct list_head orphan_inode_list;	/* orphan inode list */
-	spinlock_t orphan_inode_lock;		/* for orphan inode list */
+	/* for inode management */
+	struct radix_tree_root ino_root[MAX_INO_ENTRY];	/* ino entry array */
+	spinlock_t ino_lock[MAX_INO_ENTRY];		/* for ino entry lock */
+	struct list_head ino_list[MAX_INO_ENTRY];	/* inode list head */
+
+	/* for orphan inode, use 0'th array */
 	unsigned int n_orphans;			/* # of orphan inodes */
 	unsigned int max_orphans;		/* max orphan inodes */
 
@@ -768,7 +781,7 @@ static inline void *__bitmap_ptr(struct f2fs_sb_info *sbi, int flag)
 		if (flag == NAT_BITMAP)
 			return &ckpt->sit_nat_version_bitmap;
 		else
-			return ((unsigned char *)ckpt + F2FS_BLKSIZE);
+			return (unsigned char *)ckpt + F2FS_BLKSIZE;
 	} else {
 		offset = (flag == NAT_BITMAP) ?
 			le32_to_cpu(ckpt->sit_ver_bitmap_bytesize) : 0;
@@ -983,11 +996,15 @@ enum {
 	FI_NO_EXTENT,		/* not to use the extent cache */
 	FI_INLINE_XATTR,	/* used for inline xattr */
 	FI_INLINE_DATA,		/* used for inline data*/
+	FI_APPEND_WRITE,	/* inode has appended data */
+	FI_UPDATE_WRITE,	/* inode has in-place-update data */
+	FI_NEED_IPU,		/* used fo ipu for fdatasync */
 };
 
 static inline void set_inode_flag(struct f2fs_inode_info *fi, int flag)
 {
-	set_bit(flag, &fi->flags);
+	if (!test_bit(flag, &fi->flags))
+		set_bit(flag, &fi->flags);
 }
 
 static inline int is_inode_flag_set(struct f2fs_inode_info *fi, int flag)
@@ -997,7 +1014,8 @@ static inline int is_inode_flag_set(struct f2fs_inode_info *fi, int flag)
 
 static inline void clear_inode_flag(struct f2fs_inode_info *fi, int flag)
 {
-	clear_bit(flag, &fi->flags);
+	if (test_bit(flag, &fi->flags))
+		clear_bit(flag, &fi->flags);
 }
 
 static inline void set_acl_inode(struct f2fs_inode_info *fi, umode_t mode)
@@ -1136,6 +1154,7 @@ void f2fs_set_link(struct inode *, struct f2fs_dir_entry *,
 int update_dent_inode(struct inode *, const struct qstr *);
 int __f2fs_add_link(struct inode *, const struct qstr *, struct inode *);
 void f2fs_delete_entry(struct f2fs_dir_entry *, struct page *, struct inode *);
+int f2fs_do_tmpfile(struct inode *, struct inode *);
 int f2fs_make_empty(struct inode *, struct inode *);
 bool f2fs_empty_dir(struct inode *);
 
@@ -1155,7 +1174,7 @@ void f2fs_msg(struct super_block *, const char *, const char *, ...);
 /*
  * hash.c
  */
-f2fs_hash_t f2fs_dentry_hash(const char *, size_t);
+f2fs_hash_t f2fs_dentry_hash(const struct qstr *);
 
 /*
  * node.c
@@ -1173,7 +1192,7 @@ int truncate_inode_blocks(struct inode *, pgoff_t);
 int truncate_xattr_node(struct inode *, struct page *);
 int wait_on_node_pages_writeback(struct f2fs_sb_info *, nid_t);
 void remove_inode_page(struct inode *);
-struct page *new_inode_page(struct inode *, const struct qstr *);
+struct page *new_inode_page(struct inode *);
 struct page *new_node_page(struct dnode_of_data *, unsigned int, struct page *);
 void ra_node_page(struct f2fs_sb_info *, nid_t);
 struct page *get_node_page(struct f2fs_sb_info *, pgoff_t);
@@ -1185,6 +1204,7 @@ void alloc_nid_done(struct f2fs_sb_info *, nid_t);
 void alloc_nid_failed(struct f2fs_sb_info *, nid_t);
 void recover_node_page(struct f2fs_sb_info *, struct page *,
 		struct f2fs_summary *, struct node_info *, block_t);
+void recover_inline_xattr(struct inode *, struct page *);
 bool recover_xattr_data(struct inode *, struct page *, block_t);
 int recover_inode_page(struct f2fs_sb_info *, struct page *);
 int restore_node_summary(struct f2fs_sb_info *, unsigned int,
@@ -1206,7 +1226,7 @@ void destroy_flush_cmd_control(struct f2fs_sb_info *);
 void invalidate_blocks(struct f2fs_sb_info *, block_t);
 void refresh_sit_entry(struct f2fs_sb_info *, block_t, block_t);
 void clear_prefree_segments(struct f2fs_sb_info *);
-void discard_next_dnode(struct f2fs_sb_info *);
+void discard_next_dnode(struct f2fs_sb_info *, block_t);
 int npages_for_summary_flush(struct f2fs_sb_info *);
 void allocate_new_segments(struct f2fs_sb_info *);
 struct page *get_sum_page(struct f2fs_sb_info *, unsigned int);
@@ -1240,6 +1260,9 @@ struct page *grab_meta_page(struct f2fs_sb_info *, pgoff_t);
 struct page *get_meta_page(struct f2fs_sb_info *, pgoff_t);
 int ra_meta_pages(struct f2fs_sb_info *, int, int, int);
 long sync_meta_pages(struct f2fs_sb_info *, enum page_type, long);
+void add_dirty_inode(struct f2fs_sb_info *, nid_t, int type);
+void remove_dirty_inode(struct f2fs_sb_info *, nid_t, int type);
+bool exist_written_data(struct f2fs_sb_info *, nid_t, int);
 int acquire_orphan_inode(struct f2fs_sb_info *);
 void release_orphan_inode(struct f2fs_sb_info *);
 void add_orphan_inode(struct f2fs_sb_info *, nid_t);
@@ -1251,7 +1274,7 @@ void add_dirty_dir_inode(struct inode *);
 void remove_dirty_dir_inode(struct inode *);
 void sync_dirty_dir_inodes(struct f2fs_sb_info *);
 void write_checkpoint(struct f2fs_sb_info *, bool);
-void init_orphan_info(