Merge tag 'for-f2fs-4.12' of git://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk/f2fs

Pull f2fs updates from Jaegeuk Kim:
 "In this round, we've focused on enhancing performance with regards to
  block allocation, GC, and discard/in-place-update IO controls. There
  are a bunch of clean-ups as well as minor bug fixes.

  Enhancements:
   - disable heap-based allocation by default
   - issue small-sized discard commands by default
   - change the policy of data hotness for logging
   - distinguish IOs in terms of size and wbc type
   - start SSR earlier to avoid foreground GC
   - enhance data structures managing discard commands
   - enhance in-place update flow
   - add some more fault injection routines
   - secure one more xattr entry

  Bug fixes:
   - calculate victim cost for GC correctly
   - remain correct victim segment number for GC
   - race condition in nid allocator and initializer
   - stale pointer produced by atomic_writes
   - fix missing REQ_SYNC for flush commands
   - handle missing errors in more corner cases"

* tag 'for-f2fs-4.12' of git://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk/f2fs: (111 commits)
  f2fs: fix a mount fail for wrong next_scan_nid
  f2fs: enhance scalability of trace macro
  f2fs: relocate inode_{,un}lock in F2FS_IOC_SETFLAGS
  f2fs: Make flush bios explicitely sync
  f2fs: show available_nids in f2fs/status
  f2fs: flush dirty nats periodically
  f2fs: introduce CP_TRIMMED_FLAG to avoid unneeded discard
  f2fs: allow cpc->reason to indicate more than one reason
  f2fs: release cp and dnode lock before IPU
  f2fs: shrink size of struct discard_cmd
  f2fs: don't hold cmd_lock during waiting discard command
  f2fs: nullify fio->encrypted_page for each writes
  f2fs: sanity check segment count
  f2fs: introduce valid_ipu_blkaddr to clean up
  f2fs: lookup extent cache first under IPU scenario
  f2fs: reconstruct code to write a data page
  f2fs: introduce __wait_discard_cmd
  f2fs: introduce __issue_discard_cmd
  f2fs: enable small discard by default
  f2fs: delay awaking discard thread
  ...

20 files changed, 1694 insertions, 902 deletions

diff --git a/fs/f2fs/checkpoint.c b/fs/f2fs/checkpoint.c
index 0339daf4ca02..ea9c317b5916 100644
--- a/fs/f2fs/checkpoint.c
+++ b/fs/f2fs/checkpoint.c
@@ -275,10 +275,11 @@ static int f2fs_write_meta_pages(struct address_space *mapping,
 		get_pages(sbi, F2FS_DIRTY_META) < nr_pages_to_skip(sbi, META))
 		goto skip_write;
 
-	trace_f2fs_writepages(mapping->host, wbc, META);
+	/* if locked failed, cp will flush dirty pages instead */
+	if (!mutex_trylock(&sbi->cp_mutex))
+		goto skip_write;
 
-	/* if mounting is failed, skip writing node pages */
-	mutex_lock(&sbi->cp_mutex);
+	trace_f2fs_writepages(mapping->host, wbc, META);
 	diff = nr_pages_to_write(sbi, META, wbc);
 	written = sync_meta_pages(sbi, META, wbc->nr_to_write);
 	mutex_unlock(&sbi->cp_mutex);
@@ -567,7 +568,7 @@ static int recover_orphan_inode(struct f2fs_sb_info *sbi, nid_t ino)
 	if (ni.blk_addr != NULL_ADDR) {
 		set_sbi_flag(sbi, SBI_NEED_FSCK);
 		f2fs_msg(sbi->sb, KERN_WARNING,
-			"%s: orphan failed (ino=%x), run fsck to fix.",
+			"%s: orphan failed (ino=%x) by kernel, retry mount.",
 				__func__, ino);
 		return -EIO;
 	}
@@ -677,7 +678,7 @@ static int get_checkpoint_version(struct f2fs_sb_info *sbi, block_t cp_addr,
 	*cp_block = (struct f2fs_checkpoint *)page_address(*cp_page);
 
 	crc_offset = le32_to_cpu((*cp_block)->checksum_offset);
-	if (crc_offset >= blk_size) {
+	if (crc_offset > (blk_size - sizeof(__le32))) {
 		f2fs_msg(sbi->sb, KERN_WARNING,
 			"invalid crc_offset: %zu", crc_offset);
 		return -EINVAL;
@@ -816,7 +817,9 @@ static void __add_dirty_inode(struct inode *inode, enum inode_type type)
 		return;
 
 	set_inode_flag(inode, flag);
-	list_add_tail(&F2FS_I(inode)->dirty_list, &sbi->inode_list[type]);
+	if (!f2fs_is_volatile_file(inode))
+		list_add_tail(&F2FS_I(inode)->dirty_list,
+						&sbi->inode_list[type]);
 	stat_inc_dirty_inode(sbi, type);
 }
 
@@ -941,6 +944,19 @@ int f2fs_sync_inode_meta(struct f2fs_sb_info *sbi)
 	return 0;
 }
 
+static void __prepare_cp_block(struct f2fs_sb_info *sbi)
+{
+	struct f2fs_checkpoint *ckpt = F2FS_CKPT(sbi);
+	struct f2fs_nm_info *nm_i = NM_I(sbi);
+	nid_t last_nid = nm_i->next_scan_nid;
+
+	next_free_nid(sbi, &last_nid);
+	ckpt->valid_block_count = cpu_to_le64(valid_user_blocks(sbi));
+	ckpt->valid_node_count = cpu_to_le32(valid_node_count(sbi));
+	ckpt->valid_inode_count = cpu_to_le32(valid_inode_count(sbi));
+	ckpt->next_free_nid = cpu_to_le32(last_nid);
+}
+
 /*
  * Freeze all the FS-operations for checkpoint.
  */
@@ -964,21 +980,26 @@ retry_flush_dents:
 		err = sync_dirty_inodes(sbi, DIR_INODE);
 		if (err)
 			goto out;
+		cond_resched();
 		goto retry_flush_dents;
 	}
 
+	/*
+	 * POR: we should ensure that there are no dirty node pages
+	 * until finishing nat/sit flush. inode->i_blocks can be updated.
+	 */
+	down_write(&sbi->node_change);
+
 	if (get_pages(sbi, F2FS_DIRTY_IMETA)) {
+		up_write(&sbi->node_change);
 		f2fs_unlock_all(sbi);
 		err = f2fs_sync_inode_meta(sbi);
 		if (err)
 			goto out;
+		cond_resched();
 		goto retry_flush_dents;
 	}
 
-	/*
-	 * POR: we should ensure that there are no dirty node pages
-	 * until finishing nat/sit flush.
-	 */
 retry_flush_nodes:
 	down_write(&sbi->node_write);
 
@@ -986,11 +1007,20 @@ retry_flush_nodes:
 		up_write(&sbi->node_write);
 		err = sync_node_pages(sbi, &wbc);
 		if (err) {
+			up_write(&sbi->node_change);
 			f2fs_unlock_all(sbi);
 			goto out;
 		}
+		cond_resched();
 		goto retry_flush_nodes;
 	}
+
+	/*
+	 * sbi->node_change is used only for AIO write_begin path which produces
+	 * dirty node blocks and some checkpoint values by block allocation.
+	 */
+	__prepare_cp_block(sbi);
+	up_write(&sbi->node_change);
 out:
 	blk_finish_plug(&plug);
 	return err;
@@ -1024,16 +1054,20 @@ static void update_ckpt_flags(struct f2fs_sb_info *sbi, struct cp_control *cpc)
 
 	spin_lock(&sbi->cp_lock);
 
-	if (cpc->reason == CP_UMOUNT && ckpt->cp_pack_total_block_count >
+	if ((cpc->reason & CP_UMOUNT) &&
+			le32_to_cpu(ckpt->cp_pack_total_block_count) >
 			sbi->blocks_per_seg - NM_I(sbi)->nat_bits_blocks)
 		disable_nat_bits(sbi, false);
 
-	if (cpc->reason == CP_UMOUNT)
+	if (cpc->reason & CP_TRIMMED)
+		__set_ckpt_flags(ckpt, CP_TRIMMED_FLAG);
+
+	if (cpc->reason & CP_UMOUNT)
 		__set_ckpt_flags(ckpt, CP_UMOUNT_FLAG);
 	else
 		__clear_ckpt_flags(ckpt, CP_UMOUNT_FLAG);
 
-	if (cpc->reason == CP_FASTBOOT)
+	if (cpc->reason & CP_FASTBOOT)
 		__set_ckpt_flags(ckpt, CP_FASTBOOT_FLAG);
 	else
 		__clear_ckpt_flags(ckpt, CP_FASTBOOT_FLAG);
@@ -1057,7 +1091,6 @@ static int do_checkpoint(struct f2fs_sb_info *sbi, struct cp_control *cpc)
 	struct f2fs_checkpoint *ckpt = F2FS_CKPT(sbi);
 	struct f2fs_nm_info *nm_i = NM_I(sbi);
 	unsigned long orphan_num = sbi->im[ORPHAN_INO].ino_num;
-	nid_t last_nid = nm_i->next_scan_nid;
 	block_t start_blk;
 	unsigned int data_sum_blocks, orphan_blocks;
 	__u32 crc32 = 0;
@@ -1074,14 +1107,11 @@ static int do_checkpoint(struct f2fs_sb_info *sbi, struct cp_control *cpc)
 			return -EIO;
 	}
 
-	next_free_nid(sbi, &last_nid);
-
 	/*
 	 * modify checkpoint
 	 * version number is already updated
 	 */
 	ckpt->elapsed_time = cpu_to_le64(get_mtime(sbi));
-	ckpt->valid_block_count = cpu_to_le64(valid_user_blocks(sbi));
 	ckpt->free_segment_count = cpu_to_le32(free_segments(sbi));
 	for (i = 0; i < NR_CURSEG_NODE_TYPE; i++) {
 		ckpt->cur_node_segno[i] =
@@ -1100,10 +1130,6 @@ static int do_checkpoint(struct f2fs_sb_info *sbi, struct cp_control *cpc)
 				curseg_alloc_type(sbi, i + CURSEG_HOT_DATA);
 	}
 
-	ckpt->valid_node_count = cpu_to_le32(valid_node_count(sbi));
-	ckpt->valid_inode_count = cpu_to_le32(valid_inode_count(sbi));
-	ckpt->next_free_nid = cpu_to_le32(last_nid);
-
 	/* 2 cp  + n data seg summary + orphan inode blocks */
 	data_sum_blocks = npages_for_summary_flush(sbi, false);
 	spin_lock(&sbi->cp_lock);
@@ -1143,7 +1169,6 @@ static int do_checkpoint(struct f2fs_sb_info *sbi, struct cp_control *cpc)
 	/* write nat bits */
 	if (enabled_nat_bits(sbi, cpc)) {
 		__u64 cp_ver = cur_cp_version(ckpt);
-		unsigned int i;
 		block_t blk;
 
 		cp_ver |= ((__u64)crc32 << 32);
@@ -1250,8 +1275,8 @@ int write_checkpoint(struct f2fs_sb_info *sbi, struct cp_control *cpc)
 	mutex_lock(&sbi->cp_mutex);
 
 	if (!is_sbi_flag_set(sbi, SBI_IS_DIRTY) &&
-		(cpc->reason == CP_FASTBOOT || cpc->reason == CP_SYNC ||
-		(cpc->reason == CP_DISCARD && !sbi->discard_blks)))
+		((cpc->reason & CP_FASTBOOT) || (cpc->reason & CP_SYNC) ||
+		((cpc->reason & CP_DISCARD) && !sbi->discard_blks)))
 		goto out;
 	if (unlikely(f2fs_cp_error(sbi))) {
 		err = -EIO;
@@ -1273,7 +1298,7 @@ int write_checkpoint(struct f2fs_sb_info *sbi, struct cp_control *cpc)
 	f2fs_flush_merged_bios(sbi);
 
 	/* this is the case of multiple fstrims without any changes */
-	if (cpc->reason == CP_DISCARD) {
+	if (cpc->reason & CP_DISCARD) {
 		if (!exist_trim_candidates(sbi, cpc)) {
 			unblock_operations(sbi);
 			goto out;
@@ -1311,7 +1336,7 @@ int write_checkpoint(struct f2fs_sb_info *sbi, struct cp_control *cpc)
 	unblock_operations(sbi);
 	stat_inc_cp_count(sbi->stat_info);
 
-	if (cpc->reason == CP_RECOVERY)
+	if (cpc->reason & CP_RECOVERY)
 		f2fs_msg(sbi->sb, KERN_NOTICE,
 			"checkpoint: version = %llx", ckpt_ver);
 
diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c
index 1602b4bccae6..7c0f6bdf817d 100644
--- a/fs/f2fs/data.c
+++ b/fs/f2fs/data.c
@@ -309,7 +309,7 @@ static void __f2fs_submit_merged_bio(struct f2fs_sb_info *sbi,
 	if (type >= META_FLUSH) {
 		io->fio.type = META_FLUSH;
 		io->fio.op = REQ_OP_WRITE;
-		io->fio.op_flags = REQ_META | REQ_PRIO;
+		io->fio.op_flags = REQ_META | REQ_PRIO | REQ_SYNC;
 		if (!test_opt(sbi, NOBARRIER))
 			io->fio.op_flags |= REQ_PREFLUSH | REQ_FUA;
 	}
@@ -341,7 +341,7 @@ void f2fs_flush_merged_bios(struct f2fs_sb_info *sbi)
 
 /*
  * Fill the locked page with data located in the block address.
- * Return unlocked page.
+ * A caller needs to unlock the page on failure.
  */
 int f2fs_submit_page_bio(struct f2fs_io_info *fio)
 {
@@ -362,6 +362,9 @@ int f2fs_submit_page_bio(struct f2fs_io_info *fio)
 	bio_set_op_attrs(bio, fio->op, fio->op_flags);
 
 	__submit_bio(fio->sbi, bio, fio->type);
+
+	if (!is_read_io(fio->op))
+		inc_page_count(fio->sbi, WB_DATA_TYPE(fio->page));
 	return 0;
 }
 
@@ -787,6 +790,21 @@ int f2fs_preallocate_blocks(struct kiocb *iocb, struct iov_iter *from)
 	return err;
 }
 
+static inline void __do_map_lock(struct f2fs_sb_info *sbi, int flag, bool lock)
+{
+	if (flag == F2FS_GET_BLOCK_PRE_AIO) {
+		if (lock)
+			down_read(&sbi->node_change);
+		else
+			up_read(&sbi->node_change);
+	} else {
+		if (lock)
+			f2fs_lock_op(sbi);
+		else
+			f2fs_unlock_op(sbi);
+	}
+}
+
 /*
  * f2fs_map_blocks() now supported readahead/bmap/rw direct_IO with
  * f2fs_map_blocks structure.
@@ -829,7 +847,7 @@ int f2fs_map_blocks(struct inode *inode, struct f2fs_map_blocks *map,
 
 next_dnode:
 	if (create)
-		f2fs_lock_op(sbi);
+		__do_map_lock(sbi, flag, true);
 
 	/* When reading holes, we need its node page */
 	set_new_dnode(&dn, inode, NULL, NULL, 0);
@@ -939,7 +957,7 @@ skip:
 	f2fs_put_dnode(&dn);
 
 	if (create) {
-		f2fs_unlock_op(sbi);
+		__do_map_lock(sbi, flag, false);
 		f2fs_balance_fs(sbi, dn.node_changed);
 	}
 	goto next_dnode;
@@ -948,7 +966,7 @@ sync_out:
 	f2fs_put_dnode(&dn);
 unlock_out:
 	if (create) {
-		f2fs_unlock_op(sbi);
+		__do_map_lock(sbi, flag, false);
 		f2fs_balance_fs(sbi, dn.node_changed);
 	}
 out:
@@ -1151,9 +1169,10 @@ static int f2fs_mpage_readpages(struct address_space *mapping,
 
 	for (page_idx = 0; nr_pages; page_idx++, nr_pages--) {
 
-		prefetchw(&page->flags);
 		if (pages) {
 			page = list_last_entry(pages, struct page, lru);
+
+			prefetchw(&page->flags);
 			list_del(&page->lru);
 			if (add_to_page_cache_lru(page, mapping,
 						  page->index,
@@ -1283,17 +1302,83 @@ static int f2fs_read_data_pages(struct file *file,
 	return f2fs_mpage_readpages(mapping, pages, NULL, nr_pages);
 }
 
+static int encrypt_one_page(struct f2fs_io_info *fio)
+{
+	struct inode *inode = fio->page->mapping->host;
+	gfp_t gfp_flags = GFP_NOFS;
+
+	if (!f2fs_encrypted_inode(inode) || !S_ISREG(inode->i_mode))
+		return 0;
+
+	/* wait for GCed encrypted page writeback */
+	f2fs_wait_on_encrypted_page_writeback(fio->sbi, fio->old_blkaddr);
+
+retry_encrypt:
+	fio->encrypted_page = fscrypt_encrypt_page(inode, fio->page,
+			PAGE_SIZE, 0, fio->page->index, gfp_flags);
+	if (!IS_ERR(fio->encrypted_page))
+		return 0;
+
+	/* flush pending IOs and wait for a while in the ENOMEM case */
+	if (PTR_ERR(fio->encrypted_page) == -ENOMEM) {
+		f2fs_flush_merged_bios(fio->sbi);
+		congestion_wait(BLK_RW_ASYNC, HZ/50);
+		gfp_flags |= __GFP_NOFAIL;
+		goto retry_encrypt;
+	}
+	return PTR_ERR(fio->encrypted_page);
+}
+
+static inline bool need_inplace_update(struct f2fs_io_info *fio)
+{
+	struct inode *inode = fio->page->mapping->host;
+
+	if (S_ISDIR(inode->i_mode) || f2fs_is_atomic_file(inode))
+		return false;
+	if (is_cold_data(fio->page))
+		return false;
+	if (IS_ATOMIC_WRITTEN_PAGE(fio->page))
+		return false;
+
+	return need_inplace_update_policy(inode, fio);
+}
+
+static inline bool valid_ipu_blkaddr(struct f2fs_io_info *fio)
+{
+	if (fio->old_blkaddr == NEW_ADDR)
+		return false;
+	if (fio->old_blkaddr == NULL_ADDR)
+		return false;
+	return true;
+}
+
 int do_write_data_page(struct f2fs_io_info *fio)
 {
 	struct page *page = fio->page;
 	struct inode *inode = page->mapping->host;
 	struct dnode_of_data dn;
+	struct extent_info ei = {0,0,0};
+	bool ipu_force = false;
 	int err = 0;
 
 	set_new_dnode(&dn, inode, NULL, NULL, 0);
+	if (need_inplace_update(fio) &&
+			f2fs_lookup_extent_cache(inode, page->index, &ei)) {
+		fio->old_blkaddr = ei.blk + page->index - ei.fofs;
+
+		if (valid_ipu_blkaddr(fio)) {
+			ipu_force = true;
+			fio->need_lock = false;
+			goto got_it;
+		}
+	}
+
+	if (fio->need_lock)
+		f2fs_lock_op(fio->sbi);
+
 	err = get_dnode_of_data(&dn, page->index, LOOKUP_NODE);
 	if (err)
-		return err;
+		goto out;
 
 	fio->old_blkaddr = dn.data_blkaddr;
 
@@ -1302,31 +1387,10 @@ int do_write_data_page(struct f2fs_io_info *fio)
 		ClearPageUptodate(page);
 		goto out_writepage;
 	}
-
-	if (f2fs_encrypted_inode(inode) && S_ISREG(inode->i_mode)) {
-		gfp_t gfp_flags = GFP_NOFS;
-
-		/* wait for GCed encrypted page writeback */
-		f2fs_wait_on_encrypted_page_writeback(F2FS_I_SB(inode),
-							fio->old_blkaddr);
-retry_encrypt:
-		fio->encrypted_page = fscrypt_encrypt_page(inode, fio->page,
-							PAGE_SIZE, 0,
-							fio->page->index,
-							gfp_flags);
-		if (IS_ERR(fio->encrypted_page)) {
-			err = PTR_ERR(fio->encrypted_page);
-			if (err == -ENOMEM) {
-				/* flush pending ios and wait for a while */
-				f2fs_flush_merged_bios(F2FS_I_SB(inode));
-				congestion_wait(BLK_RW_ASYNC, HZ/50);
-				gfp_flags |= __GFP_NOFAIL;
-				err = 0;
-				goto retry_encrypt;
-			}
-			goto out_writepage;
-		}
-	}
+got_it:
+	err = encrypt_one_page(fio);
+	if (err)
+		goto out_writepage;
 
 	set_page_writeback(page);
 
@@ -1334,22 +1398,27 @@ retry_encrypt:
 	 * If current allocation needs SSR,
 	 * it had better in-place writes for updated data.
 	 */
-	if (unlikely(fio->old_blkaddr != NEW_ADDR &&
-			!is_cold_data(page) &&
-			!IS_ATOMIC_WRITTEN_PAGE(page) &&
-			need_inplace_update(inode))) {
-		rewrite_data_page(fio);
+	if (ipu_force || (valid_ipu_blkaddr(fio) && need_inplace_update(fio))) {
+		f2fs_put_dnode(&dn);
+		if (fio->need_lock)
+			f2fs_unlock_op(fio->sbi);
+		err = rewrite_data_page(fio);
+		trace_f2fs_do_write_data_page(fio->page, IPU);
 		set_inode_flag(inode, FI_UPDATE_WRITE);
-		trace_f2fs_do_write_data_page(page, IPU);
-	} else {
-		write_data_page(&dn, fio);
-		trace_f2fs_do_write_data_page(page, OPU);
-		set_inode_flag(inode, FI_APPEND_WRITE);
-		if (page->index == 0)
-			set_inode_flag(inode, FI_FIRST_BLOCK_WRITTEN);
+		return err;
 	}
+
+	/* LFS mode write path */
+	write_data_page(&dn, fio);
+	trace_f2fs_do_write_data_page(page, OPU);
+	set_inode_flag(inode, FI_APPEND_WRITE);
+	if (page->index == 0)
+		set_inode_flag(inode, FI_FIRST_BLOCK_WRITTEN);
 out_writepage:
 	f2fs_put_dnode(&dn);
+out:
+	if (fio->need_lock)
+		f2fs_unlock_op(fio->sbi);
 	return err;
 }
 
@@ -1370,9 +1439,11 @@ static int __write_data_page(struct page *page, bool *submitted,
 		.type = DATA,
 		.op = REQ_OP_WRITE,
 		.op_flags = wbc_to_write_flags(wbc),
+		.old_blkaddr = NULL_ADDR,
 		.page = page,
 		.encrypted_page = NULL,
 		.submitted = false,
+		.need_lock = true,
 	};
 
 	trace_f2fs_writepage(page, DATA);
@@ -1408,6 +1479,7 @@ write:
 
 	/* Dentry blocks are controlled by checkpoint */
 	if (S_ISDIR(inode->i_mode)) {
+		fio.need_lock = false;
 		err = do_write_data_page(&fio);
 		goto done;
 	}
@@ -1416,6 +1488,8 @@ write:
 		need_balance_fs = true;
 	else if (has_not_enough_free_secs(sbi, 0, 0))
 		goto redirty_out;
+	else
+		set_inode_flag(inode, FI_HOT_DATA);
 
 	err = -EAGAIN;
 	if (f2fs_has_inline_data(inode)) {
@@ -1423,12 +1497,12 @@ write:
 		if (!err)
 			goto out;
 	}
-	f2fs_lock_op(sbi);
+
 	if (err == -EAGAIN)
 		err = do_write_data_page(&fio);
 	if (F2FS_I(inode)->last_disk_size < psize)
 		F2FS_I(inode)->last_disk_size = psize;
-	f2fs_unlock_op(sbi);
+
 done:
 	if (err && err != -ENOENT)
 		goto redirty_out;
@@ -1441,12 +1515,14 @@ out:
 	if (wbc->for_reclaim) {
 		f2fs_submit_merged_bio_cond(sbi, inode, 0, page->index,
 						DATA, WRITE);
+		clear_inode_flag(inode, FI_HOT_DATA);
 		remove_dirty_inode(inode);
 		submitted = NULL;
 	}
 
 	unlock_page(page);
-	f2fs_balance_fs(sbi, need_balance_fs);
+	if (!S_ISDIR(inode->i_mode))
+		f2fs_balance_fs(sbi, need_balance_fs);
 
 	if (unlikely(f2fs_cp_error(sbi))) {
 		f2fs_submit_merged_bio(sbi, DATA, WRITE);
@@ -1495,6 +1571,12 @@ static int f2fs_write_cache_pages(struct address_space *mapping,
 
 	pagevec_init(&pvec, 0);
 
+	if (get_dirty_pages(mapping->host) <=
+				SM_I(F2FS_M_SB(mapping))->min_hot_blocks)
+		set_inode_flag(mapping->host, FI_HOT_DATA);
+	else
+		clear_inode_flag(mapping->host, FI_HOT_DATA);
+
 	if (wbc->range_cyclic) {
 		writeback_index = mapping->writeback_index; /* prev offset */
 		index = writeback_index;
@@ -1580,8 +1662,10 @@ continue_unlock:
 				last_idx = page->index;
 			}
 
-			if (--wbc->nr_to_write <= 0 &&
-			    wbc->sync_mode == WB_SYNC_NONE) {
+			/* give a priority to WB_SYNC threads */
+			if ((atomic_read(&F2FS_M_SB(mapping)->wb_sync_req) ||
+					--wbc->nr_to_write <= 0) &&
+					wbc->sync_mode == WB_SYNC_NONE) {
 				done = 1;
 				break;
 			}
@@ -1637,9 +1721,18 @@ static int f2fs_write_data_pages(struct address_space *mapping,
 
 	trace_f2fs_writepages(mapping->host, wbc, DATA);
 
+	/* to avoid spliting IOs due to mixed WB_SYNC_ALL and WB_SYNC_NONE */
+	if (wbc->sync_mode == WB_SYNC_ALL)
+		atomic_inc(&sbi->wb_sync_req);
+	else if (atomic_read(&sbi->wb_sync_req))
+		goto skip_write;
+
 	blk_start_plug(&plug);
 	ret = f2fs_write_cache_pages(mapping, wbc);
 	blk_finish_plug(&plug);
+
+	if (wbc->sync_mode == WB_SYNC_ALL)
+		atomic_dec(&sbi->wb_sync_req);
 	/*
 	 * if some pages were truncated, we cannot guarantee its mapping->host
 	 * to detect pending bios.
@@ -1687,7 +1780,7 @@ static int prepare_write_begin(struct f2fs_sb_info *sbi,
 
 	if (f2fs_has_inline_data(inode) ||
 			(pos & PAGE_MASK) >= i_size_read(inode)) {
-		f2fs_lock_op(sbi);
+		__do_map_lock(sbi, F2FS_GET_BLOCK_PRE_AIO, true);
 		locked = true;
 	}
 restart:
@@ -1723,7 +1816,8 @@ restart:
 			err = get_dnode_of_data(&dn, index, LOOKUP_NODE);
 			if (err || dn.data_blkaddr == NULL_ADDR) {
 				f2fs_put_dnode(&dn);
-				f2fs_lock_op(sbi);
+				__do_map_lock(sbi, F2FS_GET_BLOCK_PRE_AIO,
+								true);
 				locked = true;
 				goto restart;
 			}
@@ -1737,7 +1831,7 @@ out:
 	f2fs_put_dnode(&dn);
 unlock_out:
 	if (locked)
-		f2fs_unlock_op(sbi);
+		__do_map_lock(sbi, F2FS_GET_BLOCK_PRE_AIO, false);
 	return err;
 }
 
@@ -1951,7 +2045,7 @@ void f2fs_invalidate_page(struct page *page, unsigned int offset,
 
 	/* This is atomic written page, keep Private */
 	if (IS_ATOMIC_WRITTEN_PAGE(page))
-		return;
+		return drop_inmem_page(inode, page);
 
 	set_page_private(page, 0);
 	ClearPagePrivate(page);
diff --git a/fs/f2fs/debug.c b/fs/f2fs/debug.c
index ee2d0a485fc3..87f449845f5f 100644
--- a/fs/f2fs/debug.c
+++ b/fs/f2fs/debug.c
@@ -51,15 +51,26 @@ static void update_general_status(struct f2fs_sb_info *sbi)
 	si->ndirty_all = sbi->ndirty_inode[DIRTY_META];
 	si->inmem_pages = get_pages(sbi, F2FS_INMEM_PAGES);
 	si->aw_cnt = atomic_read(&sbi->aw_cnt);
+	si->vw_cnt = atomic_read(&sbi->vw_cnt);
 	si->max_aw_cnt = atomic_read(&sbi->max_aw_cnt);
+	si->max_vw_cnt = atomic_read(&sbi->max_vw_cnt);
 	si->nr_wb_cp_data = get_pages(sbi, F2FS_WB_CP_DATA);
 	si->nr_wb_data = get_pages(sbi, F2FS_WB_DATA);
-	if (SM_I(sbi) && SM_I(sbi)->fcc_info)
-		si->nr_flush =
-			atomic_read(&SM_I(sbi)->fcc_info->submit_flush);
-	if (SM_I(sbi) && SM_I(sbi)->dcc_info)
-		si->nr_discard =
-			atomic_read(&SM_I(sbi)->dcc_info->submit_discard);
+	if (SM_I(sbi) && SM_I(sbi)->fcc_info) {
+		si->nr_flushed =
+			atomic_read(&SM_I(sbi)->fcc_info->issued_flush);
+		si->nr_flushing =
+			atomic_read(&SM_I(sbi)->fcc_info->issing_flush);
+	}
+	if (SM_I(sbi) && SM_I(sbi)->dcc_info) {
+		si->nr_discarded =
+			atomic_read(&SM_I(sbi)->dcc_info->issued_discard);
+		si->nr_discarding =
+			atomic_read(&SM_I(sbi)->dcc_info->issing_discard);
+		si->nr_discard_cmd =
+			atomic_read(&SM_I(sbi)->dcc_info->discard_cmd_cnt);
+		si->undiscard_blks = SM_I(sbi)->dcc_info->undiscard_blks;
+	}
 	si->total_count = (int)sbi->user_block_count / sbi->blocks_per_seg;
 	si->rsvd_segs = reserved_segments(sbi);
 	si->overp_segs = overprovision_segments(sbi);
@@ -86,6 +97,7 @@ static void update_general_status(struct f2fs_sb_info *sbi)
 	si->sits = MAIN_SEGS(sbi);
 	si->dirty_sits = SIT_I(sbi)->dirty_sentries;
 	si->free_nids = NM_I(sbi)->nid_cnt[FREE_NID_LIST];
+	si->avail_nids = NM_I(sbi)->available_nids;
 	si->alloc_nids = NM_I(sbi)->nid_cnt[ALLOC_NID_LIST];
 	si->bg_gc = sbi->bg_gc;
 	si->util_free = (int)(free_user_blocks(sbi) >> sbi->log_blocks_per_seg)
@@ -99,8 +111,8 @@ static void update_general_status(struct f2fs_sb_info *sbi)
 	for (i = CURSEG_HOT_DATA; i <= CURSEG_COLD_NODE; i++) {
 		struct curseg_info *curseg = CURSEG_I(sbi, i);
 		si->curseg[i] = curseg->segno;
-		si->cursec[i] = curseg->segno / sbi->segs_per_sec;
-		si->curzone[i] = si->cursec[i] / sbi->secs_per_zone;
+		si->cursec[i] = GET_SEC_FROM_SEG(sbi, curseg->segno);
+		si->curzone[i] = GET_ZONE_FROM_SEC(sbi, si->cursec[i]);
 	}
 
 	for (i = 0; i < 2; i++) {
@@ -124,10 +136,10 @@ static void update_sit_info(struct f2fs_sb_info *sbi)
 
 	bimodal = 0;
 	total_vblocks = 0;
-	blks_per_sec = sbi->segs_per_sec * sbi->blocks_per_seg;
+	blks_per_sec = BLKS_PER_SEC(sbi);
 	hblks_per_sec = blks_per_sec / 2;
 	for (segno = 0; segno < MAIN_SEGS(sbi); segno += sbi->segs_per_sec) {
-		vblocks = get_valid_blocks(sbi, segno, sbi->segs_per_sec);
+		vblocks = get_valid_blocks(sbi, segno, true);
 		dist = abs(vblocks - hblks_per_sec);
 		bimodal += dist * dist;
 
@@ -156,7 +168,11 @@ static void update_mem_info(struct f2fs_sb_info *sbi)
 	if (si->base_mem)
 		goto get_cache;
 
-	si->base_mem = sizeof(struct f2fs_sb_info) + sbi->sb->s_blocksize;
+	/* build stat */
+	si->base_mem = sizeof(struct f2fs_stat_info);
+
+	/* build superblock */
+	si->base_mem += sizeof(struct f2fs_sb_info) + sbi->sb->s_blocksize;
 	si->base_mem += 2 * sizeof(struct f2fs_inode_info);
 	si->base_mem += sizeof(*sbi->ckpt);
 	si->base_mem += sizeof(struct percpu_counter) * NR_COUNT_TYPE;
@@ -208,8 +224,11 @@ get_cache:
 	/* build merge flush thread */
 	if (SM_I(sbi)->fcc_info)
 		si->cache_mem += sizeof(struct flush_cmd_control);
-	if (SM_I(sbi)->dcc_info)
+	if (SM_I(sbi)->dcc_info) {
 		si->cache_mem += sizeof(struct discard_cmd_control);
+		si->cache_mem += sizeof(struct discard_cmd) *
+			atomic_read(&SM_I(sbi)->dcc_info->discard_cmd_cnt);
+	}
 
 	/* free nids */
 	si->cache_mem += (NM_I(sbi)->nid_cnt[FREE_NID_LIST] +
@@ -330,11 +349,16 @@ static int stat_show(struct seq_file *s, void *v)
 		seq_printf(s, "  - Inner Struct Count: tree: %d(%d), node: %d\n",
 				si->ext_tree, si->zombie_tree, si->ext_node);
 		seq_puts(s, "\nBalancing F2FS Async:\n");
-		seq_printf(s, "  - IO (CP: %4d, Data: %4d, Flush: %4d, Discard: %4d)\n",
+		seq_printf(s, "  - IO (CP: %4d, Data: %4d, Flush: (%4d %4d), "
+			"Discard: (%4d %4d)) cmd: %4d undiscard:%4u\n",
 			   si->nr_wb_cp_data, si->nr_wb_data,
-			   si->nr_flush, si->nr_discard);
-		seq_printf(s, "  - inmem: %4d, atomic IO: %4d (Max. %4d)\n",
-			   si->inmem_pages, si->aw_cnt, si->max_aw_cnt);
+			   si->nr_flushing, si->nr_flushed,
+			   si->nr_discarding, si->nr_discarded,
+			   si->nr_discard_cmd, si->undiscard_blks);
+		seq_printf(s, "  - inmem: %4d, atomic IO: %4d (Max. %4d), "
+			"volatile IO: %4d (Max. %4d)\n",
+			   si->inmem_pages, si->aw_cnt, si->max_aw_cnt,
+			   si->vw_cnt, si->max_vw_cnt);
 		seq_printf(s, "  - nodes: %4d in %4d\n",
 			   si->ndirty_node, si->node_pages);
 		seq_printf(s, "  - dents: %4d in dirs:%4d (%4d)\n",
@@ -347,8 +371,8 @@ static int stat_show(struct seq_file *s, void *v)
 			   si->ndirty_imeta);
 		seq_printf(s, "  - NATs: %9d/%9d\n  - SITs: %9d/%9d\n",
 			   si->dirty_nats, si->nats, si->dirty_sits, si->sits);
-		seq_printf(s, "  - free_nids: %9d, alloc_nids: %9d\n",
-			   si->free_nids, si->alloc_nids);
+		seq_printf(s, "  - free_nids: %9d/%9d\n  - alloc_nids: %9d\n",
+			   si->free_nids, si->avail_nids, si->alloc_nids);
 		seq_puts(s, "\nDistribution of User Blocks:");
 		seq_puts(s, " [ valid | invalid | free ]\n");
 		seq_puts(s, "  [");
@@ -434,7 +458,9 @@ int f2fs_build_stats(struct f2fs_sb_info *sbi)
 	atomic_set(&sbi->inplace_count, 0);
 
 	atomic_set(&sbi->aw_cnt, 0);
+	atomic_set(&sbi->vw_cnt, 0);
 	atomic_set(&sbi->max_aw_cnt, 0);
+	atomic_set(&sbi->max_vw_cnt, 0);
 
 	mutex_lock(&f2fs_stat_mutex);
 	list_add_tail(&si->stat_list, &f2fs_stat_list);
diff --git a/fs/f2fs/dir.c b/fs/f2fs/dir.c
index e64087052834..94756f55a97e 100644
--- a/fs/f2fs/dir.c
+++ b/fs/f2fs/dir.c
@@ -94,7 +94,7 @@ static struct f2fs_dir_entry *find_in_block(struct page *dentry_page,
 
 	dentry_blk = (struct f2fs_dentry_block *)kmap(dentry_page);
 
-	make_dentry_ptr(NULL, &d, (void *)dentry_blk, 1);
+	make_dentry_ptr_block(NULL, &d, dentry_blk);
 	de = find_target_dentry(fname, namehash, max_slots, &d);
 	if (de)
 		*res_page = dentry_page;
@@ -192,13 +192,9 @@ static struct f2fs_dir_entry *find_in_level(struct inode *dir,
 		f2fs_put_page(dentry_page, 0);
 	}
 
-	/* This is to increase the speed of f2fs_create */
-	if (!de && room) {
-		F2FS_I(dir)->task = current;
-		if (F2FS_I(dir)->chash != namehash) {
-			F2FS_I(dir)->chash = namehash;
-			F2FS_I(dir)->clevel = level;
-		}
+	if (!de && room && F2FS_I(dir)->chash != namehash) {
+		F2FS_I(dir)->chash = namehash;
+		F2FS_I(dir)->clevel = level;
 	}
 
 	return de;
@@ -239,6 +235,9 @@ struct f2fs_dir_entry *__f2fs_find_entry(struct inode *dir,
 			break;
 	}
 out:
+	/* This is to increase the speed of f2fs_create */
+	if (!de)
+		F2FS_I(dir)->task = current;
 	return de;
 }
 
@@ -322,24 +321,6 @@ static void init_dent_inode(const struct qstr *name, struct page *ipage)
 	set_page_dirty(ipage);
 }
 
-int update_dent_inode(struct inode *inode, struct inode *to,
-					const struct qstr *name)
-{
-	struct page *page;
-
-	if (file_enc_name(to))
-		return 0;
-
-	page = get_node_page(F2FS_I_SB(inode), inode->i_ino);
-	if (IS_ERR(page))
-		return PTR_ERR(page);
-
-	init_dent_inode(name, page);
-	f2fs_put_page(page, 1);
-
-	return 0;
-}
-
 void do_make_empty_dir(struct inode *inode, struct inode *parent,
 					struct f2fs_dentry_ptr *d)
 {
@@ -369,7 +350,7 @@ static int make_empty_dir(struct inode *inode,
 
 	dentry_blk = kmap_atomic(dentry_page);
 
-	make_dentry_ptr(NULL, &d, (void *)dentry_blk, 1);
+	make_dentry_ptr_block(NULL, &d, dentry_blk);
 	do_make_empty_dir(inode, parent, &d);
 
 	kunmap_atomic(dentry_blk);
@@ -423,8 +404,11 @@ struct page *init_inode_metadata(struct inode *inode, struct inode *dir,
 		set_cold_node(inode, page);
 	}
 
-	if (new_name)
+	if (new_name) {
 		init_dent_inode(new_name, page);
+		if (f2fs_encrypted_inode(dir))
+			file_set_enc_name(inode);
+	}
 
 	/*
 	 * This file should be checkpointed during fsync.
@@ -584,11 +568,9 @@ add_dentry:
 			err = PTR_ERR(page);
 			goto fail;
 		}
-		if (f2fs_encrypted_inode(dir))
-			file_set_enc_name(inode);
 	}
 
-	make_dentry_ptr(NULL, &d, (void *)dentry_blk, 1);
+	make_dentry_ptr_block(NULL, &d, dentry_blk);
 	f2fs_update_dentry(ino, mode, &d, new_name, dentry_hash, bit_pos);
 
 	set_page_dirty(dentry_page);
@@ -896,7 +878,7 @@ static int f2fs_readdir(struct file *file, struct dir_context *ctx)
 
 		dentry_blk = kmap(dentry_page);
 
-		make_dentry_ptr(inode, &d, (void *)dentry_blk, 1);
+		make_dentry_ptr_block(inode, &d, dentry_blk);
 
 		err = f2fs_fill_dentries(ctx, &d,
 				n * NR_DENTRY_IN_BLOCK, &fstr);
diff --git a/fs/f2fs/extent_cache.c b/fs/f2fs/extent_cache.c
index c6934f014e0f..2f98d7039701 100644
--- a/fs/f2fs/extent_cache.c
+++ b/fs/f2fs/extent_cache.c
@@ -18,6 +18,179 @@
 #include "node.h"
 #include <trace/events/f2fs.h>
 
+static struct rb_entry *__lookup_rb_tree_fast(struct rb_entry *cached_re,
+							unsigned int ofs)
+{
+	if (cached_re) {
+		if (cached_re->ofs <= ofs &&
+				cached_re->ofs + cached_re->len > ofs) {
+			return cached_re;
+		}
+	}
+	return NULL;
+}
+
+static struct rb_entry *__lookup_rb_tree_slow(struct rb_root *root,
+							unsigned int ofs)
+{
+	struct rb_node *node = root->rb_node;
+	struct rb_entry *re;
+
+	while (node) {
+		re = rb_entry(node, struct rb_entry, rb_node);
+
+		if (ofs < re->ofs)
+			node = node->rb_left;
+		else if (ofs >= re->ofs + re->len)
+			node = no