Merge tag 'for-f2fs-3.13' of git://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk/f2fs

Pull f2fs updates from Jaegeuk Kim:
 "This patch-set includes the following major enhancement patches.
   - add a sysfs to control reclaiming free segments
   - enhance the f2fs global lock procedures
   - enhance the victim selection flow
   - wait for selected node blocks during fsync
   - add some tracepoints
   - add a config to remove abundant BUG_ONs

  The other bug fixes are as follows.
   - fix deadlock on acl operations
   - fix some bugs with respect to orphan inodes

  And, there are a bunch of cleanups"

* tag 'for-f2fs-3.13' of git://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk/f2fs: (42 commits)
  f2fs: issue more large discard command
  f2fs: fix memory leak after kobject init failed in fill_super
  f2fs: cleanup waiting routine for writeback pages in cp
  f2fs: avoid to use a NULL point in destroy_segment_manager
  f2fs: remove unnecessary TestClearPageError when wait pages writeback
  f2fs: update f2fs document
  f2fs: avoid to wait all the node blocks during fsync
  f2fs: check all ones or zeros bitmap with bitops for better mount performance
  f2fs: change the method of calculating the number summary blocks
  f2fs: fix calculating incorrect free size when update xattr in __f2fs_setxattr
  f2fs: add an option to avoid unnecessary BUG_ONs
  f2fs: introduce CONFIG_F2FS_CHECK_FS for BUG_ON control
  f2fs: fix a deadlock during init_acl procedure
  f2fs: clean up acl flow for better readability
  f2fs: remove unnecessary segment bitmap updates
  f2fs: add tracepoint for vm_page_mkwrite
  f2fs: add tracepoint for set_page_dirty
  f2fs: remove redundant set_page_dirty from write_compacted_summaries
  f2fs: add reclaiming control by sysfs
  f2fs: introduce f2fs_balance_fs_bg for some background jobs
  ...

Documentation/filesystems/f2fs.txt

@@ -119,6 +119,7 @@ active_logs=%u         Support configuring the number of active logs. In the
                        Default number is 6.
 disable_ext_identify   Disable the extension list configured by mkfs, so f2fs
                        does not aware of cold files such as media files.
+inline_xattr           Enable the inline xattrs feature.
 
 ================================================================================
 DEBUGFS ENTRIES
@@ -164,6 +165,12 @@ Files in /sys/fs/f2fs/<devname>
                               gc_idle = 1 will select the Cost Benefit approach
                               & setting gc_idle = 2 will select the greedy aproach.
 
+ reclaim_segments             This parameter controls the number of prefree
+                              segments to be reclaimed. If the number of prefree
+			      segments is larger than this number, f2fs tries to
+			      conduct checkpoint to reclaim the prefree segments
+			      to free segments. By default, 100 segments, 200MB.
+
 ================================================================================
 USAGE
 ================================================================================

fs/f2fs/Kconfig

@@ -63,3 +63,11 @@ config F2FS_FS_SECURITY
 	  the extended attribute support in advance.
 
 	  If you are not using a security module, say N.
+
+config F2FS_CHECK_FS
+	bool "F2FS consistency checking feature"
+	depends on F2FS_FS
+	help
+	  Enables BUG_ONs which check the file system consistency in runtime.
+
+	  If you want to improve the performance, say N.

fs/f2fs/acl.c

@@ -205,7 +205,8 @@ struct posix_acl *f2fs_get_acl(struct inode *inode, int type)
 	return acl;
 }
 
-static int f2fs_set_acl(struct inode *inode, int type, struct posix_acl *acl)
+static int f2fs_set_acl(struct inode *inode, int type,
+			struct posix_acl *acl, struct page *ipage)
 {
 	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
 	struct f2fs_inode_info *fi = F2FS_I(inode);
@@ -250,7 +251,7 @@ static int f2fs_set_acl(struct inode *inode, int type, struct posix_acl *acl)
 		}
 	}
 
-	error = f2fs_setxattr(inode, name_index, "", value, size, NULL);
+	error = f2fs_setxattr(inode, name_index, "", value, size, ipage);
 
 	kfree(value);
 	if (!error)
@@ -260,10 +261,10 @@ static int f2fs_set_acl(struct inode *inode, int type, struct posix_acl *acl)
 	return error;
 }
 
-int f2fs_init_acl(struct inode *inode, struct inode *dir)
+int f2fs_init_acl(struct inode *inode, struct inode *dir, struct page *ipage)
 {
-	struct posix_acl *acl = NULL;
 	struct f2fs_sb_info *sbi = F2FS_SB(dir->i_sb);
+	struct posix_acl *acl = NULL;
 	int error = 0;
 
 	if (!S_ISLNK(inode->i_mode)) {
@@ -276,19 +277,19 @@ int f2fs_init_acl(struct inode *inode, struct inode *dir)
 			inode->i_mode &= ~current_umask();
 	}
 
-	if (test_opt(sbi, POSIX_ACL) && acl) {
+	if (!test_opt(sbi, POSIX_ACL) || !acl)
+		goto cleanup;
 
-		if (S_ISDIR(inode->i_mode)) {
-			error = f2fs_set_acl(inode, ACL_TYPE_DEFAULT, acl);
-			if (error)
-				goto cleanup;
-		}
-		error = posix_acl_create(&acl, GFP_KERNEL, &inode->i_mode);
-		if (error < 0)
-			return error;
-		if (error > 0)
-			error = f2fs_set_acl(inode, ACL_TYPE_ACCESS, acl);
+	if (S_ISDIR(inode->i_mode)) {
+		error = f2fs_set_acl(inode, ACL_TYPE_DEFAULT, acl, ipage);
+		if (error)
+			goto cleanup;
 	}
+	error = posix_acl_create(&acl, GFP_KERNEL, &inode->i_mode);
+	if (error < 0)
+		return error;
+	if (error > 0)
+		error = f2fs_set_acl(inode, ACL_TYPE_ACCESS, acl, ipage);
 cleanup:
 	posix_acl_release(acl);
 	return error;
@@ -313,7 +314,8 @@ int f2fs_acl_chmod(struct inode *inode)
 	error = posix_acl_chmod(&acl, GFP_KERNEL, mode);
 	if (error)
 		return error;
-	error = f2fs_set_acl(inode, ACL_TYPE_ACCESS, acl);
+
+	error = f2fs_set_acl(inode, ACL_TYPE_ACCESS, acl, NULL);
 	posix_acl_release(acl);
 	return error;
 }
@@ -388,7 +390,7 @@ static int f2fs_xattr_set_acl(struct dentry *dentry, const char *name,
 		acl = NULL;
 	}
 
-	error = f2fs_set_acl(inode, type, acl);
+	error = f2fs_set_acl(inode, type, acl, NULL);
 
 release_and_out:
 	posix_acl_release(acl);

fs/f2fs/acl.h

@@ -36,9 +36,9 @@ struct f2fs_acl_header {
 
 #ifdef CONFIG_F2FS_FS_POSIX_ACL
 
-extern struct posix_acl *f2fs_get_acl(struct inode *inode, int type);
-extern int f2fs_acl_chmod(struct inode *inode);
-extern int f2fs_init_acl(struct inode *inode, struct inode *dir);
+extern struct posix_acl *f2fs_get_acl(struct inode *, int);
+extern int f2fs_acl_chmod(struct inode *);
+extern int f2fs_init_acl(struct inode *, struct inode *, struct page *);
 #else
 #define f2fs_check_acl	NULL
 #define f2fs_get_acl	NULL
@@ -49,7 +49,8 @@ static inline int f2fs_acl_chmod(struct inode *inode)
 	return 0;
 }
 
-static inline int f2fs_init_acl(struct inode *inode, struct inode *dir)
+static inline int f2fs_init_acl(struct inode *inode, struct inode *dir,
+							struct page *page)
 {
 	return 0;
 }

fs/f2fs/checkpoint.c

@@ -81,7 +81,7 @@ static int f2fs_write_meta_page(struct page *page,
 	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
 
 	/* Should not write any meta pages, if any IO error was occurred */
-	if (wbc->for_reclaim ||
+	if (wbc->for_reclaim || sbi->por_doing ||
 			is_set_ckpt_flags(F2FS_CKPT(sbi), CP_ERROR_FLAG)) {
 		dec_page_count(sbi, F2FS_DIRTY_META);
 		wbc->pages_skipped++;
@@ -142,8 +142,8 @@ long sync_meta_pages(struct f2fs_sb_info *sbi, enum page_type type,
 		for (i = 0; i < nr_pages; i++) {
 			struct page *page = pvec.pages[i];
 			lock_page(page);
-			BUG_ON(page->mapping != mapping);
-			BUG_ON(!PageDirty(page));
+			f2fs_bug_on(page->mapping != mapping);
+			f2fs_bug_on(!PageDirty(page));
 			clear_page_dirty_for_io(page);
 			if (f2fs_write_meta_page(page, &wbc)) {
 				unlock_page(page);
@@ -167,6 +167,8 @@ static int f2fs_set_meta_page_dirty(struct page *page)
 	struct address_space *mapping = page->mapping;
 	struct f2fs_sb_info *sbi = F2FS_SB(mapping->host->i_sb);
 
+	trace_f2fs_set_page_dirty(page, META);
+
 	SetPageUptodate(page);
 	if (!PageDirty(page)) {
 		__set_page_dirty_nobuffers(page);
@@ -206,6 +208,7 @@ int acquire_orphan_inode(struct f2fs_sb_info *sbi)
 void release_orphan_inode(struct f2fs_sb_info *sbi)
 {
 	mutex_lock(&sbi->orphan_inode_mutex);
+	f2fs_bug_on(sbi->n_orphans == 0);
 	sbi->n_orphans--;
 	mutex_unlock(&sbi->orphan_inode_mutex);
 }
@@ -225,12 +228,8 @@ void add_orphan_inode(struct f2fs_sb_info *sbi, nid_t ino)
 			break;
 		orphan = NULL;
 	}
-retry:
-	new = kmem_cache_alloc(orphan_entry_slab, GFP_ATOMIC);
-	if (!new) {
-		cond_resched();
-		goto retry;
-	}
+
+	new = f2fs_kmem_cache_alloc(orphan_entry_slab, GFP_ATOMIC);
 	new->ino = ino;
 
 	/* add new_oentry into list which is sorted by inode number */
@@ -253,6 +252,7 @@ void remove_orphan_inode(struct f2fs_sb_info *sbi, nid_t ino)
 		if (orphan->ino == ino) {
 			list_del(&orphan->list);
 			kmem_cache_free(orphan_entry_slab, orphan);
+			f2fs_bug_on(sbi->n_orphans == 0);
 			sbi->n_orphans--;
 			break;
 		}
@@ -263,7 +263,7 @@ void remove_orphan_inode(struct f2fs_sb_info *sbi, nid_t ino)
 static void recover_orphan_inode(struct f2fs_sb_info *sbi, nid_t ino)
 {
 	struct inode *inode = f2fs_iget(sbi->sb, ino);
-	BUG_ON(IS_ERR(inode));
+	f2fs_bug_on(IS_ERR(inode));
 	clear_nlink(inode);
 
 	/* truncate all the data during iput */
@@ -277,7 +277,7 @@ int recover_orphan_inodes(struct f2fs_sb_info *sbi)
 	if (!is_set_ckpt_flags(F2FS_CKPT(sbi), CP_ORPHAN_PRESENT_FLAG))
 		return 0;
 
-	sbi->por_doing = 1;
+	sbi->por_doing = true;
 	start_blk = __start_cp_addr(sbi) + 1;
 	orphan_blkaddr = __start_sum_addr(sbi) - 1;
 
@@ -294,7 +294,7 @@ int recover_orphan_inodes(struct f2fs_sb_info *sbi)
 	}
 	/* clear Orphan Flag */
 	clear_ckpt_flags(F2FS_CKPT(sbi), CP_ORPHAN_PRESENT_FLAG);
-	sbi->por_doing = 0;
+	sbi->por_doing = false;
 	return 0;
 }
 
@@ -469,9 +469,7 @@ static int __add_dirty_inode(struct inode *inode, struct dir_inode_entry *new)
 			return -EEXIST;
 	}
 	list_add_tail(&new->list, head);
-#ifdef CONFIG_F2FS_STAT_FS
-	sbi->n_dirty_dirs++;
-#endif
+	stat_inc_dirty_dir(sbi);
 	return 0;
 }
 
@@ -482,12 +480,8 @@ void set_dirty_dir_page(struct inode *inode, struct page *page)
 
 	if (!S_ISDIR(inode->i_mode))
 		return;
-retry:
-	new = kmem_cache_alloc(inode_entry_slab, GFP_NOFS);
-	if (!new) {
-		cond_resched();
-		goto retry;
-	}
+
+	new = f2fs_kmem_cache_alloc(inode_entry_slab, GFP_NOFS);
 	new->inode = inode;
 	INIT_LIST_HEAD(&new->list);
 
@@ -504,13 +498,9 @@ void set_dirty_dir_page(struct inode *inode, struct page *page)
 void add_dirty_dir_inode(struct inode *inode)
 {
 	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
-	struct dir_inode_entry *new;
-retry:
-	new = kmem_cache_alloc(inode_entry_slab, GFP_NOFS);
-	if (!new) {
-		cond_resched();
-		goto retry;
-	}
+	struct dir_inode_entry *new =
+			f2fs_kmem_cache_alloc(inode_entry_slab, GFP_NOFS);
+
 	new->inode = inode;
 	INIT_LIST_HEAD(&new->list);
 
@@ -541,9 +531,7 @@ void remove_dirty_dir_inode(struct inode *inode)
 		if (entry->inode == inode) {
 			list_del(&entry->list);
 			kmem_cache_free(inode_entry_slab, entry);
-#ifdef CONFIG_F2FS_STAT_FS
-			sbi->n_dirty_dirs--;
-#endif
+			stat_dec_dirty_dir(sbi);
 			break;
 		}
 	}
@@ -617,11 +605,10 @@ static void block_operations(struct f2fs_sb_info *sbi)
 	blk_start_plug(&plug);
 
 retry_flush_dents:
-	mutex_lock_all(sbi);
-
+	f2fs_lock_all(sbi);
 	/* write all the dirty dentry pages */
 	if (get_pages(sbi, F2FS_DIRTY_DENTS)) {
-		mutex_unlock_all(sbi);
+		f2fs_unlock_all(sbi);
 		sync_dirty_dir_inodes(sbi);
 		goto retry_flush_dents;
 	}
@@ -644,7 +631,22 @@ static void block_operations(struct f2fs_sb_info *sbi)
 static void unblock_operations(struct f2fs_sb_info *sbi)
 {
 	mutex_unlock(&sbi->node_write);
-	mutex_unlock_all(sbi);
+	f2fs_unlock_all(sbi);
+}
+
+static void wait_on_all_pages_writeback(struct f2fs_sb_info *sbi)
+{
+	DEFINE_WAIT(wait);
+
+	for (;;) {
+		prepare_to_wait(&sbi->cp_wait, &wait, TASK_UNINTERRUPTIBLE);
+
+		if (!get_pages(sbi, F2FS_WRITEBACK))
+			break;
+
+		io_schedule();
+	}
+	finish_wait(&sbi->cp_wait, &wait);
 }
 
 static void do_checkpoint(struct f2fs_sb_info *sbi, bool is_umount)
@@ -756,8 +758,7 @@ static void do_checkpoint(struct f2fs_sb_info *sbi, bool is_umount)
 	f2fs_put_page(cp_page, 1);
 
 	/* wait for previous submitted node/meta pages writeback */
-	while (get_pages(sbi, F2FS_WRITEBACK))
-		congestion_wait(BLK_RW_ASYNC, HZ / 50);
+	wait_on_all_pages_writeback(sbi);
 
 	filemap_fdatawait_range(sbi->node_inode->i_mapping, 0, LONG_MAX);
 	filemap_fdatawait_range(sbi->meta_inode->i_mapping, 0, LONG_MAX);

fs/f2fs/data.c

@@ -68,9 +68,6 @@ static int check_extent_cache(struct inode *inode, pgoff_t pgofs,
 					struct buffer_head *bh_result)
 {
 	struct f2fs_inode_info *fi = F2FS_I(inode);
-#ifdef CONFIG_F2FS_STAT_FS
-	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
-#endif
 	pgoff_t start_fofs, end_fofs;
 	block_t start_blkaddr;
 
@@ -80,9 +77,8 @@ static int check_extent_cache(struct inode *inode, pgoff_t pgofs,
 		return 0;
 	}
 
-#ifdef CONFIG_F2FS_STAT_FS
-	sbi->total_hit_ext++;
-#endif
+	stat_inc_total_hit(inode->i_sb);
+
 	start_fofs = fi->ext.fofs;
 	end_fofs = fi->ext.fofs + fi->ext.len - 1;
 	start_blkaddr = fi->ext.blk_addr;
@@ -100,9 +96,7 @@ static int check_extent_cache(struct inode *inode, pgoff_t pgofs,
 		else
 			bh_result->b_size = UINT_MAX;
 
-#ifdef CONFIG_F2FS_STAT_FS
-		sbi->read_hit_ext++;
-#endif
+		stat_inc_read_hit(inode->i_sb);
 		read_unlock(&fi->ext.ext_lock);
 		return 1;
 	}
@@ -116,7 +110,7 @@ void update_extent_cache(block_t blk_addr, struct dnode_of_data *dn)
 	pgoff_t fofs, start_fofs, end_fofs;
 	block_t start_blkaddr, end_blkaddr;
 
-	BUG_ON(blk_addr == NEW_ADDR);
+	f2fs_bug_on(blk_addr == NEW_ADDR);
 	fofs = start_bidx_of_node(ofs_of_node(dn->node_page), fi) +
 							dn->ofs_in_node;
 
@@ -442,7 +436,7 @@ static int get_data_block_ro(struct inode *inode, sector_t iblock,
 	}
 
 	/* It does not support data allocation */
-	BUG_ON(create);
+	f2fs_bug_on(create);
 
 	if (dn.data_blkaddr != NEW_ADDR && dn.data_blkaddr != NULL_ADDR) {
 		int i;
@@ -560,9 +554,9 @@ static int f2fs_write_data_page(struct page *page,
 		inode_dec_dirty_dents(inode);
 		err = do_write_data_page(page);
 	} else {
-		int ilock = mutex_lock_op(sbi);
+		f2fs_lock_op(sbi);
 		err = do_write_data_page(page);
-		mutex_unlock_op(sbi, ilock);
+		f2fs_unlock_op(sbi);
 		need_balance_fs = true;
 	}
 	if (err == -ENOENT)
@@ -641,7 +635,6 @@ static int f2fs_write_begin(struct file *file, struct address_space *mapping,
 	pgoff_t index = ((unsigned long long) pos) >> PAGE_CACHE_SHIFT;
 	struct dnode_of_data dn;
 	int err = 0;
-	int ilock;
 
 	f2fs_balance_fs(sbi);
 repeat:
@@ -650,7 +643,7 @@ static int f2fs_write_begin(struct file *file, struct address_space *mapping,
 		return -ENOMEM;
 	*pagep = page;
 
-	ilock = mutex_lock_op(sbi);
+	f2fs_lock_op(sbi);
 
 	set_new_dnode(&dn, inode, NULL, NULL, 0);
 	err = get_dnode_of_data(&dn, index, ALLOC_NODE);
@@ -664,7 +657,7 @@ static int f2fs_write_begin(struct file *file, struct address_space *mapping,
 	if (err)
 		goto err;
 
-	mutex_unlock_op(sbi, ilock);
+	f2fs_unlock_op(sbi);
 
 	if ((len == PAGE_CACHE_SIZE) || PageUptodate(page))
 		return 0;
@@ -700,7 +693,7 @@ static int f2fs_write_begin(struct file *file, struct address_space *mapping,
 	return 0;
 
 err:
-	mutex_unlock_op(sbi, ilock);
+	f2fs_unlock_op(sbi);
 	f2fs_put_page(page, 1);
 	return err;
 }
@@ -763,6 +756,8 @@ static int f2fs_set_data_page_dirty(struct page *page)
 	struct address_space *mapping = page->mapping;
 	struct inode *inode = mapping->host;
 
+	trace_f2fs_set_page_dirty(page, DATA);
+
 	SetPageUptodate(page);
 	if (!PageDirty(page)) {
 		__set_page_dirty_nobuffers(page);

fs/f2fs/dir.c

@@ -139,7 +139,7 @@ static struct f2fs_dir_entry *find_in_level(struct inode *dir,
 	bool room = false;
 	int max_slots = 0;
 
-	BUG_ON(level > MAX_DIR_HASH_DEPTH);
+	f2fs_bug_on(level > MAX_DIR_HASH_DEPTH);
 
 	nbucket = dir_buckets(level);
 	nblock = bucket_blocks(level);
@@ -346,7 +346,7 @@ static struct page *init_inode_metadata(struct inode *inode,
 				goto error;
 		}
 
-		err = f2fs_init_acl(inode, dir);
+		err = f2fs_init_acl(inode, dir, page);
 		if (err)
 			goto error;
 

fs/f2fs/f2fs.h

@@ -18,6 +18,13 @@
 #include <linux/crc32.h>
 #include <linux/magic.h>
 #include <linux/kobject.h>
+#include <linux/sched.h>
+
+#ifdef CONFIG_F2FS_CHECK_FS
+#define f2fs_bug_on(condition)	BUG_ON(condition)
+#else
+#define f2fs_bug_on(condition)
+#endif
 
 /*
  * For mount options
@@ -298,6 +305,9 @@ struct f2fs_sm_info {
 	unsigned int main_segments;	/* # of segments in main area */
 	unsigned int reserved_segments;	/* # of reserved segments */
 	unsigned int ovp_segments;	/* # of overprovision segments */
+
+	/* a threshold to reclaim prefree segments */
+	unsigned int rec_prefree_segments;
 };
 
 /*
@@ -317,14 +327,6 @@ enum count_type {
 	NR_COUNT_TYPE,
 };
 
-/*
- * Uses as sbi->fs_lock[NR_GLOBAL_LOCKS].
- * The checkpoint procedure blocks all the locks in this fs_lock array.
- * Some FS operations grab free locks, and if there is no free lock,
- * then wait to grab a lock in a round-robin manner.
- */
-#define NR_GLOBAL_LOCKS	8
-
 /*
  * The below are the page types of bios used in submti_bio().
  * The available types are:
@@ -365,12 +367,12 @@ struct f2fs_sb_info {
 	struct f2fs_checkpoint *ckpt;		/* raw checkpoint pointer */
 	struct inode *meta_inode;		/* cache meta blocks */
 	struct mutex cp_mutex;			/* checkpoint procedure lock */
-	struct mutex fs_lock[NR_GLOBAL_LOCKS];	/* blocking FS operations */
+	struct rw_semaphore cp_rwsem;		/* blocking FS operations */
 	struct mutex node_write;		/* locking node writes */
 	struct mutex writepages;		/* mutex for writepages() */
-	unsigned char next_lock_num;		/* round-robin global locks */
-	int por_doing;				/* recovery is doing or not */
-	int on_build_free_nids;			/* build_free_nids is doing */
+	bool por_doing;				/* recovery is doing or not */
+	bool on_build_free_nids;		/* build_free_nids is doing */
+	wait_queue_head_t cp_wait;
 
 	/* for orphan inode management */
 	struct list_head orphan_inode_list;	/* orphan inode list */
@@ -520,48 +522,24 @@ static inline void clear_ckpt_flags(struct f2fs_checkpoint *cp, unsigned int f)
 	cp->ckpt_flags = cpu_to_le32(ckpt_flags);
 }
 
-static inline void mutex_lock_all(struct f2fs_sb_info *sbi)
+static inline void f2fs_lock_op(struct f2fs_sb_info *sbi)
 {
-	int i;
-
-	for (i = 0; i < NR_GLOBAL_LOCKS; i++) {
-		/*
-		 * This is the only time we take multiple fs_lock[]
-		 * instances; the order is immaterial since we
-		 * always hold cp_mutex, which serializes multiple
-		 * such operations.
-		 */
-		mutex_lock_nest_lock(&sbi->fs_lock[i], &sbi->cp_mutex);
-	}
+	down_read(&sbi->cp_rwsem);
 }
 
-static inline void mutex_unlock_all(struct f2fs_sb_info *sbi)
+static inline void f2fs_unlock_op(struct f2fs_sb_info *sbi)
 {
-	int i = 0;
-	for (; i < NR_GLOBAL_LOCKS; i++)
-		mutex_unlock(&sbi->fs_lock[i]);
+	up_read(&sbi->cp_rwsem);
 }
 
-static inline int mutex_lock_op(struct f2fs_sb_info *sbi)
+static inline void f2fs_lock_all(struct f2fs_sb_info *sbi)
 {
-	unsigned char next_lock = sbi->next_lock_num % NR_GLOBAL_LOCKS;
-	int i = 0;
-
-	for (; i < NR_GLOBAL_LOCKS; i++)
-		if (mutex_trylock(&sbi->fs_lock[i]))
-			return i;
-
-	mutex_lock(&sbi->fs_lock[next_lock]);
-	sbi->next_lock_num++;
-	return next_lock;
+	down_write_nest_lock(&sbi->cp_rwsem, &sbi->cp_mutex);
 }
 
-static inline void mutex_unlock_op(struct f2fs_sb_info *sbi, int ilock)
+static inline void f2fs_unlock_all(struct f2fs_sb_info *sbi)
 {
-	if (ilock < 0)
-		return;
-	BUG_ON(ilock >= NR_GLOBAL_LOCKS);
-	mutex_unlock(&sbi->fs_lock[ilock]);
+	up_write(&sbi->cp_rwsem);
 }
 
 /*
@@ -612,8 +590,8 @@ static inline int dec_valid_block_count(struct f2fs_sb_info *sbi,
 						blkcnt_t count)
 {
 	spin_lock(&sbi->stat_lock);
-	BUG_ON(sbi->total_valid_block_count < (block_t) count);
-	BUG_ON(inode->i_blocks < count);
+	f2fs_bug_on(sbi->total_valid_block_count < (block_t) count);
+	f2fs_bug_on(inode->i_blocks < count);
 	inode->i_blocks -= count;
 	sbi->total_valid_block_count -= (block_t)count;
 	spin_unlock(&sbi->stat_lock);
@@ -745,9 +723,9 @@ static inline void dec_valid_node_count(struct f2fs_sb_info *sbi,
 {
 	spin_lock(&sbi->stat_lock);
 
-	BUG_ON(sbi->total_valid_block_count < count);
-	BUG_ON(sbi->total_valid_node_count < count);
-	BUG_ON(inode->i_blocks < count);
+	f2fs_bug_on(sbi->total_valid_block_count < count);
+	f2fs_bug_on(sbi->total_valid_node_count < count);
+	f2fs_bug_on(inode->i_blocks < count);
 
 	inode->i_blocks -= count;
 	sbi->total_valid_node_count -= count;
@@ -768,7 +746,7 @@ static inline unsigned int valid_node_count(struct f2fs_sb_info *sbi)
 static inline void inc_valid_inode_count(struct f2fs_sb_info *sbi)
 {
 	spin_lock(&sbi->stat_lock);
-	BUG_ON(sbi->total_valid_inode_count == sbi->total_node_count);
+	f2fs_bug_on(sbi->total_valid_inode_count == sbi->total_node_count);
 	sbi->total_valid_inode_count++;
 	spin_unlock(&sbi->stat_lock);
 }
@@ -776,7 +754,7 @@ static inline void inc_valid_inode_count(struct f2fs_sb_info *sbi)
 static inline int dec_valid_inode_count(struct f2fs_sb_info *sbi)
 {
 	spin_lock(&sbi->stat_lock);
-	BUG_ON(!sbi->total_valid_inode_count);
+	f2fs_bug_on(!sbi->total_valid_inode_count);
 	sbi->total_valid_inode_count--;
 	spin_unlock(&sbi->stat_lock);
 	return 0;
@@ -797,7 +775,7 @@ static inline void f2fs_put_page(struct page *page, int unlock)
 		return;
 
 	if (unlock) {
-		BUG_ON(!PageLocked(page));
+		f2fs_bug_on(!PageLocked(page));
 		unlock_page(page);
 	}
 	page_cache_release(page);
@@ -819,6 +797,20 @@ static inline struct kmem_cache *f2fs_kmem_cache_create(const char *name,
 	return kmem_cache_create(name, size, 0, SLAB_RECLAIM_ACCOUNT, ctor);
 }
 
+static inline void *f2fs_kmem_cache_alloc(struct kmem_cache *cachep,
+						gfp_t flags)
+{
+	void *entry;
+retry:
+	entry = kmem_cache_alloc(cachep, flags);
+	if (!entry) {
+		cond_resched();
+		goto retry;
+	}
+
+	return entry;
+}
+
 #define RAW_IS_INODE(p)	((p)->footer.nid == (p)->footer.ino)
 
 static inline bool IS_INODE(struct page *page)
@@ -979,6 +971,7 @@ long f2fs_compat_ioctl(struct file *, unsigned int, unsigned long);
  */
 void f2fs_set_inode_flags(struct inode *);
 struct inode *f2fs_iget(struct super_block *, unsigned long);
+int try_to_free_nats(struct f2fs_sb_info *, int);
 void update_inode(struct inode *, struct page *);
 int update_inode_page(struct inode *);
 int f2fs_write_inode(struct inode *, struct writeback_control *);
@@ -1033,6 +1026,7 @@ void get_node_info(struct f2fs_sb_info *, nid_t, struct node_info *);
 int get_dnode_of_data(struct dnode_of_data *, pgoff_t, int);
 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);
 int remove_inode_page(struct inode *);
 struct page *new_inode_page(struct inode *, const struct qstr *);
 struct page *new_node_page(struct dnode_of_data *, unsigned int, struct page *);
@@ -1059,6 +1053,7 @@ void destroy_node_manager_caches(void);
  * segment.c
  */
 void f2fs_balance_fs(struct f2fs_sb_info *);
+void f2fs_balance_fs_bg(struct f2fs_sb_info *);
 void invalidate_blocks(struct f2fs_sb_info *, block_t);
 void clear_prefree_segments(struct f2fs_sb_info *);
 int npages_for_summary_flush(struct f2fs_sb_info *);
@@ -1172,7 +1167,16 @@ static inline struct f2fs_stat_info *F2FS_STAT(struct f2fs_sb_info *sbi)
 	return (struct f2fs_stat_info*)sbi->stat_info;
 }
 
-#define stat_inc_call_count(si)	((si)->call_count++)
+#define stat_inc_call_count(si)		((si)->call_count++)
+#define stat_inc_bggc_count(sbi)	((sbi)->bg_gc++)
+#define stat_inc_dirty_dir(sbi)		((sbi)->n_dirty_dirs++)
+#define stat_dec_dirty_dir(sbi)		((sbi)->n_dirty_dirs--)
+#define stat_inc_total_hit(sb)		((F2FS_SB(sb))->total_hit_ext++)
+#define stat_inc_read_hit(sb)		((F2FS_SB(sb))->read_hit_ext++)
+#define stat_inc_seg_type(sbi, curseg)					\
+		((sbi)->segment_count[(curseg)->alloc_type]++)
+#define stat_inc_block_count(sbi, curseg)				\
+		((sbi)->block_count[(curseg)->alloc_type]++)
 
 #define stat_inc_seg_count(sbi, type)					\
 	do {								\
@@ -1207,6 +1211,13 @@ void __init f2fs_create_root_stats(void);
 void f2fs_destroy_root_stats(void);
 #else
 #define stat_inc_call_count(si)
+#define stat_inc_bggc_count(si)
+#define stat_inc_dirty_dir(sbi)
+#define stat_dec_dirty_dir(sbi)
+#define stat_inc_total_hit(sb)
+#define stat_inc_read_hit(sb)
+#define stat_inc_seg_type(sbi, curseg)
+#define stat_inc_block_count(sbi, curseg)
 #define stat_inc_seg_count(si, type)
 #define stat_inc_tot_blk_count(si, blks)
 #define stat_inc_data_blk_count(si, blks)

fs/f2fs/file.c

@@ -35,18 +35,18 @@ static int f2fs_vm_page_mkwrite(struct vm_area_struct *vma,
 	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
 	block_t old_blk_addr;
 	struct dnode_of_data dn;
-	int err, ilock;
+	int err;
 
 	f2fs_balance_fs(sbi);
 
 	sb_start_pagefault(inode->i_sb);
 
 	/* block allocation */
-	ilock = mutex_lock_op(sbi);
+	f2fs_lock_op(sbi);
 	set_new_dnode(&dn, inode, NULL, NULL, 0);
 	err = get_dnode_of_data(&dn, page->index, ALLOC_NODE);
 	if (err) {
-		mutex_unlock_op(sbi, ilock);
+		f2fs_unlock_op(sbi);
 		goto out;
 	}
 
@@ -56,12 +56,12 @@ static int f2fs_vm_page_mkwrite(struct vm_area_struct *vma,
 		err = reserve_new_block(&dn);
 		if (err) {
 			f2fs_put_dnode(&dn);
-			mutex_unlock_op(sbi, ilock);
+			f2fs_unlock_op(sbi);
 			goto out;
 		}
 	}
 	f2fs_put_dnode(&dn);
-	mutex_unlock_op(sbi, ilock);
+	f2fs_unlock_op(sbi);
 
 	file_update_time(vma->vm_file);
 	lock_page(page);
@@ -88,6 +88,7 @@ static int f2fs_vm_page_mkwrite(struct vm_area_struct *vma,
 	set_page_dirty(page);
 	SetPageUptodate(page);
 
+	trace_f2fs_vm_page_mkwrite(page, DATA);
 mapped:
 	/* fill the page */
 	wait_on_page_writeback(page);
@@ -188,8 +189,9 @@ int f2fs_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
 			if (ret)
 				goto out;
 		}
-		filemap_fdatawait_range(sbi->node_inode->i_mapping,
-							0, LONG_MAX);
+		ret = wait_on_node_pages_writeback(sbi, inode->i_ino);
+		if (ret)
+			goto out;
 		ret = blkdev_issue_flush(inode->i_sb->s_bdev, GFP_KERNEL, NULL);
 	}
 out:
@@ -270,7 +272,7 @@ static int truncate_blocks(struct inode *inode, u64 from)
 	unsigned int blocksize = inode->i_sb->s_blocksize;
 	struct dnode_of_data dn;
 	pgoff_t free_from;
-	int count = 0, ilock = -1;
+	int count = 0;
 	int err;
 
 	trace_f2fs_truncate_blocks_enter(inode, from);
@@ -278,13 +280,13 @@ static int truncate_blocks(struct inode *inode, u64 from)
 	free_from = (pgoff_t)
 			((from + blocksize - 1) >> (sbi->log_blocksize));
 
-	ilock = mutex_lock_op(sbi);
+	f2fs_lock_op(sbi);
 	set_new_dnode(&dn, inode, NULL, NULL, 0);
 	err = get_dnode_of_data(&dn, free_from, LOOKUP_NODE);
 	if (err) {
 		if (err == -ENOENT)
 			goto free_next;
-		mutex_unlock_op(sbi, ilock);
+		f2fs_unlock_op(sbi);
 		trace_f2fs_truncate_blocks_exit(inode, err);
 		return err;
 	}
@@ -295,7 +297,7 @@ static int truncate_blocks(struct inode *inode, u64 from)
 		count = ADDRS_PER_BLOCK;
 
 	count -= dn.ofs_in_node;
-	BUG_ON(count < 0);
+	f2fs_bug_on(count < 0);
 
 	if (dn.ofs_in_node || IS_INODE(dn.node_page)) {
 		truncate_data_blocks_range(&dn, count);
@@ -305,7 +307,7 @@ static int truncate_blocks(struct inode *inode, u64 from)
 	f2fs_put_dnode(&dn);
 free_next:
 	err = truncate_inode_blocks(inode, free_from);
-	mutex_unlock_op(sbi, ilock);
+	f2fs_unlock_op(sbi);
 
 	/* lastly zero out the first data page */
 	truncate_partial_data_page(inode, from);
@@ -416,16 +418,15 @@ static void fill_zero(struct inode *inode, pgoff_t index,
 {
 	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
 	struct page *page;
-	int ilock;
 
 	if (!len)
 		return;
 
 	f2fs_balance_fs(sbi);
 
-	ilock = mutex_lock_op(sbi);
+	f2fs_lock_op(sbi);
 	page = get_new_data_page(inode, NULL, index, false);
-	mutex_unlock_op(sbi, ilock);
+	f2fs_unlock_op(sbi);
 
 	if (!IS_ERR(page)) {
 		wait_on_page_writeback(page);
@@ -484,7 +485,6 @@ static int punch_hole(struct inode *inode, loff_t offset, loff_t len, int mode)
 			struct address_space *mapping = inode->i_mapping;
 			loff_t blk_start, blk_end;
 			struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
-			int ilock;
 
 			f2fs_balance_fs(sbi);
 
@@ -493,9 +493,9 @@ static int punch_hole(struct inode *inode, loff_t offset, loff_t len, int mode)
 			truncate_inode_pages_range(mapping, blk_start,
 					blk_end - 1);
 
-			ilock = mutex_lock_op(sbi);
+			f2fs_lock_op(sbi);
 			ret = truncate_hole(inode, pg_start, pg_end);
-			mutex_unlock_op(sbi, ilock);
+			f2fs_unlock_op(sbi);
 		}
 	}
 
@@ -529,13 +529,12 @@ static int expand_inode_data(struct inode *inode, loff_t offset,
 
 	for (index = pg_start; index <= pg_end; index++) {
 		struct dnode_of_data dn;
-		int ilock;
 
-		ilock = mutex_lock_op(sbi);
+		f2fs_lock_op(sbi);
 		set_new_dnode(&dn, inode, NULL, NULL, 0);
 		ret = get_dnode_of_data(&dn, index, ALLOC_NODE);
 		if (ret) {
-			mutex_unlock_op(sbi, ilock);
+			f2fs_unlock_op(sbi);
 			break;
 		}
 
@@ -543,12 +542,12 @@ static int expand_inode_data(struct inode *inode, loff_t offset,
 			ret = reserve_new_block(&dn);
 			if (ret) {
 				f2fs_put_dnode(&dn);
-				mutex_unlock_op(sbi, ilock);
+				f2fs_unlock_op(sbi);
 				break;
 			}
 		}
 		f2fs_put_dnode(&dn);
-		mutex_unlock_op(sbi, ilock);
+		f2fs_unlock_op(sbi);
 
 		if (pg_start == pg_end)
 			new_size = offset + len;

fs/f2fs/gc.c

@@ -77,13 +77,15 @@ static int gc_thread_func(void *data)
 		else
 			wait_ms = increase_sleep_time(gc_th, wait_ms);
 
-#ifdef CONFIG_F2FS_STAT_FS
-		sbi->bg_gc++;
-#endif
+		stat_inc_bggc_count(sbi);
 
 		/* if return value is not zero, no victim was selected */
 		if (f2fs_gc(sbi))
 			wait_ms = gc_th->no_gc_sleep_time;
+
+		/* balancing f2fs's metadata periodically */
+		f2fs_balance_fs_bg(sbi);
+
 	} while (!kthread_should_stop());
 	return 0;
 }
@@ -236,8 +238,8 @@ static unsigned int get_cb_cost(struct f2fs_sb_info *sbi, unsigned int segno)
 	return UINT_MAX - ((100 * (100 - u) * age) / (100 + u));
 }
 
-static unsigned int get_gc_cost(struct f2fs_sb_info *sbi, unsigned int segno,
-					struct victim_sel_policy *p)
+static inline unsigned int get_gc_cost(struct f2fs_sb_info *sbi,
+			unsigned int segno, struct victim_sel_policy *p)
 {
 	if (p->alloc_mode == SSR)
 		return get_seg_entry(sbi, segno)->ckpt_valid_blocks;
@@ -293,7 +295,11 @@ static int get_victim_by_default(struct f2fs_sb_info *sbi,
 			}
 			break;
 		}
-		p.offset = ((segno / p.ofs_unit) * p.ofs_unit) + p.ofs_unit;
+
+		p.offset = segno + p.ofs_unit;
+		if (p.ofs_unit > 1)
+			p.offset -= segno % p.ofs_unit;
+
 		secno = GET_SECNO(sbi, segno);
 
 		if (sec_usage_check(sbi, secno))
@@ -306,10 +312,9 @@ static int get_victim_by_default(struct f2fs_sb_info *sbi,
 		if (p.min_cost > cost) {
 			p.min_segno = segno;
 			p.min_cost = cost;
-		}
-
-		if (cost == max_cost)
+		} else if (unlikely(cost == max_cost)) {
 			continue;
+		}
 
 		if (nsearched++ >= p.max_search) {
 			sbi->last_victim[p.gc_mode] = segno;
@@ -358,12 +363,8 @@ static void add_gc_inode(struct inode *inode, struct list_head *ilist)
 		iput(inode);
 		return;
 	}
-repeat:
-	new_ie = kmem_cache_alloc(winode_slab, GFP_NOFS);
-	if (!new_ie) {
-		cond_resched();
-		goto repeat;
-	}
+
+	new_ie = f2fs_kmem_cache_alloc(winode_slab, GFP_NOFS);
 	new_ie->inode = inode;
 	list_add_tail(&new_ie->list, ilist);
 }

fs/f2fs/inode.c

@@ -37,6 +37,31 @@ void f2fs_set_inode_flags(struct inode *inode)
 		inode->i_flags |= S_DIRSYNC;
 }
 
+static void __get_inode_rdev(struct inode *inode, struct f2fs_inode *ri)
+{
+	if (S_ISCHR(inode->i_mode) || S_ISBLK(inode->i_mode) ||
+			S_ISFIFO(inode->i_mode) || S_ISSOCK(inode->i_mode)) {
+		if (ri->i_addr[0])
+			inode->i_rdev = old_decode_dev(le32_to_cpu(ri->i_addr[0]));
+		else
+			inode->i_rdev = new_decode_dev(le32_to_cpu(ri->i_addr[1]));
+	}
+}
+
+static void __set_inode_rdev(struct inode *inode, struct f2fs_inode *ri)
+{
+	if (S_ISCHR(inode->i_mode) || S_ISBLK(inode->i_mode)) {
+		if (old_valid_dev(inode->i_rdev)) {
+			ri->i_addr[0] = cpu_to_le32(old_encode_dev(inode->i_rdev));
+			ri->i_addr[1] = 0;
+		} else {
+			ri->i_addr[0] = 0;
+			ri->i_addr[1] = cpu_to_le32(new_encode_dev(inode->i_rdev));
+			ri->i_addr[2] = 0;
+		}
+	}
+}
+
 static int do_read_inode(struct inode *inode)
 {
 	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
@@ -73,10 +98,6 @@ static int do_read_inode(struct inode *inode)
 	inode->i_ctime.tv_nsec = le32_to_cpu(ri->i_ctime_nsec);
 	inode->i_mtime.tv_nsec = le32_to_cpu(ri->i_mtime_nsec);
 	inode->i_generation = le32_to_cpu(ri->i_generation);
-	if (ri->i_addr[0])
-		inode->i_rdev = old_decode_dev(le32_to_cpu(ri->i_addr[0]));
-	else
-		inode->i_rdev = new_decode_dev(le32_to_cpu(ri->i_addr[1]));
 
 	fi->i_current_depth = le32_to_cpu(ri->i_current_depth);
 	fi->i_xattr_nid = le32_to_cpu(ri->i_xattr_nid);
@@ -84,8 +105,13 @@ static int do_read_inode(struct inode *inode)
 	fi->flags = 0;
 	fi->i_advise = ri->i_advise;
 	fi->i_pino = le32_to_cpu(ri->i_pino);
+
 	get_extent_info(&fi->ext, ri->i_ext);
 	get_inline_info(fi, ri);
+
+	/* get rdev by using inline_info */
+	__get_inode_rdev(inode, ri);
+
 	f2fs_put_page(node_page, 1);
 	return 0;
 }
@@ -179,21 +205,10 @@ void update_inode(struct inode *inode, struct page *node_page)
 	ri->i_pino = cpu_to_le32(F2FS_I(inode)->i_pino);
 	ri->i_generation = cpu_to_le32(inode->i_generation);
 
-	if (S_ISCHR(inode->i_mode) || S_ISBLK(inode->i_mode)) {
-		if (old_valid_dev(inode->i_rdev)) {
-			ri->i_addr[0] =
-				cpu_to_le32(old_encode_dev(inode->i_rdev));
-			ri->i_addr[1] = 0;
-		} else {
-			ri->i_addr[0] = 0;
-			ri->i_addr[1] =
-				cpu_to_le32(new_encode_dev(inode->i_rdev));
-			ri->i_addr[2] = 0;
-		}
-	}
-
+	__set_inode_rdev(inode, ri);
 	set_cold_node(inode, node_page);
 	set_page_dirty(node_page);
+
 	clear_inode_flag(F2FS_I(inode), FI_DIRTY_INODE);
 }
 
@@ -214,7 +229,7 @@ int update_inode_page(struct inode *inode)
 int f2fs_write_inode(struct inode *inode, struct writeback_control *wbc)
 {
 	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
-	int ret, ilock;
+	int ret;
 
 	if (inode->i_ino == F2FS_NODE_INO(sbi) ||
 			inode->i_ino == F2FS_META_INO(sbi))
@@ -227,9 +242,9 @@ int f2fs_write_inode(struct inode *inode, struct writeback_control *wbc)
 	 * We need to lock here to prevent from producing dirty node pages
 	 * during the urgent cleaning time when runing out of free sections.
 	 */
-	ilock = mutex_lock_op(sbi);
+	f2fs_lock_op(sbi);
 	ret = update_inode_page(inode);
-	mutex_unlock_op(sbi, ilock);
+	f2fs_unlock_op(sbi);
 
 	if (wbc)
 		f2fs_balance_fs(sbi);
@@ -243,7 +258,6 @@ int f2fs_write_inode(struct inode *inode, struct writeback_control *wbc)
 void f2fs_evict_inode(struct inode *inode)
 {
 	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
-	int ilock;
 
 	trace_f2fs_evict_inode(inode);
 	truncate_inode_pages(&inode->i_data, 0);
@@ -252,7 +266,7 @@ void f2fs_evict_inode(struct inode *inode)
 			inode->i_ino == F2FS_META_INO(sbi))
 		goto no_delete;
 
-	BUG_ON(atomic_read(&F2FS_I(inode)->dirty_dents));
+	f2fs_bug_on(atomic_read(&F2FS_I(inode)->dirty_dents));
 	remove_dirty_dir_inode(inode);
 
 	if (inode->i_nlink || is_bad_inode(inode))
@@ -265,9 +279,9 @@ void f2fs_evict_inode(struct inode *inode)
 	if (F2FS_HAS_BLOCKS(inode))
 		f2fs_truncate(inode);
 
-	ilock = mutex_lock_op(sbi);
+	f2fs_lock_op(sbi);
 	remove_inode_page(inode);
-	mutex_unlock_op(sbi, ilock);
+	f2fs_unlock_op(sbi);
 
 	sb_end_intwrite(inode->i_sb);
 no_delete:

fs/f2fs/namei.c

@@ -27,19 +27,19 @@ static struct inode *f2fs_new_inode(struct inode *dir, umode_t mode)
 	nid_t ino;
 	struct inode *inode;
 	bool nid_free = false;
-	int err, ilock;
+	int err;
 
 	inode = new_inode(sb);
 	if (!inode)
 		return ERR_PTR(-ENOMEM);
 
-	ilock = mutex_lock_op(sbi);
+	f2fs_lock_op(sbi);
 	if (!alloc_nid(sbi, &ino)) {
-		mutex_unlock_op(sbi, ilock);
+		f2fs_unlock_op(sbi);
 		err = -ENOSPC;
 		goto fail;
 	}
-	mutex_unlock_op(sbi, ilock);
+	f2fs_unlock_op(sbi);
 
 	inode->i_uid = current_fsuid();
 
@@ -115,7 +115,7 @@ static int f2fs_create(struct inode *dir, struct dentry *dentry, umode_t mode,
 	struct f2fs_sb_info *sbi = F2FS_SB(sb);
 	struct inode *inode;
 	nid_t ino = 0;
-	int err, ilock;
+	int err;
 
 	f2fs_balance_fs(sbi);
 
@@ -131,9 +131,9 @@ static int f2fs_create(struct inode *dir, struct dentry *dentry, umode_t mode,
 	inode->i_mapping->a_ops = &f2fs_dblock_aops;
 	ino = inode->i_ino;
 
-	ilock = mutex_lock_op(sbi);
+	f2fs_lock_op(sbi);
 	err = f2fs_add_link(dentry, inode);
-	mutex_unlock_op(sbi, ilock);
+	f2fs_unlock_op(sbi);
 	if (err)
 		goto out;
 
@@ -157,7 +157,7 @@ static int f2fs_link(struct dentry *old_dentry, struct inode *dir,
 	struct inode *inode = old_dentry->d_inode;
 	struct super_block *sb = dir->i_sb;
 	struct f2fs_sb_info *sbi = F2FS_SB(sb);
-	int err, ilock;
+	int err;
 
 	f2fs_balance_fs(sbi);
 
@@ -165,9 +165,9 @@ static int f2fs_link(struct dentry *old_dentry, struct inode *dir,
 	ihold(inode);
 
 	set_inode_flag(F2FS_I(inode), FI_INC_LINK);
-	ilock = mutex_lock_op(sbi);
+	f2fs_lock_op(sbi);
 	err = f2fs_add_link(dentry, inode);
-	mutex_unlock_op(sbi, ilock);
+	f2fs_unlock_op(sbi);
 	if (err)
 		goto out;
 
@@ -220,7 +220,6 @@ static int f2fs_unlink(struct inode *dir, struct dentry *dentry)
 	struct f2fs_dir_entry *de;
 	struct page *page;
 	int err = -ENOENT;
-	int ilock;
 
 	trace_f2fs_unlink_enter(dir, dentry);
 	f2fs_balance_fs(sbi);
@@ -229,16 +228,16 @@ static int f2fs_unlink(struct inode *dir, struct dentry *dentry)
 	if (!de)
 		goto fail;
 
+	f2fs_lock_op(sbi);
 	err = acquire_orphan_inode(sbi);
 	if (err) {
+		f2fs_unlock_op(sbi);
 		kunmap(page);
 		f2fs_put_page(page, 0);
 		goto fail;
 	}
-
-	ilock = mutex_lock_op(sbi);
 	f2fs_delete_entry(de, page, inode);
-	mutex_unlock_op(sbi, ilock);
+	f2fs_unlock_op(sbi);
 
 	/* In order to evict this inode,  we set it dirty */
 	mark_inode_dirty(inode);
@@ -254,7 +253,7 @@ static int f2fs_symlink(struct inode *dir, struct dentry *dentry,
 	struct f2fs_sb_info *sbi = F2FS_SB(sb);
 	struct inode *inode;
 	size_t symlen = strlen(symname) + 1;
-	int err, ilock;
+	int err;
 
 	f2fs_balance_fs(sbi);
 
@@ -265,9 +264,9 @@ static int f2fs_symlink(struct inode *dir, struct dentry *dentry,
 	inode->i_op = &f2fs_symlink_inode_operations;
 	inode->i_mapping->a_ops = &f2fs_dblock_aops;
 
-	ilock = mutex_lock_op(sbi);
+	f2fs_lock_op(sbi);
 	err = f2fs_add_link(dentry, inode);
-	mutex_unlock_op(sbi, ilock);
+	f2fs_unlock_op(sbi);
 	if (err)
 		goto out;
 
@@ -290,7 +289,7 @@ static int f2fs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode)
 {
 	struct f2fs_sb_info *sbi = F2FS_SB(dir->i_sb);
 	struct inode *inode;
-	int err, ilock;
+	int err;
 
 	f2fs_balance_fs(sbi);
 
@@ -304,9 +303,9 @@ static int f2fs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode)
 	mapping_set_gfp_mask(inode->i_mapping, GFP_F2FS_ZERO);
 
 	set_inode_flag(F2FS_I(inode), FI_INC_LINK);
-	ilock = mutex_lock_op(sbi);
+	f2fs_lock_op(sbi);
 	err = f2fs_add_link(dentry, inode);
-	mutex_unlock_op(sbi, ilock);
+	f2fs_unlock_op(sbi);
 	if (err)
 		goto out_fail;
 
@@ -342,7 +341,6 @@ static int f2fs_mknod(struct inode *dir, struct dentry *dentry,
 	struct f2fs_sb_info *sbi = F2FS_SB(sb);
 	struct inode *inode;
 	int err = 0;
-	int ilock;
 
 	if (!new_valid_dev(rdev))
 		return -EINVAL;
@@ -356,9 +354,9 @@ static int f2fs_mknod(struct inode *dir, struct dentry *dentry,
 	init_special_inode(inode, inode->i_mode, rdev);
 	inode->i_op = &f2fs_special_inode_operations;
 
-	ilock = mutex_lock_op(sbi);
+	f2fs_lock_op(sbi);
 	err = f2fs_add_link(dentry, inode);
-	mutex_unlock_op(sbi, ilock);
+	f2fs_unlock_op(sbi);
 	if (err)
 		goto out;
 
@@ -387,7 +385,7 @@ static int f2fs_rename(struct inode *old_dir, struct dentry *old_dentry,
 	struct f2fs_dir_entry *old_dir_entry = NULL;
 	struct f2fs_dir_entry *old_entry;
 	struct f2fs_dir_entry *new_entry;
-	int err = -ENOENT, ilock = -1;
+	int err = -ENOENT;
 
 	f2fs_balance_fs(sbi);
 
@@ -402,7 +400,7 @@ static int f2fs_rename(struct inode *old_dir, struct dentry *old_dentry,
 			goto out_old;
 	}
 
-	ilock = mutex_lock_op(sbi);
+	f2fs_lock_op(sbi);
 
 	if (new_inode) {
 
@@ -467,7 +465,7 @@ static int f2fs_rename(struct inode *old_dir, struct dentry *old_dentry,
 		update_inode_page(old_dir);
 	}
 
-	mutex_unlock_op(sbi, ilock);
+	f2fs_unlock_op(sbi);
 	return 0;
 
 put_out_dir:
@@ -477,7 +475,7 @@ static int f2fs_rename(struct inode *old_dir, struct dentry *old_dentry,
 		kunmap(old_dir_page);
 		f2fs_put_page(old_dir_page, 0);
 	}
-	mutex_unlock_op(sbi, ilock);
+	f2fs_unlock_op(sbi);
 out_old:
 	kunmap(old_page);
 	f2fs_put_page(old_page, 0);

fs/f2fs/node.c

@@ -204,7 +204,7 @@ static void set_node_addr(struct f2fs_sb_info *sbi, struct node_info *ni,
 		}
 		e->ni = *ni;
 		e->checkpointed = true;
-		BUG_ON(ni->blk_addr == NEW_ADDR);
+		f2fs_bug_on(ni->blk_addr == NEW_ADDR);
 	} else if (new_blkaddr == NEW_ADDR) {
 		/*
 		 * when nid is reallocated,
@@ -212,19 +212,19 @@ static void set_node_addr(struct f2fs_sb_info *sbi, struct node_info *ni,
 		 * So, reinitialize it with new information.
 		 */
 		e->ni = *ni;
-		BUG_ON(ni->blk_addr != NULL_ADDR);
+		f2fs_bug_on(ni->blk_addr != NULL_ADDR);
 	}
 
 	if (new_blkaddr == NEW_ADDR)
 		e->checkpointed = false;
 
 	/* sanity check */
-	BUG_ON(nat_get_blkaddr(e) != ni->blk_addr);
-	BUG_ON(nat_get_blkaddr(e) == NULL_ADDR &&
+	f2fs_bug_on(nat_get_blkaddr(e) != ni->blk_addr);
+	f2fs_bug_on(nat_get_blkaddr(e) == NULL_ADDR &&
 			new_blkaddr == NULL_ADDR);
-	BUG_ON(nat_get_blkaddr(e) == NEW_ADDR &&
+	f2fs_bug_on(nat_get_blkaddr(e) == NEW_ADDR &&
 			new_blkaddr == NEW_ADDR);
-	BUG_ON(nat_get_blkaddr(e) != NEW_ADDR &&
+	f2fs_bug_on(nat_get_blkaddr(e) != NEW_ADDR &&
 			nat_get_blkaddr(e) != NULL_ADDR &&
 			new_blkaddr == NEW_ADDR);
 
@@ -240,7 +240,7 @@ static void set_node_addr(struct f2fs_sb_info *sbi, struct node_info *ni,
 	write_unlock(&nm_i->nat_tree_lock);
 }
 
-static int try_to_free_nats(struct f2fs_sb_info *sbi, int nr_shrink)
+int try_to_free_nats(struct f2fs_sb_info *sbi, int nr_shrink)
 {
 	struct f2fs_nm_info *nm_i = NM_I(sbi);
 
@@ -495,10 +495,10 @@ static void truncate_node(struct dnode_of_data *dn)
 
 	get_node_info(sbi, dn->nid, &ni);
 	if (dn->inode->i_blocks == 0) {
-		BUG_ON(ni.blk_addr != NULL_ADDR);
+		f2fs_bug_on(ni.blk_addr != NULL_ADDR);
 		goto invalidate;
 	}
-	BUG_ON(ni.blk_addr == NULL_ADDR);
+	f2fs_bug_on(ni.blk_addr == NULL_ADDR);
 
 	/* Deallocate node address */
 	invalidate_blocks(sbi, ni.blk_addr);
@@ -822,7 +822,7 @@ int remove_inode_page(struct inode *inode)
 	}
 
 	/* 0 is possible, after f2fs_new_inode() is failed */
-	BUG_ON(inode->i_blocks != 0 && inode->i_blocks != 1);
+	f2fs_bug_on(inode->i_blocks != 0 && inode->i_blocks != 1);
 	set_new_dnode(&dn, inode, page, page, ino);
 	truncate_node(&dn);
 	return 0;
@@ -863,7 +863,7 @@ struct page *new_node_page(struct dnode_of_data *dn,
 	get_node_info(sbi, dn->nid, &old_ni);
 
 	/* Reinitialize old_ni with new node page */
-	BUG_ON(old_ni.blk_addr != NULL_ADDR);
+	f2fs_bug_on(old_ni.blk_addr != NULL_ADDR);
 	new_ni = old_ni;
 	new_ni.ino = dn->inode->i_ino;
 	set_node_addr(sbi, &new_ni, NEW_ADDR);
@@ -969,7 +969,7 @@ struct page *get_node_page(struct f2fs_sb_info *sbi, pgoff_t nid)
 		goto repeat;
 	}
 got_it:
-	BUG_ON(nid != nid_of_node(page));
+	f2fs_bug_on(nid != nid_of_node(page));
 	mark_page_accessed(page);
 	return page;
 }
@@ -1148,6 +1148,47 @@ int sync_node_pages(struct f2fs_sb_info *sbi, nid_t ino,
 	return nwritten;
 }
 
+int wait_on_node_pages_writeback(struct f2fs_sb_info *sbi, nid_t ino)
+{
+	struct address_space *mapping = sbi->node_inode->i_mapping;
+	pgoff_t index = 0, end = LONG_MAX;
+	struct pagevec pvec;
+	int nr_pages;
+	int ret2 = 0, ret = 0;
+
+	pagevec_init(&pvec, 0);
+	while ((index <= end) &&
+			(nr_pages = pagevec_lookup_tag(&pvec, mapping, &index,
+			PAGECACHE_TAG_WRITEBACK,
+			min(end - index, (pgoff_t)PAGEVEC_SIZE-1) + 1)) != 0) {
+		unsigned i;
+
+		for (i = 0; i < nr_pages; i++) {
+			struct page *page = pvec.pages[i];
+
+			/* until radix tree lookup accepts end_index */
+			if (page->index > end)
+				continue;
+
+			if (ino && ino_of_node(page) == ino) {
+				wait_on_page_writeback(page);
+				if (TestClearPageError(page))
+					ret = -EIO;
+			}
+		}
+		pagevec_release(&pvec);
+		cond_resched();
+	}
+
+	if (test_and_clear_bit(AS_ENOSPC, &mapping->flags))
+		ret2 = -ENOSPC;
+	if (test_and_clear_bit(AS_EIO, &mapping->flags))
+		ret2 = -EIO;
+	if (!ret)
+		ret = ret2;
+	return ret;
+}
+
 static int f2fs_write_node_page(struct page *page,
 				struct writeback_control *wbc)
 {
@@ -1156,11 +1197,14 @@ static int f2fs_write_node_page(struct page *page,
 	block_t new_addr;
 	struct node_info ni;
 
+	if (sbi->por_doing)
+		goto redirty_out;
+
 	wait_on_page_writeback(page);
 
 	/* get old block addr of this node page */
 	nid = nid_of_node(page);
-	BUG_ON(page->index != nid);
+	f2fs_bug_on(page->index != nid);
 
 	get_node_info(sbi, nid, &ni);
 
@@ -1171,12 +1215,8 @@ static int f2fs_write_node_page(struct page *page,
 		return 0;
 	}
 
-	if (wbc->for_reclaim) {
-		dec_page_count(sbi, F2FS_DIRTY_NODES);
-		wbc->pages_skipped++;
-		set_page_dirty(page);
-		return AOP_WRITEPAGE_ACTIVATE;
-	}
+	if (wbc->for_reclaim)
+		goto redirty_out;
 
 	mutex_lock(&sbi->node_write);
 	set_page_writeback(page);
@@ -1186,6 +1226,12 @@ static int f2fs_write_node_page(struct page *page,
 	mutex_unlock(&sbi->node_write);
 	unlock_page(page);
 	return 0;
+
+redirty_out:
+	dec_page_count(sbi, F2FS_DIRTY_NODES);
+	wbc->pages_skipped++;
+	set_page_dirty(page);
+	return AOP_WRITEPAGE_ACTIVATE;
 }
 
 /*
@@ -1200,11 +1246,8 @@ static int f2fs_write_node_pages(struct address_space *mapping,
 	struct f2fs_sb_info *sbi = F2FS_SB(mapping->host->i_sb);
 	long nr_to_write = wbc->nr_to_write;
 
-	/* First check balancing cached NAT entries */
-	if (try_to_free_nats(sbi, NAT_ENTRY_PER_BLOCK)) {
-		f2fs_sync_fs(sbi->sb, true);
-		return 0;
-	}
+	/* balancing f2fs's metadata in background */
+	f2fs_balance_fs_bg(sbi);
 
 	/* collect a number of dirty node pages and write together */
 	if (get_pages(sbi, F2FS_DIRTY_NODES) < COLLECT_DIRTY_NODES)
@@ -1223,6 +1266,8 @@ static int f2fs_set_node_page_dirty(struct page *page)
 	struct address_space *mapping = page->mapping;
 	struct f2fs_sb_info *sbi = F2FS_SB(mapping->host->i_sb);
 
+	trace_f2fs_set_page_dirty(page, NODE);
+
 	SetPageUptodate(page);
 	if (!PageDirty(page)) {
 		__set_page_dirty_nobuffers(page);
@@ -1291,23 +1336,18 @@ static int add_free_nid(struct f2fs_nm_info *nm_i, nid_t nid, bool build)
 	if (nid == 0)
 		return 0;
 
-	if (!build)
-		goto retry;
-
-	/* do not add allocated nids */
-	read_lock(&nm_i->nat_tree_lock);
-	ne = __lookup_nat_cache(nm_i, nid);
-	if (ne && nat_get_blkaddr(ne) != NULL_ADDR)
-		allocated = true;
-	read_unlock(&nm_i->nat_tree_lock);
-	if (allocated)
-		return 0;
-retry:
-	i = kmem_cache_alloc(free_nid_slab, GFP_NOFS);
-	if (!i) {
-		cond_resched();
-		goto retry;
+	if (build) {
+		/* do not add allocated nids */
+		read_lock(&nm_i->nat_tree_lock);
+		ne = __lookup_nat_cache(nm_i, nid);
+		if (ne && nat_get_blkaddr(ne) != NULL_ADDR)
+			allocated = true;
+		read_unlock(&nm_i->nat_tree_lock);
+		if (allocated)
+			return 0;
 	}
+
+	i = f2fs_kmem_cache_alloc(free_nid_slab, GFP_NOFS);
 	i->nid = nid;
 	i->state = NID_NEW;
 
@@ -1350,7 +1390,7 @@ static void scan_nat_page(struct f2fs_nm_info *nm_i,
 			break;
 
 		blk_addr = le32_to_cpu(nat_blk->entries[i].block_addr);
-		BUG_ON(blk_addr == NEW_ADDR);
+		f2fs_bug_on(blk_addr == NEW_ADDR);
 		if (blk_addr == NULL_ADDR) {
 			if (add_free_nid(nm_i, start_nid, true) < 0)
 				break;
@@ -1421,14 +1461,14 @@ bool alloc_nid(struct f2fs_sb_info *sbi, nid_t *nid)
 
 	/* We should not use stale free nids created by build_free_nids */
 	if (nm_i->fcnt && !sbi->on_build_free_nids) {
-		BUG_ON(list_empty(&nm_i->free_nid_list));
+		f2fs_bug_on(list_empty(&nm_i->free_nid_list));
 		list_for_each(this, &nm_i->free_nid_list) {
 			i = list_entry(this, struct free_nid, list);
 			if (i->state == NID_NEW)
 				break;
 		}
 
-		BUG_ON(i->state != NID_NEW);
+		f2fs_bug_on(i->state != NID_NEW);
 		*nid = i->nid;
 		i->state = NID_ALLOC;
 		nm_i->fcnt--;
@@ -1439,9 +1479,9 @@ bool alloc_nid(struct f2fs_sb_info *sbi, nid_t *nid)
 
 	/* Let's scan nat pages and its caches to get free nids */
 	mutex_lock(&nm_i->build_lock);
-	sbi->on_build_free_nids = 1;
+	sbi->on_build_free_nids = true;
 	build_free_nids(sbi);
-	sbi->on_build_free_nids = 0;
+	sbi->on_build_free_nids = false;
 	mutex_unlock(&nm_i->build_lock);
 	goto retry;
 }
@@ -1456,7 +1496,7 @@ void alloc_nid_done(struct f2fs_sb_info *sbi, nid_t nid)
 
 	spin_lock(&nm_i->free_nid_list_lock);
 	i = __lookup_free_nid_list(nid, &nm_i->free_nid_list);
-	BUG_ON(!i || i->state != NID_ALLOC);
+	f2fs_bug_on(!i || i->state != NID_ALLOC);
 	__del_from_free_nid_list(i);
 	spin_unlock(&nm_i->free_nid_list_lock);
 }
@@ -1474,7 +1514,7 @@ void alloc_nid_failed(struct f2fs_sb_info *sbi, nid_t nid)
 
 	spin_lock(&nm_i->free_nid_list_lock);
 	i = __lookup_free_nid_list(nid, &nm_i->free_nid_list);
-	BUG_ON(!i || i->state != NID_ALLOC);
+	f2fs_bug_on(!i || i->state != NID_ALLOC);
 	if (nm_i->fcnt > 2 * MAX_FREE_NIDS) {
 		__del_from_free_nid_list(i);
 	} else {
@@ -1677,7 +1717,7 @@ void flush_nat_entries(struct f2fs_sb_info *sbi)
 			nat_blk = page_address(page);
 		}
 
-		BUG_ON(!nat_blk);
+		f2fs_bug_on(!nat_blk);
 		raw_ne = nat_blk->entries[nid - start_nid];
 flush_now:
 		new_blkaddr = nat_get_blkaddr(ne);
@@ -1781,11 +1821,11 @@ void destroy_node_manager(struct f2fs_sb_info *sbi)
 	/* destroy free nid list */
 	spin_lock(&nm_i->free_nid_list_lock);
 	list_for_each_entry_safe(i, next_i, &nm_i->free_nid_list, list) {
-		BUG_ON(i->state == NID_ALLOC);
+		f2fs_bug_on(i->state == NID_ALLOC);
 		__del_from_free_nid_list(i);
 		nm_i->fcnt--;
 	}
-	BUG_ON(nm_i->fcnt);
+	f2fs_bug_on(nm_i->fcnt);
 	spin_unlock(&nm_i->free_nid_list_lock);
 
 	/* destroy nat cache */
@@ -1799,7 +1839,7 @@ void destroy_node_manager(struct f2fs_sb_info *sbi)
 			__del_from_nat_cache(nm_i, e);
 		}
 	}
-	BUG_ON(nm_i->nat_cnt);
+	f2fs_bug_on(nm_i->nat_cnt);
 	write_unlock(&nm_i->nat_tree_lock);
 
 	kfree(nm_i->nat_bitmap);

fs/f2fs/recovery.c

@@ -64,24 +64,31 @@ static int recover_dentry(struct page *ipage, struct inode *inode)
 	name.name = raw_inode->i_name;
 retry:
 	de = f2fs_find_entry(dir, &name, &page);
-	if (de && inode->i_ino == le32_to_cpu(de->ino)) {
-		kunmap(page);
-		f2fs_put_page(page, 0);
-		goto out;
-	}
+	if (de && inode->i_ino == le32_to_cpu(de->ino))
+		goto out_unmap_put;
 	if (de) {
 		einode = f2fs_iget(inode->i_sb, le32_to_cpu(de->ino));
 		if (IS_ERR(einode)) {
 			WARN_ON(1);
 			if (PTR_ERR(einode) == -ENOENT)
 				err = -EEXIST;
-			goto out;
+			goto out_unmap_put;
+		}
+		err = acquire_orphan_inode(F2FS_SB(inode->i_sb));
+		if (err) {
+			iput(einode);
+			goto out_unmap_put;
 		}
 		f2fs_delete_entry(de, page, einode);
 		iput(einode);
 		goto retry;
 	}
 	err = __f2fs_add_link(dir, &name, inode);
+	goto out;
+
+out_unmap_put:
+	kunmap(page);
+	f2fs_put_page(page, 0);
 out:
 	f2fs_msg(inode->i_sb, KERN_NOTICE, "recover_inode and its dentry: "
 			"ino = %x, name = %s, dir = %lx, err = %d",
@@ -285,7 +292,6 @@ static int do_recover_data(struct f2fs_sb_info *sbi, struct inode *inode,
 	struct f2fs_summary sum;
 	struct node_info ni;
 	int err = 0, recovered = 0;
-	int ilock;
 
 	start = start_bidx_of_node(ofs_of_node(page), fi);
 	if (IS_INODE(page))
@@ -293,20 +299,20 @@ static int do_recover_data(struct f2fs_sb_info *sbi, struct inode *inode,
 	else
 		end = start + ADDRS_PER_BLOCK;
 
-	ilock = mutex_lock_op(sbi);
+	f2fs_lock_op(sbi);
 	set_new_dnode(&dn, inode, NULL, NULL, 0);
 
 	err = get_dnode_of_data(&dn, start, ALLOC_NODE);
 	if (err) {
-		mutex_unlock_op(sbi, ilock);
+		f2fs_unlock_op(sbi);
 		return err;
 	}
 
 	wait_on_page_writeback(dn.node_page);
 
 	get_node_info(sbi, dn.nid, &ni);
-	BUG_ON(ni.ino != ino_of_node(page));
-	BUG_ON(ofs_of_node(dn.node_page) != ofs_of_node(page));
+	f2fs_bug_on(ni.ino != ino_of_node(page));
+	f2fs_bug_on(ofs_of_node(dn.node_page) != ofs_of_node(page));
 
 	for (; start < end; start++) {
 		block_t src, dest;
@@ -316,9 +322,9 @@ static int do_recover_data(struct f2fs_sb_info *sbi, struct inode *inode,
 
 		if (src != dest && dest != NEW_ADDR && dest != NULL_ADDR) {
 			if (src == NULL_ADDR) {
-				int err = reserve_new_block(&dn);
+				err = reserve_new_block(&dn);
 				/* We should not get -ENOSPC */
-				BUG_ON(err);
+				f2fs_bug_on(err);
 			}
 
 			/* Check the previous node page having this index */
@@ -349,7 +355,7 @@ static int do_recover_data(struct f2fs_sb_info *sbi, struct inode *inode,
 	recover_node_page(sbi, dn.node_page, &sum, &ni, blkaddr);
 err:
 	f2fs_put_dnode(&dn);
-	mutex_unlock_op(sbi, ilock);
+	f2fs_unlock_op(sbi);
 
 	f2fs_msg(sbi->sb, KERN_NOTICE, "recover_data: ino = %lx, "
 			"recovered_data = %d blocks, err = %d",
@@ -419,6 +425,7 @@ int recover_fsync_data(struct f2fs_sb_info *sbi)
 {
 	struct list_head inode_list;
 	int err;
+	bool need_writecp = false;
 
 	fsync_entry_slab = f2fs_kmem_cache_create("f2fs_fsync_inode_entry",
 			sizeof(struct fsync_inode_entry), NULL);
@@ -428,7 +435,7 @@ int recover_fsync_data(struct f2fs_sb_info *sbi)
 	INIT_LIST_HEAD(&inode_list);
 
 	/* step #1: find fsynced inode numbers */
-	sbi->por_doing = 1;
+	sbi->por_doing = true;
 	err = find_fsync_dnodes(sbi, &inode_list);
 	if (err)
 		goto out;
@@ -436,14 +443,16 @@ int recover_fsync_data(struct f2fs_sb_info *sbi)
 	if (list_empty(&inode_list))
 		goto out;
 
+	need_writecp = true;
+
 	/* step #2: recover data */
 	err = recover_data(sbi, &inode_list, CURSEG_WARM_NODE);
-	BUG_ON(!list_empty(&inode_list));
+	f2fs_bug_on(!list_empty(&inode_list));
 out:
 	destroy_fsync_dnodes(&inode_list);
 	kmem_cache_destroy(fsync_entry_slab);
-	sbi->por_doing = 0;
-	if (!err)
+	sbi->por_doing = false;
+	if (!err && need_writecp)
 		write_checkpoint(sbi, false);
 	return err;
 }

fs/f2fs/segment.h

@@ -14,6 +14,8 @@
 #define NULL_SEGNO			((unsigned int)(~0))
 #define NULL_SECNO			((unsigned int)(~0))
 
+#define DEF_RECLAIM_PREFREE_SEGMENTS	100	/* 200MB of prefree segments */
+
 /* L: Logical segment # in volume, R: Relative segment # in main area */
 #define GET_L2R_SEGNO(free_i, segno)	(segno - free_i->start_segno)
 #define GET_R2L_SEGNO(free_i, segno)	(segno + free_i->start_segno)
@@ -90,6 +92,8 @@
 	(blk_addr << ((sbi)->log_blocksize - F2FS_LOG_SECTOR_SIZE))
 #define SECTOR_TO_BLOCK(sbi, sectors)					\
 	(sectors >> ((sbi)->log_blocksize - F2FS_LOG_SECTOR_SIZE))
+#define MAX_BIO_BLOCKS(max_hw_blocks)					\
+	(min((int)max_hw_blocks, BIO_MAX_PAGES))
 
 /* during checkpoint, bio_private is used to synchronize the last bio */
 struct bio_private {
@@ -470,6 +474,11 @@ static inline bool has_not_enough_free_secs(struct f2fs_sb_info *sbi, int freed)
 						reserved_sections(sbi)));
 }
 
+static inline bool excess_prefree_segs(struct f2fs_sb_info *sbi)
+{
+	return (prefree_segments(sbi) > SM_I(sbi)->rec_prefree_segments);
+}
+
 static inline int utilization(struct f2fs_sb_info *sbi)
 {
 	return div_u64((u64)valid_user_blocks(sbi) * 100, sbi->user_block_count);
@@ -513,16 +522,13 @@ static inline unsigned short curseg_blkoff(struct f2fs_sb_info *sbi, int type)
 	return curseg->next_blkoff;
 }
 
+#ifdef CONFIG_F2FS_CHECK_FS
 static inline void check_seg_range(struct f2fs_sb_info *sbi, unsigned int segno)
 {
 	unsigned int end_segno = SM_I(sbi)->segment_count - 1;
 	BUG_ON(segno > end_segno);
 }
 
-/*
- * This function is used for only debugging.
- * NOTE: In future, we have to remove this function.
- */
 static inline void verify_block_addr(struct f2fs_sb_info *sbi, block_t blk_addr)
 {
 	struct f2fs_sm_info *sm_info = SM_I(sbi);
@@ -541,8 +547,9 @@ static inline void check_block_count(struct f2fs_sb_info *sbi,
 {
 	struct f2fs_sm_info *sm_info = SM_I(sbi);
 	unsigned int end_segno = sm_info->segment_count - 1;
+	bool is_valid  = test_bit_le(0, raw_sit->valid_map) ? true : false;
 	int valid_blocks = 0;
-	int i;
+	int cur_pos = 0, next_pos;
 
 	/* check segment usage */
 	BUG_ON(GET_SIT_VBLOCKS(raw_sit) > sbi->blocks_per_seg);
@@ -551,11 +558,26 @@ static inline void check_block_count(struct f2fs_sb_info *sbi,
 	BUG_ON(segno > end_segno);
 
 	/* check bitmap with valid block count */
-	for (i = 0; i < sbi->blocks_per_seg; i++)
-		if (f2fs_test_bit(i, raw_sit->valid_map))
-			valid_blocks++;
+	do {
+		if (is_valid) {
+			next_pos = find_next_zero_bit_le(&raw_sit->valid_map,
+					sbi->blocks_per_seg,
+					cur_pos);
+			valid_blocks += next_pos - cur_pos;
+		} else
+			next_pos = find_next_bit_le(&raw_sit->valid_map,
+					sbi->blocks_per_seg,
+					cur_pos);
+		cur_pos = next_pos;
+		is_valid = !is_valid;
+	} while (cur_pos < sbi->blocks_per_seg);
 	BUG_ON(GET_SIT_VBLOCKS(raw_sit) != valid_blocks);
 }
+#else
+#define check_seg_range(sbi, segno)
+#define verify_block_addr(sbi, blk_addr)
+#define check_block_count(sbi, segno, raw_sit)
+#endif
 
 static inline pgoff_t current_sit_addr(struct f2fs_sb_info *sbi,
 						unsigned int start)

fs/f2fs/super.c

@@ -43,7 +43,9 @@ enum {
 	Opt_disable_roll_forward,
 	Opt_discard,
 	Opt_noheap,
+	Opt_user_xattr,
 	Opt_nouser_xattr,
+	Opt_acl,
 	Opt_noacl,
 	Opt_active_logs,
 	Opt_disable_ext_identify,
@@ -56,7 +58,9 @@ static match_table_t f2fs_tokens = {
 	{Opt_disable_roll_forward, "disable_roll_forward"},
 	{Opt_discard, "discard"},
 	{Opt_noheap, "no_heap"},
+	{Opt_user_xattr, "user_xattr"},
 	{Opt_nouser_xattr, "nouser_xattr"},
+	{Opt_acl, "acl"},
 	{Opt_noacl, "noacl"},
 	{Opt_active_logs, "active_logs=%u"},
 	{Opt_disable_ext_identify, "disable_ext_identify"},
@@ -65,24 +69,40 @@ static match_table_t f2fs_tokens = {
 };
 
 /* Sysfs support for f2fs */
+enum {
+	GC_THREAD,	/* struct f2fs_gc_thread */
+	SM_INFO,	/* struct f2fs_sm_info */
+};
+
 struct f2fs_attr {
 	struct attribute attr;
 	ssize_t (*show)(struct f2fs_attr *, struct f2fs_sb_info *, char *);
 	ssize_t (*store)(struct f2fs_attr *, struct f2fs_sb_info *,
 			 const char *, size_t);
+	int struct_type;
 	int offset;
 };
 
+static unsigned char *__struct_ptr(struct f2fs_sb_info *sbi, int struct_type)
+{
+	if (struct_type == GC_THREAD)
+		return (unsigned char *)sbi->gc_thread;
+	else if (struct_type == SM_INFO)
+		return (unsigned char *)SM_I(sbi);
+	return NULL;
+}
+
 static ssize_t f2fs_sbi_show(struct f2fs_attr *a,
 			struct f2fs_sb_info *sbi, char *buf)
 {
-	struct f2fs_gc_kthread *gc_kth = sbi->gc_thread;
+	unsigned char *ptr = NULL;
 	unsigned int *ui;
 
-	if (!gc_kth)
+	ptr = __struct_ptr(sbi, a->struct_type);
+	if (!ptr)
 		return -EINVAL;
 
-	ui = (unsigned int *)(((char *)gc_kth) + a->offset);
+	ui = (unsigned int *)(ptr + a->offset);
 
 	return snprintf(buf, PAGE_SIZE, "%u\n", *ui);
 }
@@ -91,15 +111,16 @@ static ssize_t f2fs_sbi_store(struct f2fs_attr *a,
 			struct f2fs_sb_info *sbi,
 			const char *buf, size_t count)
 {
-	struct f2fs_gc_kthread *gc_kth = sbi->gc_thread;
+	unsigned char *ptr;
 	unsigned long t;
 	unsigned int *ui;
 	ssize_t ret;
 
-	if (!gc_kth)
+	ptr = __struct_ptr(sbi, a->struct_type);
+	if (!ptr)
 		return -EINVAL;
 
-	ui = (unsigned int *)(((char *)gc_kth) + a->offset);
+	ui = (unsigned int *)(ptr + a->offset);
 
 	ret = kstrtoul(skip_spaces(buf), 0, &t);
 	if (ret < 0)
@@ -135,21 +156,25 @@ static void f2fs_sb_release(struct kobject *kobj)
 	complete(&sbi->s_kobj_unregister);
 }
 
-#define F2FS_ATTR_OFFSET(_name, _mode, _show, _store, _elname) \
+#define F2FS_ATTR_OFFSET(_struct_type, _name, _mode, _show, _store, _offset) \
 static struct f2fs_attr f2fs_attr_##_name = {			\
 	.attr = {.name = __stringify(_name), .mode = _mode },	\
 	.show	= _show,					\
 	.store	= _store,					\
-	.offset = offsetof(struct f2fs_gc_kthread, _elname),	\
+	.struct_type = _struct_type,				\
+	.offset = _offset					\
 }
 
-#define F2FS_RW_ATTR(name, elname)	\
-	F2FS_ATTR_OFFSET(name, 0644, f2fs_sbi_show, f2fs_sbi_store, elname)
+#define F2FS_RW_ATTR(struct_type, struct_name, name, elname)	\
+	F2FS_ATTR_OFFSET(struct_type, name, 0644,		\
+		f2fs_sbi_show, f2fs_sbi_store,			\
+		offsetof(struct struct_name, elname))
 
-F2FS_RW_ATTR(gc_min_sleep_time, min_sleep_time);
-F2FS_RW_ATTR(gc_max_sleep_time, max_sleep_time);
-F2FS_RW_ATTR(gc_no_gc_sleep_time, no_gc_sleep_time);
-F2FS_RW_ATTR(gc_idle, gc_idle);
+F2FS_RW_ATTR(GC_THREAD, f2fs_gc_kthread, gc_min_sleep_time, min_sleep_time);
+F2FS_RW_ATTR(GC_THREAD, f2fs_gc_kthread, gc_max_sleep_time, max_sleep_time);
+F2FS_RW_ATTR(GC_THREAD, f2fs_gc_kthread, gc_no_gc_sleep_time, no_gc_sleep_time);
+F2FS_RW_ATTR(GC_THREAD, f2fs_gc_kthread, gc_idle, gc_idle);
+F2FS_RW_ATTR(SM_INFO, f2fs_sm_info, reclaim_segments, rec_prefree_segments);
 
 #define ATTR_LIST(name) (&f2fs_attr_##name.attr)
 static struct attribute *f2fs_attrs[] = {
@@ -157,6 +182,7 @@ static struct attribute *f2fs_attrs[] = {
 	ATTR_LIST(gc_max_sleep_time),
 	ATTR_LIST(gc_no_gc_sleep_time),
 	ATTR_LIST(gc_idle),
+	ATTR_LIST(reclaim_segments),
 	NULL,
 };
 
@@ -237,6 +263,9 @@ static int parse_options(struct super_block *sb, char *options)
 			set_opt(sbi, NOHEAP);
 			break;
 #ifdef CONFIG_F2FS_FS_XATTR
+		case Opt_user_xattr:
+			set_opt(sbi, XATTR_USER);
+			break;
 		case Opt_nouser_xattr:
 			clear_opt(sbi, XATTR_USER);
 			break;
@@ -244,6 +273,10 @@ static int parse_options(struct super_block *sb, char *options)
 			set_opt(sbi, INLINE_XATTR);
 			break;
 #else
+		case Opt_user_xattr:
+			f2fs_msg(sb, KERN_INFO,
+				"user_xattr options not supported");
+			break;
 		case Opt_nouser_xattr:
 			f2fs_msg(sb, KERN_INFO,
 				"nouser_xattr options not supported");
@@ -254,10 +287,16 @@ static int parse_options(struct super_block *sb, char *options)
 			break;
 #endif
 #ifdef CONFIG_F2FS_FS_POSIX_ACL
+		case Opt_acl:
+			set_opt(sbi, POSIX_ACL);
+			break;
 		case Opt_noacl:
 			clear_opt(sbi, POSIX_ACL);
 			break;
 #else
+		case Opt_acl:
+			f2fs_msg(sb, KERN_INFO, "acl options not supported");
+			break;
 		case Opt_noacl:
 			f2fs_msg(sb, KERN_INFO, "noacl options not supported");
 			break;
@@ -355,7 +394,9 @@ static void f2fs_put_super(struct super_block *sb)
 	f2fs_destroy_stats(sbi);
 	stop_gc_thread(sbi);
 
-	write_checkpoint(sbi, true);
+	/* We don't need to do checkpoint when it's clean */
+	if (sbi->s_dirty && get_pages(sbi, F2FS_DIRTY_NODES))
+		write_checkpoint(sbi, true);
 
 	iput(sbi->node_inode);
 	iput(sbi->meta_inode);
@@ -727,30 +768,47 @@ static void init_sb_info(struct f2fs_sb_info *sbi)
 		atomic_set(&sbi->nr_pages[i], 0);
 }
 
-static int validate_superblock(struct super_block *sb,
-		struct f2fs_super_block **raw_super,
-		struct buffer_head **raw_super_buf, sector_t block)
+/*
+ * Read f2fs raw super block.
+ * Because we have two copies of super block, so read the first one at first,
+ * if the first one is invalid, move to read the second one.
+ */
+static int read_raw_super_block(struct super_block *sb,
+			struct f2fs_super_block **raw_super,
+			struct buffer_head **raw_super_buf)
 {
-	const char *super = (block == 0 ? "first" : "second");
+	int block = 0;
 
-	/* read f2fs raw super block */
+retry:
 	*raw_super_buf = sb_bread(sb, block);
 	if (!*raw_super_buf) {
-		f2fs_msg(sb, KERN_ERR, "unable to read %s superblock",
-				super);
-		return -EIO;
+		f2fs_msg(sb, KERN_ERR, "Unable to read %dth superblock",
+				block + 1);
+		if (block == 0) {
+			block++;
+			goto retry;
+		} else {
+			return -EIO;
+		}
 	}
 
 	*raw_super = (struct f2fs_super_block *)
 		((char *)(*raw_super_buf)->b_data + F2FS_SUPER_OFFSET);
 
 	/* sanity checking of raw super */
-	if (!sanity_check_raw_super(sb, *raw_super))
-		return 0;
+	if (sanity_check_raw_super(sb, *raw_super)) {
+		brelse(*raw_super_buf);
+		f2fs_msg(sb, KERN_ERR, "Can't find a valid F2FS filesystem "
+				"in %dth superblock", block + 1);
+		if(block == 0) {
+			block++;
+			goto retry;
+		} else {
+			return -EINVAL;
+		}
+	}
 
-	f2fs_msg(sb, KERN_ERR, "Can't find a valid F2FS filesystem "
-				"in %s superblock", super);
-	return -EINVAL;
+	return 0;
 }
 
 static int f2fs_fill_super(struct super_block *sb, void *data, int silent)
@@ -760,7 +818,6 @@ static int f2fs_fill_super(struct super_block *sb, void *data, int silent)
 	struct buffer_head *raw_super_buf;
 	struct inode *root;
 	long err = -EINVAL;
-	int i;
 
 	/* allocate memory for f2fs-specific super block info */
 	sbi = kzalloc(sizeof(struct f2fs_sb_info), GFP_KERNEL);
@@ -773,14 +830,10 @@ static int f2fs_fill_super(struct super_block *sb, void *data, int silent)
 		goto free_sbi;
 	}
 
-	err = validate_superblock(sb, &raw_super, &raw_super_buf, 0);
-	if (err) {
-		brelse(raw_super_buf);
-		/* check secondary superblock when primary failed */
-		err = validate_superblock(sb, &raw_super, &raw_super_buf, 1);
-		if (err)
-			goto free_sb_buf;
-	}
+	err = read_raw_super_block(sb, &raw_super, &raw_super_buf);
+	if (err)
+		goto free_sbi;
+
 	sb->s_fs_info = sbi;
 	/* init some FS parameters */
 	sbi->active_logs = NR_CURSEG_TYPE;
@@ -818,12 +871,12 @@ static int f2fs_fill_super(struct super_block *sb, void *data, int silent)
 	mutex_init(&sbi->gc_mutex);
 	mutex_init(&sbi->writepages);
 	mutex_init(&sbi->cp_mutex);
-	for (i = 0; i < NR_GLOBAL_LOCKS; i++)
-		mutex_init(&sbi->fs_lock[i]);
 	mutex_init(&sbi->node_write);
-	sbi->por_doing = 0;
+	sbi->por_doing = false;
 	spin_lock_init(&sbi->stat_lock);
 	init_rwsem(&sbi->bio_sem);
+	init_rwsem(&sbi->cp_rwsem);
+	init_waitqueue_head(&sbi->cp_wait);
 	init_sb_info(sbi);
 
 	/* get an inode for meta space */
@@ -922,12 +975,12 @@ static int f2fs_fill_super(struct super_block *sb, void *data, int silent)
 		/* After POR, we can run background GC thread.*/
 		err = start_gc_thread(sbi);
 		if (err)
-			goto fail;
+			goto free_gc;
 	}
 
 	err = f2fs_build_stats(sbi);
 	if (err)
-		goto fail;
+		goto free_gc;
 
 	if (f2fs_proc_root)
 		sbi->s_proc = proc_mkdir(sb->s_id, f2fs_proc_root);
@@ -953,6 +1006,12 @@ static int f2fs_fill_super(struct super_block *sb, void *data, int silent)
 
 	return 0;
 fail:
+	if (sbi->s_proc) {
+		remove_proc_entry("segment_info", sbi->s_proc);
+		remove_proc_entry(sb->s_id, f2fs_proc_root);
+	}
+	f2fs_destroy_stats(sbi);
+free_gc:
 	stop_gc_thread(sbi);
 free_root_inode:
 	dput(sb->s_root);

fs/f2fs/xattr.c

@@ -154,6 +154,9 @@ static int f2fs_xattr_advise_set(struct dentry *dentry, const char *name,
 }
 
 #ifdef CONFIG_F2FS_FS_SECURITY
+static int __f2fs_setxattr(struct inode *inode, int name_index,
+			const char *name, const void *value, size_t value_len,
+			struct page *ipage);
 static int f2fs_initxattrs(struct inode *inode, const struct xattr *xattr_array,
 		void *page)
 {
@@ -161,7 +164,7 @@ static int f2fs_initxattrs(struct inode *inode, const struct xattr *xattr_array,
 	int err = 0;
 
 	for (xattr = xattr_array; xattr->name != NULL; xattr++) {
-		err = f2fs_setxattr(inode, F2FS_XATTR_INDEX_SECURITY,
+		err = __f2fs_setxattr(inode, F2FS_XATTR_INDEX_SECURITY,
 				xattr->name, xattr->value,
 				xattr->value_len, (struct page *)page);
 		if (err < 0)
@@ -369,7 +372,7 @@ static inline int write_all_xattrs(struct inode *inode, __u32 hsize,
 			alloc_nid_failed(sbi, new_nid);
 			return PTR_ERR(xpage);
 		}
-		BUG_ON(new_nid);
+		f2fs_bug_on(new_nid);
 	} else {
 		struct dnode_of_data dn;
 		set_new_dnode(&dn, inode, NULL, NULL, new_nid);
@@ -469,16 +472,15 @@ ssize_t f2fs_listxattr(struct dentry *dentry, char *buffer, size_t buffer_size)
 	return error;
 }
 
-int f2fs_setxattr(struct inode *inode, int name_index, const char *name,
-			const void *value, size_t value_len, struct page *ipage)
+static int __f2fs_setxattr(struct inode *inode, int name_index,
+			const char *name, const void *value, size_t value_len,
+			struct page *ipage)
 {
-	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
 	struct f2fs_inode_info *fi = F2FS_I(inode);
 	struct f2fs_xattr_entry *here, *last;
 	void *base_addr;
 	int found, newsize;
 	size_t name_len;
-	int ilock;
 	__u32 new_hsize;
 	int error = -ENOMEM;
 
@@ -493,10 +495,6 @@ int f2fs_setxattr(struct inode *inode, int name_index, const char *name,
 	if (name_len > F2FS_NAME_LEN || value_len > MAX_VALUE_LEN(inode))
 		return -ERANGE;
 
-	f2fs_balance_fs(sbi);
-
-	ilock = mutex_lock_op(sbi);
-
 	base_addr = read_all_xattrs(inode, ipage);
 	if (!base_addr)
 		goto exit;
@@ -522,7 +520,7 @@ int f2fs_setxattr(struct inode *inode, int name_index, const char *name,
 		 */
 		free = MIN_OFFSET(inode) - ((char *)last - (char *)base_addr);
 		if (found)
-			free = free - ENTRY_SIZE(here);
+			free = free + ENTRY_SIZE(here);
 
 		if (free < newsize) {
 			error = -ENOSPC;
@@ -578,7 +576,21 @@ int f2fs_setxattr(struct inode *inode, int name_index, const char *name,
 	else
 		update_inode_page(inode);
 exit:
-	mutex_unlock_op(sbi, ilock);
 	kzfree(base_addr);
 	return error;
 }
+
+int f2fs_setxattr(struct inode *inode, int name_index, const char *name,
+			const void *value, size_t value_len, struct page *ipage)
+{
+	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
+	int err;
+
+	f2fs_balance_fs(sbi);
+
+	f2fs_lock_op(sbi);
+	err = __f2fs_setxattr(inode, name_index, name, value, value_len, ipage);
+	f2fs_unlock_op(sbi);
+
+	return err;
+}

include/trace/events/f2fs.h

@@ -36,6 +36,11 @@
 		{ CURSEG_COLD_NODE, 	"Cold NODE" },			\
 		{ NO_CHECK_TYPE, 	"No TYPE" })
 
+#define show_file_type(type)						\
+	__print_symbolic(type,						\
+		{ 0,		"FILE" },				\
+		{ 1,		"DIR" })
+
 #define show_gc_type(type)						\
 	__print_symbolic(type,						\
 		{ FG_GC,	"Foreground GC" },			\
@@ -623,6 +628,52 @@ TRACE_EVENT(f2fs_do_submit_bio,
 		__entry->size)
 );
 
+DECLARE_EVENT_CLASS(f2fs__page,
+
+	TP_PROTO(struct page *page, int type),
+
+	TP_ARGS(page, type),
+
+	TP_STRUCT__entry(
+		__field(dev_t,	dev)
+		__field(ino_t,	ino)
+		__field(int, type)
+		__field(int, dir)
+		__field(pgoff_t, index)
+		__field(int, dirty)
+	),
+
+	TP_fast_assign(
+		__entry->dev	= page->mapping->host->i_sb->s_dev;
+		__entry->ino	= page->mapping->host->i_ino;
+		__entry->type	= type;
+		__entry->dir	= S_ISDIR(page->mapping->host->i_mode);
+		__entry->index	= page->index;
+		__entry->dirty	= PageDirty(page);
+	),
+
+	TP_printk("dev = (%d,%d), ino = %lu, %s, %s, index = %lu, dirty = %d",
+		show_dev_ino(__entry),
+		show_block_type(__entry->type),
+		show_file_type(__entry->dir),
+		(unsigned long)__entry->index,
+		__entry->dirty)
+);
+
+DEFINE_EVENT(f2fs__page, f2fs_set_page_dirty,
+
+	TP_PROTO(struct page *page, int type),
+
+	TP_ARGS(page, type)
+);
+
+DEFINE_EVENT(f2fs__page, f2fs_vm_page_mkwrite,
+
+	TP_PROTO(struct page *page, int type),
+
+	TP_ARGS(page, type)
+);
+
 TRACE_EVENT(f2fs_submit_write_page,
 
 	TP_PROTO(struct page *page, block_t blk_addr, int type),