Merge tag 'f2fs-for-5.18' of git://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk/f2fs

Pull f2fs updates from Jaegeuk Kim:
 "In this cycle, f2fs has some performance improvements for Android
  workloads such as using read-unfair rwsems and adding some sysfs
  entries to control GCs and discard commands in more details. In
  addtiion, it has some tunings to improve the recovery speed after
  sudden power-cut.

  Enhancement:
   - add reader-unfair rwsems with F2FS_UNFAIR_RWSEM: will replace with
     generic API support
   - adjust to make the readahead/recovery flow more efficiently
   - sysfs entries to control issue speeds of GCs and Discard commands
   - enable idmapped mounts

  Bug fix:
   - correct wrong error handling routines
   - fix missing conditions in quota
   - fix a potential deadlock between writeback and block plug routines
   - fix a deadlock btween freezefs and evict_inode

  We've added some boundary checks to avoid kernel panics on corrupted
  images, and several minor code clean-ups"

* tag 'f2fs-for-5.18' of git://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk/f2fs: (27 commits)
  f2fs: fix to do sanity check on .cp_pack_total_block_count
  f2fs: make gc_urgent and gc_segment_mode sysfs node readable
  f2fs: use aggressive GC policy during f2fs_disable_checkpoint()
  f2fs: fix compressed file start atomic write may cause data corruption
  f2fs: initialize sbi->gc_mode explicitly
  f2fs: introduce gc_urgent_mid mode
  f2fs: compress: fix to print raw data size in error path of lz4 decompression
  f2fs: remove redundant parameter judgment
  f2fs: use spin_lock to avoid hang
  f2fs: don't get FREEZE lock in f2fs_evict_inode in frozen fs
  f2fs: remove unnecessary read for F2FS_FITS_IN_INODE
  f2fs: introduce F2FS_UNFAIR_RWSEM to support unfair rwsem
  f2fs: avoid an infinite loop in f2fs_sync_dirty_inodes
  f2fs: fix to do sanity check on curseg->alloc_type
  f2fs: fix to avoid potential deadlock
  f2fs: quota: fix loop condition at f2fs_quota_sync()
  f2fs: Restore rwsem lockdep support
  f2fs: fix missing free nid in f2fs_handle_failed_inode
  f2fs: support idmapped mounts
  f2fs: add a way to limit roll forward recovery time
  ...

Documentation/ABI/testing/sysfs-fs-f2fs

@@ -55,8 +55,9 @@ Description:	Controls the in-place-update policy.
 		0x04  F2FS_IPU_UTIL
 		0x08  F2FS_IPU_SSR_UTIL
 		0x10  F2FS_IPU_FSYNC
-		0x20  F2FS_IPU_ASYNC,
+		0x20  F2FS_IPU_ASYNC
 		0x40  F2FS_IPU_NOCACHE
+		0x80  F2FS_IPU_HONOR_OPU_WRITE
 		====  =================
 
 		Refer segment.h for details.
@@ -98,6 +99,33 @@ Description:	Controls the issue rate of discard commands that consist of small
 		checkpoint is triggered, and issued during the checkpoint.
 		By default, it is disabled with 0.
 
+What:		/sys/fs/f2fs/<disk>/max_discard_request
+Date:		December 2021
+Contact:	"Konstantin Vyshetsky" <vkon@google.com>
+Description:	Controls the number of discards a thread will issue at a time.
+		Higher number will allow the discard thread to finish its work
+		faster, at the cost of higher latency for incomming I/O.
+
+What:		/sys/fs/f2fs/<disk>/min_discard_issue_time
+Date:		December 2021
+Contact:	"Konstantin Vyshetsky" <vkon@google.com>
+Description:	Controls the interval the discard thread will wait between
+		issuing discard requests when there are discards to be issued and
+		no I/O aware interruptions occur.
+
+What:		/sys/fs/f2fs/<disk>/mid_discard_issue_time
+Date:		December 2021
+Contact:	"Konstantin Vyshetsky" <vkon@google.com>
+Description:	Controls the interval the discard thread will wait between
+		issuing discard requests when there are discards to be issued and
+		an I/O aware interruption occurs.
+
+What:		/sys/fs/f2fs/<disk>/max_discard_issue_time
+Date:		December 2021
+Contact:	"Konstantin Vyshetsky" <vkon@google.com>
+Description:	Controls the interval the discard thread will wait when there are
+		no discard operations to be issued.
+
 What:		/sys/fs/f2fs/<disk>/discard_granularity
 Date:		July 2017
 Contact:	"Chao Yu" <yuchao0@huawei.com>
@@ -269,11 +297,16 @@ Description:	Shows current reserved blocks in system, it may be temporarily
 What:		/sys/fs/f2fs/<disk>/gc_urgent
 Date:		August 2017
 Contact:	"Jaegeuk Kim" <jaegeuk@kernel.org>
-Description:	Do background GC aggressively when set. When gc_urgent = 1,
-		background thread starts to do GC by given gc_urgent_sleep_time
-		interval. When gc_urgent = 2, F2FS will lower the bar of
-		checking idle in order to process outstanding discard commands
-		and GC a little bit aggressively. It is set to 0 by default.
+Description:	Do background GC aggressively when set. Set to 0 by default.
+		gc urgent high(1): does GC forcibly in a period of given
+		gc_urgent_sleep_time and ignores I/O idling check. uses greedy
+		GC approach and turns SSR mode on.
+		gc urgent low(2): lowers the bar of checking I/O idling in
+		order to process outstanding discard commands and GC a
+		little bit aggressively. uses cost benefit GC approach.
+		gc urgent mid(3): does GC forcibly in a period of given
+		gc_urgent_sleep_time and executes a mid level of I/O idling check.
+		uses cost benefit GC approach.
 
 What:		/sys/fs/f2fs/<disk>/gc_urgent_sleep_time
 Date:		August 2017
@@ -430,6 +463,7 @@ Description:	Show status of f2fs superblock in real time.
 		0x800  SBI_QUOTA_SKIP_FLUSH  skip flushing quota in current CP
 		0x1000 SBI_QUOTA_NEED_REPAIR quota file may be corrupted
 		0x2000 SBI_IS_RESIZEFS       resizefs is in process
+		0x4000 SBI_IS_FREEZING       freefs is in process
 		====== ===================== =================================
 
 What:		/sys/fs/f2fs/<disk>/ckpt_thread_ioprio
@@ -503,7 +537,7 @@ Date:		July 2021
 Contact:	"Daeho Jeong" <daehojeong@google.com>
 Description:	Show how many segments have been reclaimed by GC during a specific
 		GC mode (0: GC normal, 1: GC idle CB, 2: GC idle greedy,
-		3: GC idle AT, 4: GC urgent high, 5: GC urgent low)
+		3: GC idle AT, 4: GC urgent high, 5: GC urgent low 6: GC urgent mid)
 		You can re-initialize this value to "0".
 
 What:		/sys/fs/f2fs/<disk>/gc_segment_mode
@@ -540,3 +574,9 @@ Contact:	"Daeho Jeong" <daehojeong@google.com>
 Description:	You can set the trial count limit for GC urgent high mode with this value.
 		If GC thread gets to the limit, the mode will turn back to GC normal mode.
 		By default, the value is zero, which means there is no limit like before.
+
+What:		/sys/fs/f2fs/<disk>/max_roll_forward_node_blocks
+Date:		January 2022
+Contact:	"Jaegeuk Kim" <jaegeuk@kernel.org>
+Description:	Controls max # of node block writes to be used for roll forward
+		recovery. This can limit the roll forward recovery time.

fs/f2fs/Kconfig

@@ -143,3 +143,10 @@ config F2FS_IOSTAT
 	  Support getting IO statistics through sysfs and printing out periodic
 	  IO statistics tracepoint events. You have to turn on "iostat_enable"
 	  sysfs node to enable this feature.
+
+config F2FS_UNFAIR_RWSEM
+	bool "F2FS unfair rw_semaphore"
+	depends on F2FS_FS && BLK_CGROUP
+	help
+	  Use unfair rw_semaphore, if system configured IO priority by block
+	  cgroup.

fs/f2fs/acl.c

@@ -204,8 +204,9 @@ struct posix_acl *f2fs_get_acl(struct inode *inode, int type, bool rcu)
 	return __f2fs_get_acl(inode, type, NULL);
 }
 
-static int f2fs_acl_update_mode(struct inode *inode, umode_t *mode_p,
-			  struct posix_acl **acl)
+static int f2fs_acl_update_mode(struct user_namespace *mnt_userns,
+				struct inode *inode, umode_t *mode_p,
+				struct posix_acl **acl)
 {
 	umode_t mode = inode->i_mode;
 	int error;
@@ -218,14 +219,15 @@ static int f2fs_acl_update_mode(struct inode *inode, umode_t *mode_p,
 		return error;
 	if (error == 0)
 		*acl = NULL;
-	if (!in_group_p(i_gid_into_mnt(&init_user_ns, inode)) &&
-	    !capable_wrt_inode_uidgid(&init_user_ns, inode, CAP_FSETID))
+	if (!in_group_p(i_gid_into_mnt(mnt_userns, inode)) &&
+	    !capable_wrt_inode_uidgid(mnt_userns, inode, CAP_FSETID))
 		mode &= ~S_ISGID;
 	*mode_p = mode;
 	return 0;
 }
 
-static int __f2fs_set_acl(struct inode *inode, int type,
+static int __f2fs_set_acl(struct user_namespace *mnt_userns,
+			struct inode *inode, int type,
 			struct posix_acl *acl, struct page *ipage)
 {
 	int name_index;
@@ -238,7 +240,8 @@ static int __f2fs_set_acl(struct inode *inode, int type,
 	case ACL_TYPE_ACCESS:
 		name_index = F2FS_XATTR_INDEX_POSIX_ACL_ACCESS;
 		if (acl && !ipage) {
-			error = f2fs_acl_update_mode(inode, &mode, &acl);
+			error = f2fs_acl_update_mode(mnt_userns, inode,
+								&mode, &acl);
 			if (error)
 				return error;
 			set_acl_inode(inode, mode);
@@ -279,7 +282,7 @@ int f2fs_set_acl(struct user_namespace *mnt_userns, struct inode *inode,
 	if (unlikely(f2fs_cp_error(F2FS_I_SB(inode))))
 		return -EIO;
 
-	return __f2fs_set_acl(inode, type, acl, NULL);
+	return __f2fs_set_acl(mnt_userns, inode, type, acl, NULL);
 }
 
 /*
@@ -419,7 +422,7 @@ int f2fs_init_acl(struct inode *inode, struct inode *dir, struct page *ipage,
 	f2fs_mark_inode_dirty_sync(inode, true);
 
 	if (default_acl) {
-		error = __f2fs_set_acl(inode, ACL_TYPE_DEFAULT, default_acl,
+		error = __f2fs_set_acl(NULL, inode, ACL_TYPE_DEFAULT, default_acl,
 				       ipage);
 		posix_acl_release(default_acl);
 	} else {
@@ -427,7 +430,7 @@ int f2fs_init_acl(struct inode *inode, struct inode *dir, struct page *ipage,
 	}
 	if (acl) {
 		if (!error)
-			error = __f2fs_set_acl(inode, ACL_TYPE_ACCESS, acl,
+			error = __f2fs_set_acl(NULL, inode, ACL_TYPE_ACCESS, acl,
 					       ipage);
 		posix_acl_release(acl);
 	} else {

fs/f2fs/checkpoint.c

@@ -98,6 +98,13 @@ static struct page *__get_meta_page(struct f2fs_sb_info *sbi, pgoff_t index,
 	}
 
 	if (unlikely(!PageUptodate(page))) {
+		if (page->index == sbi->metapage_eio_ofs &&
+			sbi->metapage_eio_cnt++ == MAX_RETRY_META_PAGE_EIO) {
+			set_ckpt_flags(sbi, CP_ERROR_FLAG);
+		} else {
+			sbi->metapage_eio_ofs = page->index;
+			sbi->metapage_eio_cnt = 0;
+		}
 		f2fs_put_page(page, 1);
 		return ERR_PTR(-EIO);
 	}
@@ -282,18 +289,22 @@ int f2fs_ra_meta_pages(struct f2fs_sb_info *sbi, block_t start, int nrpages,
 	return blkno - start;
 }
 
-void f2fs_ra_meta_pages_cond(struct f2fs_sb_info *sbi, pgoff_t index)
+void f2fs_ra_meta_pages_cond(struct f2fs_sb_info *sbi, pgoff_t index,
+							unsigned int ra_blocks)
 {
 	struct page *page;
 	bool readahead = false;
 
+	if (ra_blocks == RECOVERY_MIN_RA_BLOCKS)
+		return;
+
 	page = find_get_page(META_MAPPING(sbi), index);
 	if (!page || !PageUptodate(page))
 		readahead = true;
 	f2fs_put_page(page, 0);
 
 	if (readahead)
-		f2fs_ra_meta_pages(sbi, index, BIO_MAX_VECS, META_POR, true);
+		f2fs_ra_meta_pages(sbi, index, ra_blocks, META_POR, true);
 }
 
 static int __f2fs_write_meta_page(struct page *page,
@@ -351,13 +362,13 @@ static int f2fs_write_meta_pages(struct address_space *mapping,
 		goto skip_write;
 
 	/* if locked failed, cp will flush dirty pages instead */
-	if (!down_write_trylock(&sbi->cp_global_sem))
+	if (!f2fs_down_write_trylock(&sbi->cp_global_sem))
 		goto skip_write;
 
 	trace_f2fs_writepages(mapping->host, wbc, META);
 	diff = nr_pages_to_write(sbi, META, wbc);
 	written = f2fs_sync_meta_pages(sbi, META, wbc->nr_to_write, FS_META_IO);
-	up_write(&sbi->cp_global_sem);
+	f2fs_up_write(&sbi->cp_global_sem);
 	wbc->nr_to_write = max((long)0, wbc->nr_to_write - written - diff);
 	return 0;
 
@@ -864,6 +875,7 @@ static struct page *validate_checkpoint(struct f2fs_sb_info *sbi,
 	struct page *cp_page_1 = NULL, *cp_page_2 = NULL;
 	struct f2fs_checkpoint *cp_block = NULL;
 	unsigned long long cur_version = 0, pre_version = 0;
+	unsigned int cp_blocks;
 	int err;
 
 	err = get_checkpoint_version(sbi, cp_addr, &cp_block,
@@ -871,15 +883,16 @@ static struct page *validate_checkpoint(struct f2fs_sb_info *sbi,
 	if (err)
 		return NULL;
 
-	if (le32_to_cpu(cp_block->cp_pack_total_block_count) >
-					sbi->blocks_per_seg) {
+	cp_blocks = le32_to_cpu(cp_block->cp_pack_total_block_count);
+
+	if (cp_blocks > sbi->blocks_per_seg || cp_blocks <= F2FS_CP_PACKS) {
 		f2fs_warn(sbi, "invalid cp_pack_total_block_count:%u",
 			  le32_to_cpu(cp_block->cp_pack_total_block_count));
 		goto invalid_cp;
 	}
 	pre_version = *version;
 
-	cp_addr += le32_to_cpu(cp_block->cp_pack_total_block_count) - 1;
+	cp_addr += cp_blocks - 1;
 	err = get_checkpoint_version(sbi, cp_addr, &cp_block,
 					&cp_page_2, version);
 	if (err)
@@ -1159,7 +1172,7 @@ static bool __need_flush_quota(struct f2fs_sb_info *sbi)
 	if (!is_journalled_quota(sbi))
 		return false;
 
-	if (!down_write_trylock(&sbi->quota_sem))
+	if (!f2fs_down_write_trylock(&sbi->quota_sem))
 		return true;
 	if (is_sbi_flag_set(sbi, SBI_QUOTA_SKIP_FLUSH)) {
 		ret = false;
@@ -1171,7 +1184,7 @@ static bool __need_flush_quota(struct f2fs_sb_info *sbi)
 	} else if (get_pages(sbi, F2FS_DIRTY_QDATA)) {
 		ret = true;
 	}
-	up_write(&sbi->quota_sem);
+	f2fs_up_write(&sbi->quota_sem);
 	return ret;
 }
 
@@ -1228,10 +1241,10 @@ static int block_operations(struct f2fs_sb_info *sbi)
 	 * 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);
+	f2fs_down_write(&sbi->node_change);
 
 	if (get_pages(sbi, F2FS_DIRTY_IMETA)) {
-		up_write(&sbi->node_change);
+		f2fs_up_write(&sbi->node_change);
 		f2fs_unlock_all(sbi);
 		err = f2fs_sync_inode_meta(sbi);
 		if (err)
@@ -1241,15 +1254,15 @@ static int block_operations(struct f2fs_sb_info *sbi)
 	}
 
 retry_flush_nodes:
-	down_write(&sbi->node_write);
+	f2fs_down_write(&sbi->node_write);
 
 	if (get_pages(sbi, F2FS_DIRTY_NODES)) {
-		up_write(&sbi->node_write);
+		f2fs_up_write(&sbi->node_write);
 		atomic_inc(&sbi->wb_sync_req[NODE]);
 		err = f2fs_sync_node_pages(sbi, &wbc, false, FS_CP_NODE_IO);
 		atomic_dec(&sbi->wb_sync_req[NODE]);
 		if (err) {
-			up_write(&sbi->node_change);
+			f2fs_up_write(&sbi->node_change);
 			f2fs_unlock_all(sbi);
 			return err;
 		}
@@ -1262,13 +1275,13 @@ static int block_operations(struct f2fs_sb_info *sbi)
 	 * dirty node blocks and some checkpoint values by block allocation.
 	 */
 	__prepare_cp_block(sbi);
-	up_write(&sbi->node_change);
+	f2fs_up_write(&sbi->node_change);
 	return err;
 }
 
 static void unblock_operations(struct f2fs_sb_info *sbi)
 {
-	up_write(&sbi->node_write);
+	f2fs_up_write(&sbi->node_write);
 	f2fs_unlock_all(sbi);
 }
 
@@ -1543,6 +1556,7 @@ static int do_checkpoint(struct f2fs_sb_info *sbi, struct cp_control *cpc)
 	/* update user_block_counts */
 	sbi->last_valid_block_count = sbi->total_valid_block_count;
 	percpu_counter_set(&sbi->alloc_valid_block_count, 0);
+	percpu_counter_set(&sbi->rf_node_block_count, 0);
 
 	/* Here, we have one bio having CP pack except cp pack 2 page */
 	f2fs_sync_meta_pages(sbi, META, LONG_MAX, FS_CP_META_IO);
@@ -1612,7 +1626,7 @@ int f2fs_write_checkpoint(struct f2fs_sb_info *sbi, struct cp_control *cpc)
 		f2fs_warn(sbi, "Start checkpoint disabled!");
 	}
 	if (cpc->reason != CP_RESIZE)
-		down_write(&sbi->cp_global_sem);
+		f2fs_down_write(&sbi->cp_global_sem);
 
 	if (!is_sbi_flag_set(sbi, SBI_IS_DIRTY) &&
 		((cpc->reason & CP_FASTBOOT) || (cpc->reason & CP_SYNC) ||
@@ -1693,7 +1707,7 @@ int f2fs_write_checkpoint(struct f2fs_sb_info *sbi, struct cp_control *cpc)
 	trace_f2fs_write_checkpoint(sbi->sb, cpc->reason, "finish checkpoint");
 out:
 	if (cpc->reason != CP_RESIZE)
-		up_write(&sbi->cp_global_sem);
+		f2fs_up_write(&sbi->cp_global_sem);
 	return err;
 }
 
@@ -1741,9 +1755,9 @@ static int __write_checkpoint_sync(struct f2fs_sb_info *sbi)
 	struct cp_control cpc = { .reason = CP_SYNC, };
 	int err;
 
-	down_write(&sbi->gc_lock);
+	f2fs_down_write(&sbi->gc_lock);
 	err = f2fs_write_checkpoint(sbi, &cpc);
-	up_write(&sbi->gc_lock);
+	f2fs_up_write(&sbi->gc_lock);
 
 	return err;
 }
@@ -1831,9 +1845,9 @@ int f2fs_issue_checkpoint(struct f2fs_sb_info *sbi)
 	if (!test_opt(sbi, MERGE_CHECKPOINT) || cpc.reason != CP_SYNC) {
 		int ret;
 
-		down_write(&sbi->gc_lock);
+		f2fs_down_write(&sbi->gc_lock);
 		ret = f2fs_write_checkpoint(sbi, &cpc);
-		up_write(&sbi->gc_lock);
+		f2fs_up_write(&sbi->gc_lock);
 
 		return ret;
 	}

fs/f2fs/compress.c

@@ -314,10 +314,9 @@ static int lz4_decompress_pages(struct decompress_io_ctx *dic)
 	}
 
 	if (ret != PAGE_SIZE << dic->log_cluster_size) {
-		printk_ratelimited("%sF2FS-fs (%s): lz4 invalid rlen:%zu, "
+		printk_ratelimited("%sF2FS-fs (%s): lz4 invalid ret:%d, "
 					"expected:%lu\n", KERN_ERR,
-					F2FS_I_SB(dic->inode)->sb->s_id,
-					dic->rlen,
+					F2FS_I_SB(dic->inode)->sb->s_id, ret,
 					PAGE_SIZE << dic->log_cluster_size);
 		return -EIO;
 	}
@@ -1267,7 +1266,7 @@ static int f2fs_write_compressed_pages(struct compress_ctx *cc,
 		 * checkpoint. This can only happen to quota writes which can cause
 		 * the below discard race condition.
 		 */
-		down_read(&sbi->node_write);
+		f2fs_down_read(&sbi->node_write);
 	} else if (!f2fs_trylock_op(sbi)) {
 		goto out_free;
 	}
@@ -1384,7 +1383,7 @@ static int f2fs_write_compressed_pages(struct compress_ctx *cc,
 
 	f2fs_put_dnode(&dn);
 	if (IS_NOQUOTA(inode))
-		up_read(&sbi->node_write);
+		f2fs_up_read(&sbi->node_write);
 	else
 		f2fs_unlock_op(sbi);
 
@@ -1410,7 +1409,7 @@ static int f2fs_write_compressed_pages(struct compress_ctx *cc,
 	f2fs_put_dnode(&dn);
 out_unlock_op:
 	if (IS_NOQUOTA(inode))
-		up_read(&sbi->node_write);
+		f2fs_up_read(&sbi->node_write);
 	else
 		f2fs_unlock_op(sbi);
 out_free:

fs/f2fs/data.c

@@ -584,7 +584,7 @@ static void __f2fs_submit_merged_write(struct f2fs_sb_info *sbi,
 	enum page_type btype = PAGE_TYPE_OF_BIO(type);
 	struct f2fs_bio_info *io = sbi->write_io[btype] + temp;
 
-	down_write(&io->io_rwsem);
+	f2fs_down_write(&io->io_rwsem);
 
 	/* change META to META_FLUSH in the checkpoint procedure */
 	if (type >= META_FLUSH) {
@@ -594,7 +594,7 @@ static void __f2fs_submit_merged_write(struct f2fs_sb_info *sbi,
 			io->bio->bi_opf |= REQ_PREFLUSH | REQ_FUA;
 	}
 	__submit_merged_bio(io);
-	up_write(&io->io_rwsem);
+	f2fs_up_write(&io->io_rwsem);
 }
 
 static void __submit_merged_write_cond(struct f2fs_sb_info *sbi,
@@ -609,9 +609,9 @@ static void __submit_merged_write_cond(struct f2fs_sb_info *sbi,
 			enum page_type btype = PAGE_TYPE_OF_BIO(type);
 			struct f2fs_bio_info *io = sbi->write_io[btype] + temp;
 
-			down_read(&io->io_rwsem);
+			f2fs_down_read(&io->io_rwsem);
 			ret = __has_merged_page(io->bio, inode, page, ino);
-			up_read(&io->io_rwsem);
+			f2fs_up_read(&io->io_rwsem);
 		}
 		if (ret)
 			__f2fs_submit_merged_write(sbi, type, temp);
@@ -732,9 +732,9 @@ static void add_bio_entry(struct f2fs_sb_info *sbi, struct bio *bio,
 	if (bio_add_page(bio, page, PAGE_SIZE, 0) != PAGE_SIZE)
 		f2fs_bug_on(sbi, 1);
 
-	down_write(&io->bio_list_lock);
+	f2fs_down_write(&io->bio_list_lock);
 	list_add_tail(&be->list, &io->bio_list);
-	up_write(&io->bio_list_lock);
+	f2fs_up_write(&io->bio_list_lock);
 }
 
 static void del_bio_entry(struct bio_entry *be)
@@ -756,7 +756,7 @@ static int add_ipu_page(struct f2fs_io_info *fio, struct bio **bio,
 		struct list_head *head = &io->bio_list;
 		struct bio_entry *be;
 
-		down_write(&io->bio_list_lock);
+		f2fs_down_write(&io->bio_list_lock);
 		list_for_each_entry(be, head, list) {
 			if (be->bio != *bio)
 				continue;
@@ -780,7 +780,7 @@ static int add_ipu_page(struct f2fs_io_info *fio, struct bio **bio,
 			__submit_bio(sbi, *bio, DATA);
 			break;
 		}
-		up_write(&io->bio_list_lock);
+		f2fs_up_write(&io->bio_list_lock);
 	}
 
 	if (ret) {
@@ -806,7 +806,7 @@ void f2fs_submit_merged_ipu_write(struct f2fs_sb_info *sbi,
 		if (list_empty(head))
 			continue;
 
-		down_read(&io->bio_list_lock);
+		f2fs_down_read(&io->bio_list_lock);
 		list_for_each_entry(be, head, list) {
 			if (target)
 				found = (target == be->bio);
@@ -816,14 +816,14 @@ void f2fs_submit_merged_ipu_write(struct f2fs_sb_info *sbi,
 			if (found)
 				break;
 		}
-		up_read(&io->bio_list_lock);
+		f2fs_up_read(&io->bio_list_lock);
 
 		if (!found)
 			continue;
 
 		found = false;
 
-		down_write(&io->bio_list_lock);
+		f2fs_down_write(&io->bio_list_lock);
 		list_for_each_entry(be, head, list) {
 			if (target)
 				found = (target == be->bio);
@@ -836,7 +836,7 @@ void f2fs_submit_merged_ipu_write(struct f2fs_sb_info *sbi,
 				break;
 			}
 		}
-		up_write(&io->bio_list_lock);
+		f2fs_up_write(&io->bio_list_lock);
 	}
 
 	if (found)
@@ -894,7 +894,7 @@ void f2fs_submit_page_write(struct f2fs_io_info *fio)
 
 	f2fs_bug_on(sbi, is_read_io(fio->op));
 
-	down_write(&io->io_rwsem);
+	f2fs_down_write(&io->io_rwsem);
 next:
 	if (fio->in_list) {
 		spin_lock(&io->io_lock);
@@ -961,7 +961,7 @@ void f2fs_submit_page_write(struct f2fs_io_info *fio)
 	if (is_sbi_flag_set(sbi, SBI_IS_SHUTDOWN) ||
 				!f2fs_is_checkpoint_ready(sbi))
 		__submit_merged_bio(io);
-	up_write(&io->io_rwsem);
+	f2fs_up_write(&io->io_rwsem);
 }
 
 static struct bio *f2fs_grab_read_bio(struct inode *inode, block_t blkaddr,
@@ -1371,9 +1371,9 @@ void f2fs_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);
+			f2fs_down_read(&sbi->node_change);
 		else
-			up_read(&sbi->node_change);
+			f2fs_up_read(&sbi->node_change);
 	} else {
 		if (lock)
 			f2fs_lock_op(sbi);
@@ -2448,6 +2448,9 @@ static inline bool check_inplace_update_policy(struct inode *inode,
 	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
 	unsigned int policy = SM_I(sbi)->ipu_policy;
 
+	if (policy & (0x1 << F2FS_IPU_HONOR_OPU_WRITE) &&
+			is_inode_flag_set(inode, FI_OPU_WRITE))
+		return false;
 	if (policy & (0x1 << F2FS_IPU_FORCE))
 		return true;
 	if (policy & (0x1 << F2FS_IPU_SSR) && f2fs_need_SSR(sbi))
@@ -2518,6 +2521,9 @@ bool f2fs_should_update_outplace(struct inode *inode, struct f2fs_io_info *fio)
 	if (is_inode_flag_set(inode, FI_ALIGNED_WRITE))
 		return true;
 
+	if (is_inode_flag_set(inode, FI_OPU_WRITE))
+		return true;
+
 	if (fio) {
 		if (page_private_gcing(fio->page))
 			return true;
@@ -2737,13 +2743,13 @@ int f2fs_write_single_data_page(struct page *page, int *submitted,
 		 * the below discard race condition.
 		 */
 		if (IS_NOQUOTA(inode))
-			down_read(&sbi->node_write);
+			f2fs_down_read(&sbi->node_write);
 
 		fio.need_lock = LOCK_DONE;
 		err = f2fs_do_write_data_page(&fio);
 
 		if (IS_NOQUOTA(inode))
-			up_read(&sbi->node_write);
+			f2fs_up_read(&sbi->node_write);
 
 		goto done;
 	}
@@ -3142,8 +3148,8 @@ static int __f2fs_write_data_pages(struct address_space *mapping,
 			f2fs_available_free_memory(sbi, DIRTY_DENTS))
 		goto skip_write;
 
-	/* skip writing during file defragment */
-	if (is_inode_flag_set(inode, FI_DO_DEFRAG))
+	/* skip writing in file defragment preparing stage */
+	if (is_inode_flag_set(inode, FI_SKIP_WRITES))
 		goto skip_write;
 
 	trace_f2fs_writepages(mapping->host, wbc, DATA);
@@ -3151,8 +3157,12 @@ static int __f2fs_write_data_pages(struct address_space *mapping,
 	/* 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[DATA]);
-	else if (atomic_read(&sbi->wb_sync_req[DATA]))
+	else if (atomic_read(&sbi->wb_sync_req[DATA])) {
+		/* to avoid potential deadlock */
+		if (current->plug)
+			blk_finish_plug(current->plug);
 		goto skip_write;
+	}
 
 	if (__should_serialize_io(inode, wbc)) {
 		mutex_lock(&sbi->writepages);
@@ -3201,14 +3211,14 @@ void f2fs_write_failed(struct inode *inode, loff_t to)
 
 	/* In the fs-verity case, f2fs_end_enable_verity() does the truncate */
 	if (to > i_size && !f2fs_verity_in_progress(inode)) {
-		down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+		f2fs_down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
 		filemap_invalidate_lock(inode->i_mapping);
 
 		truncate_pagecache(inode, i_size);
 		f2fs_truncate_blocks(inode, i_size, true);
 
 		filemap_invalidate_unlock(inode->i_mapping);
-		up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+		f2fs_up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
 	}
 }
 
@@ -3341,7 +3351,7 @@ static int f2fs_write_begin(struct file *file, struct address_space *mapping,
 
 		*fsdata = NULL;
 
-		if (len == PAGE_SIZE)
+		if (len == PAGE_SIZE && !(f2fs_is_atomic_file(inode)))
 			goto repeat;
 
 		ret = f2fs_prepare_compress_overwrite(inode, pagep,
@@ -3709,19 +3719,20 @@ static int f2fs_migrate_blocks(struct inode *inode, block_t start_blk,
 	unsigned int end_sec = secidx + blkcnt / blk_per_sec;
 	int ret = 0;
 
-	down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+	f2fs_down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
 	filemap_invalidate_lock(inode->i_mapping);
 
 	set_inode_flag(inode, FI_ALIGNED_WRITE);
+	set_inode_flag(inode, FI_OPU_WRITE);
 
 	for (; secidx < end_sec; secidx++) {
-		down_write(&sbi->pin_sem);
+		f2fs_down_write(&sbi->pin_sem);
 
 		f2fs_lock_op(sbi);
 		f2fs_allocate_new_section(sbi, CURSEG_COLD_DATA_PINNED, false);
 		f2fs_unlock_op(sbi);
 
-		set_inode_flag(inode, FI_DO_DEFRAG);
+		set_inode_flag(inode, FI_SKIP_WRITES);
 
 		for (blkofs = 0; blkofs < blk_per_sec; blkofs++) {
 			struct page *page;
@@ -3729,7 +3740,7 @@ static int f2fs_migrate_blocks(struct inode *inode, block_t start_blk,
 
 			page = f2fs_get_lock_data_page(inode, blkidx, true);
 			if (IS_ERR(page)) {
-				up_write(&sbi->pin_sem);
+				f2fs_up_write(&sbi->pin_sem);
 				ret = PTR_ERR(page);
 				goto done;
 			}
@@ -3738,22 +3749,23 @@ static int f2fs_migrate_blocks(struct inode *inode, block_t start_blk,
 			f2fs_put_page(page, 1);
 		}
 
-		clear_inode_flag(inode, FI_DO_DEFRAG);
+		clear_inode_flag(inode, FI_SKIP_WRITES);
 
 		ret = filemap_fdatawrite(inode->i_mapping);
 
-		up_write(&sbi->pin_sem);
+		f2fs_up_write(&sbi->pin_sem);
 
 		if (ret)
 			break;
 	}
 
 done:
-	clear_inode_flag(inode, FI_DO_DEFRAG);
+	clear_inode_flag(inode, FI_SKIP_WRITES);
+	clear_inode_flag(inode, FI_OPU_WRITE);
 	clear_inode_flag(inode, FI_ALIGNED_WRITE);
 
 	filemap_invalidate_unlock(inode->i_mapping);
-	up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+	f2fs_up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
 
 	return ret;
 }

fs/f2fs/debug.c

@@ -21,7 +21,7 @@
 #include "gc.h"
 
 static LIST_HEAD(f2fs_stat_list);
-static DEFINE_MUTEX(f2fs_stat_mutex);
+static DEFINE_RAW_SPINLOCK(f2fs_stat_lock);
 #ifdef CONFIG_DEBUG_FS
 static struct dentry *f2fs_debugfs_root;
 #endif
@@ -338,14 +338,16 @@ static char *s_flag[] = {
 	[SBI_QUOTA_SKIP_FLUSH]	= " quota_skip_flush",
 	[SBI_QUOTA_NEED_REPAIR]	= " quota_need_repair",
 	[SBI_IS_RESIZEFS]	= " resizefs",
+	[SBI_IS_FREEZING]	= " freezefs",
 };
 
 static int stat_show(struct seq_file *s, void *v)
 {
 	struct f2fs_stat_info *si;
 	int i = 0, j = 0;
+	unsigned long flags;
 
-	mutex_lock(&f2fs_stat_mutex);
+	raw_spin_lock_irqsave(&f2fs_stat_lock, flags);
 	list_for_each_entry(si, &f2fs_stat_list, stat_list) {
 		update_general_status(si->sbi);
 
@@ -474,12 +476,14 @@ static int stat_show(struct seq_file *s, void *v)
 				si->node_segs, si->bg_node_segs);
 		seq_printf(s, "  - Reclaimed segs : Normal (%d), Idle CB (%d), "
 				"Idle Greedy (%d), Idle AT (%d), "
-				"Urgent High (%d), Urgent Low (%d)\n",
+				"Urgent High (%d), Urgent Mid (%d), "
+				"Urgent Low (%d)\n",
 				si->sbi->gc_reclaimed_segs[GC_NORMAL],
 				si->sbi->gc_reclaimed_segs[GC_IDLE_CB],
 				si->sbi->gc_reclaimed_segs[GC_IDLE_GREEDY],
 				si->sbi->gc_reclaimed_segs[GC_IDLE_AT],
 				si->sbi->gc_reclaimed_segs[GC_URGENT_HIGH],
+				si->sbi->gc_reclaimed_segs[GC_URGENT_MID],
 				si->sbi->gc_reclaimed_segs[GC_URGENT_LOW]);
 		seq_printf(s, "Try to move %d blocks (BG: %d)\n", si->tot_blks,
 				si->bg_data_blks + si->bg_node_blks);
@@ -532,6 +536,9 @@ static int stat_show(struct seq_file *s, void *v)
 			   si->ndirty_meta, si->meta_pages);
 		seq_printf(s, "  - imeta: %4d\n",
 			   si->ndirty_imeta);
+		seq_printf(s, "  - fsync mark: %4lld\n",
+			   percpu_counter_sum_positive(
+					&si->sbi->rf_node_block_count));
 		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/%9d\n  - alloc_nids: %9d\n",
@@ -573,7 +580,7 @@ static int stat_show(struct seq_file *s, void *v)
 		seq_printf(s, "  - paged : %llu KB\n",
 				si->page_mem >> 10);
 	}
-	mutex_unlock(&f2fs_stat_mutex);
+	raw_spin_unlock_irqrestore(&f2fs_stat_lock, flags);
 	return 0;
 }
 
@@ -584,6 +591,7 @@ int f2fs_build_stats(struct f2fs_sb_info *sbi)
 {
 	struct f2fs_super_block *raw_super = F2FS_RAW_SUPER(sbi);
 	struct f2fs_stat_info *si;
+	unsigned long flags;
 	int i;
 
 	si = f2fs_kzalloc(sbi, sizeof(struct f2fs_stat_info), GFP_KERNEL);
@@ -619,9 +627,9 @@ int f2fs_build_stats(struct f2fs_sb_info *sbi)
 	atomic_set(&sbi->max_aw_cnt, 0);
 	atomic_set(&sbi->max_vw_cnt, 0);
 
-	mutex_lock(&f2fs_stat_mutex);
+	raw_spin_lock_irqsave(&f2fs_stat_lock, flags);
 	list_add_tail(&si->stat_list, &f2fs_stat_list);
-	mutex_unlock(&f2fs_stat_mutex);
+	raw_spin_unlock_irqrestore(&f2fs_stat_lock, flags);
 
 	return 0;
 }
@@ -629,10 +637,11 @@ int f2fs_build_stats(struct f2fs_sb_info *sbi)
 void f2fs_destroy_stats(struct f2fs_sb_info *sbi)
 {
 	struct f2fs_stat_info *si = F2FS_STAT(sbi);
+	unsigned long flags;
 
-	mutex_lock(&f2fs_stat_mutex);
+	raw_spin_lock_irqsave(&f2fs_stat_lock, flags);
 	list_del(&si->stat_list);
-	mutex_unlock(&f2fs_stat_mutex);
+	raw_spin_unlock_irqrestore(&f2fs_stat_lock, flags);
 
 	kfree(si);
 }

fs/f2fs/dir.c

@@ -766,7 +766,7 @@ int f2fs_add_regular_entry(struct inode *dir, const struct f2fs_filename *fname,
 	f2fs_wait_on_page_writeback(dentry_page, DATA, true, true);
 
 	if (inode) {
-		down_write(&F2FS_I(inode)->i_sem);
+		f2fs_down_write(&F2FS_I(inode)->i_sem);
 		page = f2fs_init_inode_metadata(inode, dir, fname, NULL);
 		if (IS_ERR(page)) {
 			err = PTR_ERR(page);
@@ -793,7 +793,7 @@ int f2fs_add_regular_entry(struct inode *dir, const struct f2fs_filename *fname,
 	f2fs_update_parent_metadata(dir, inode, current_depth);
 fail:
 	if (inode)
-		up_write(&F2FS_I(inode)->i_sem);
+		f2fs_up_write(&F2FS_I(inode)->i_sem);
 
 	f2fs_put_page(dentry_page, 1);
 
@@ -858,7 +858,7 @@ int f2fs_do_tmpfile(struct inode *inode, struct inode *dir)
 	struct page *page;
 	int err = 0;
 
-	down_write(&F2FS_I(inode)->i_sem);
+	f2fs_down_write(&F2FS_I(inode)->i_sem);
 	page = f2fs_init_inode_metadata(inode, dir, NULL, NULL);
 	if (IS_ERR(page)) {
 		err = PTR_ERR(page);
@@ -869,7 +869,7 @@ int f2fs_do_tmpfile(struct inode *inode, struct inode *dir)
 	clear_inode_flag(inode, FI_NEW_INODE);
 	f2fs_update_time(F2FS_I_SB(inode), REQ_TIME);
 fail:
-	up_write(&F2FS_I(inode)->i_sem);
+	f2fs_up_write(&F2FS_I(inode)->i_sem);
 	return err;
 }
 
@@ -877,7 +877,7 @@ void f2fs_drop_nlink(struct inode *dir, struct inode *inode)
 {
 	struct f2fs_sb_info *sbi = F2FS_I_SB(dir);
 
-	down_write(&F2FS_I(inode)->i_sem);
+	f2fs_down_write(&F2FS_I(inode)->i_sem);
 
 	if (S_ISDIR(inode->i_mode))
 		f2fs_i_links_write(dir, false);
@@ -888,7 +888,7 @@ void f2fs_drop_nlink(struct inode *dir, struct inode *inode)
 		f2fs_i_links_write(inode, false);
 		f2fs_i_size_write(inode, 0);
 	}
-	up_write(&F2FS_I(inode)->i_sem);
+	f2fs_up_write(&F2FS_I(inode)->i_sem);
 
 	if (inode->i_nlink == 0)
 		f2fs_add_orphan_inode(inode);

fs/f2fs/gc.c

@@ -103,23 +103,26 @@ static int gc_thread_func(void *data)
 				sbi->gc_urgent_high_remaining--;
 			}
 			spin_unlock(&sbi->gc_urgent_high_lock);
+		}
 
+		if (sbi->gc_mode == GC_URGENT_HIGH ||
+				sbi->gc_mode == GC_URGENT_MID) {
 			wait_ms = gc_th->urgent_sleep_time;
-			down_write(&sbi->gc_lock);
+			f2fs_down_write(&sbi->gc_lock);
 			goto do_gc;
 		}
 
 		if (foreground) {
-			down_write(&sbi->gc_lock);
+			f2fs_down_write(&sbi->gc_lock);
 			goto do_gc;
-		} else if (!down_write_trylock(&sbi->gc_lock)) {
+		} else if (!f2fs_down_write_trylock(&sbi->gc_lock)) {
 			stat_other_skip_bggc_count(sbi);
 			goto next;
 		}
 
 		if (!is_idle(sbi, GC_TIME)) {
 			increase_sleep_time(gc_th, &wait_ms);
-			up_write(&sbi->gc_lock);
+			f2fs_up_write(&sbi->gc_lock);
 			stat_io_skip_bggc_count(sbi);
 			goto next;
 		}
@@ -1038,8 +1041,10 @@ static bool is_alive(struct f2fs_sb_info *sbi, struct f2fs_summary *sum,
 		set_sbi_flag(sbi, SBI_NEED_FSCK);
 	}
 
-	if (f2fs_check_nid_range(sbi, dni->ino))
+	if (f2fs_check_nid_range(sbi, dni->ino)) {
+		f2fs_put_page(node_page, 1);
 		return false;
+	}
 
 	*nofs = ofs_of_node(node_page);
 	source_blkaddr = data_blkaddr(NULL, node_page, ofs_in_node);
@@ -1230,7 +1235,7 @@ static int move_data_block(struct inode *inode, block_t bidx,
 	fio.new_blkaddr = fio.old_blkaddr = dn.data_blkaddr;
 
 	if (lfs_mode)
-		down_write(&fio.sbi->io_order_lock);
+		f2fs_down_write(&fio.sbi->io_order_lock);
 
 	mpage = f2fs_grab_cache_page(META_MAPPING(fio.sbi),
 					fio.old_blkaddr, false);
@@ -1316,7 +1321,7 @@ static int move_data_block(struct inode *inode, block_t bidx,
 							true, true, true);
 up_out:
 	if (lfs_mode)
-		up_write(&fio.sbi->io_order_lock);
+		f2fs_up_write(&fio.sbi->io_order_lock);
 put_out:
 	f2fs_put_dnode(&dn);
 out:
@@ -1475,7 +1480,7 @@ static int gc_data_segment(struct f2fs_sb_info *sbi, struct f2fs_summary *sum,
 					special_file(inode->i_mode))
 				continue;
 
-			if (!down_write_trylock(
+			if (!f2fs_down_write_trylock(
 				&F2FS_I(inode)->i_gc_rwsem[WRITE])) {
 				iput(inode);
 				sbi->skipped_gc_rwsem++;
@@ -1488,7 +1493,7 @@ static int gc_data_segment(struct f2fs_sb_info *sbi, struct f2fs_summary *sum,
 			if (f2fs_post_read_required(inode)) {
 				int err = ra_data_block(inode, start_bidx);
 
-				up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+				f2fs_up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
 				if (err) {
 					iput(inode);
 					continue;
@@ -1499,7 +1504,7 @@ static int gc_data_segment(struct f2fs_sb_info *sbi, struct f2fs_summary *sum,
 
 			data_page = f2fs_get_read_data_page(inode,
 						start_bidx, REQ_RAHEAD, true);
-			up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+			f2fs_up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
 			if (IS_ERR(data_page)) {
 				iput(inode);
 				continue;
@@ -1518,14 +1523,14 @@ static int gc_data_segment(struct f2fs_sb_info *sbi, struct f2fs_summary *sum,
 			int err;
 
 			if (S_ISREG(inode->i_mode)) {
-				if (!down_write_trylock(&fi->i_gc_rwsem[READ])) {
+				if (!f2fs_down_write_trylock(&fi->i_gc_rwsem[READ])) {
 					sbi->skipped_gc_rwsem++;
 					continue;
 				}
-				if (!down_write_trylock(
+				if (!f2fs_down_write_trylock(
 						&fi->i_gc_rwsem[WRITE])) {
 					sbi->skipped_gc_rwsem++;
-					up_write(&fi->i_gc_rwsem[READ]);
+					f2fs_up_write(&fi->i_gc_rwsem[READ]);
 					continue;
 				}
 				locked = true;
@@ -1548,8 +1553,8 @@ static int gc_data_segment(struct f2fs_sb_info *sbi, struct f2fs_summary *sum,
 				submitted++;
 
 			if (locked) {
-				up_write(&fi->i_gc_rwsem[WRITE]);
-				up_write(&fi->i_gc_rwsem[READ]);
+				f2fs_up_write(&fi->i_gc_rwsem[WRITE]);
+				f2fs_up_write(&fi->i_gc_rwsem[READ]);
 			}
 
 			stat_inc_data_blk_count(sbi, 1, gc_type);
@@ -1807,7 +1812,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync,
 				reserved_segments(sbi),
 				prefree_segments(sbi));
 
-	up_write(&sbi->gc_lock);
+	f2fs_up_write(&sbi->gc_lock);
 
 	put_gc_inode(&gc_list);
 
@@ -1936,7 +1941,7 @@ static void update_sb_metadata(struct f2fs_sb_info *sbi, int secs)
 	long long block_count;
 	int segs = secs * sbi->segs_per_sec;
 
-	down_write(&sbi->sb_lock);
+	f2fs_down_write(&sbi->sb_lock);
 
 	section_count = le32_to_cpu(raw_sb->section_count);
 	segment_count = le32_to_cpu(raw_sb->segment_count);
@@ -1957,7 +1962,7 @@ static void update_sb_metadata(struct f2fs_sb_info *sbi, int secs)
 						cpu_to_le32(dev_segs + segs);
 	}
 
-	up_write(&sbi->sb_lock);
+	f2fs_up_write(&sbi->sb_lock);
 }
 
 static void update_fs_metadata(struct f2fs_sb_info *sbi, int secs)
@@ -2031,7 +2036,7 @@ int f2fs_resize_fs(struct f2fs_sb_info *sbi, __u64 block_count)
 	secs = div_u64(shrunk_blocks, BLKS_PER_SEC(sbi));
 
 	/* stop other GC */
-	if (!down_write_trylock(&sbi->gc_lock))
+	if (!f2fs_down_write_trylock(&sbi->gc_lock))
 		return -EAGAIN;
 
 	/* stop CP to protect MAIN_SEC in free_segment_range */
@@ -2051,15 +2056,15 @@ int f2fs_resize_fs(struct f2fs_sb_info *sbi, __u64 block_count)
 
 out_unlock:
 	f2fs_unlock_op(sbi);
-	up_write(&sbi->gc_lock);
+	f2fs_up_write(&sbi->gc_lock);
 	if (err)
 		return err;
 
 	set_sbi_flag(sbi, SBI_IS_RESIZEFS);
 
 	freeze_super(sbi->sb);
-	down_write(&sbi->gc_lock);
-	down_write(&sbi->cp_global_sem);
+	f2fs_down_write(&sbi->gc_lock);
+	f2fs_down_write(&sbi->cp_global_sem);
 
 	spin_lock(&sbi->stat_lock);
 	if (shrunk_blocks + valid_user_blocks(sbi) +
@@ -2104,8 +2109,8 @@ int f2fs_resize_fs(struct f2fs_sb_info *sbi, __u64 block_count)
 		spin_unlock(&sbi->stat_lock);
 	}
 out_err:
-	up_write(&sbi->cp_global_sem);
-	up_write(&sbi->gc_lock);
+	f2fs_up_write(&sbi->cp_global_sem);
+	f2fs_up_write(&sbi->gc_lock);
 	thaw_super(sbi->sb);
 	clear_sbi_flag(sbi, SBI_IS_RESIZEFS);
 	return err;

fs/f2fs/inline.c

@@ -629,7 +629,7 @@ int f2fs_add_inline_entry(struct inode *dir, const struct f2fs_filename *fname,
 	}
 
 	if (inode) {
-		down_write(&F2FS_I(inode)->i_sem);
+		f2fs_down_write(&F2FS_I(inode)->i_sem);
 		page = f2fs_init_inode_metadata(inode, dir, fname, ipage);
 		if (IS_ERR(page)) {
 			err = PTR_ERR(page);
@@ -658,7 +658,7 @@ int f2fs_add_inline_entry(struct inode *dir, const struct f2fs_filename *fname,
 	f2fs_update_parent_metadata(dir, inode, 0);
 fail:
 	if (inode)
-		up_write(&F2FS_I(inode)->i_sem);
+		f2fs_up_write(&F2FS_I(inode)->i_sem);
 out:
 	f2fs_put_page(ipage, 1);
 	return err;

fs/f2fs/inode.c

@@ -778,7 +778,8 @@ void f2fs_evict_inode(struct inode *inode)
 	f2fs_remove_ino_entry(sbi, inode->i_ino, UPDATE_INO);
 	f2fs_remove_ino_entry(sbi, inode->i_ino, FLUSH_INO);
 
-	sb_start_intwrite(inode->i_sb);
+	if (!is_sbi_flag_set(sbi, SBI_IS_FREEZING))
+		sb_start_intwrite(inode->i_sb);
 	set_inode_flag(inode, FI_NO_ALLOC);
 	i_size_write(inode, 0);
 retry:
@@ -809,7 +810,8 @@ void f2fs_evict_inode(struct inode *inode)
 		if (dquot_initialize_needed(inode))
 			set_sbi_flag(sbi, SBI_QUOTA_NEED_REPAIR);
 	}
-	sb_end_intwrite(inode->i_sb);
+	if (!is_sbi_flag_set(sbi, SBI_IS_FREEZING))
+		sb_end_intwrite(inode->i_sb);
 no_delete:
 	dquot_drop(inode);
 
@@ -885,6 +887,7 @@ void f2fs_handle_failed_inode(struct inode *inode)
 	err = f2fs_get_node_info(sbi, inode->i_ino, &ni, false);
 	if (err) {
 		set_sbi_flag(sbi, SBI_NEED_FSCK);
+		set_inode_flag(inode, FI_FREE_NID);
 		f2fs_warn(sbi, "May loss orphan inode, run fsck to fix.");
 		goto out;
 	}

fs/f2fs/namei.c

@@ -22,7 +22,8 @@
 #include "acl.h"
 #include <trace/events/f2fs.h>
 
-static struct inode *f2fs_new_inode(struct inode *dir, umode_t mode)
+static struct inode *f2fs_new_inode(struct user_namespace *mnt_userns,
+						struct inode *dir, umode_t mode)
 {
 	struct f2fs_sb_info *sbi = F2FS_I_SB(dir);
 	nid_t ino;
@@ -46,7 +47,7 @@ static struct inode *f2fs_new_inode(struct inode *dir, umode_t mode)
 
 	nid_free = true;
 
-	inode_init_owner(&init_user_ns, inode, dir, mode);
+	inode_init_owner(mnt_userns, inode, dir, mode);
 
 	inode->i_ino = ino;
 	inode->i_blocks = 0;
@@ -67,7 +68,7 @@ static struct inode *f2fs_new_inode(struct inode *dir, umode_t mode)
 		(F2FS_I(dir)->i_flags & F2FS_PROJINHERIT_FL))
 		F2FS_I(inode)->i_projid = F2FS_I(dir)->i_projid;
 	else
-		F2FS_I(inode)->i_projid = make_kprojid(&init_user_ns,
+		F2FS_I(inode)->i_projid = make_kprojid(mnt_userns,
 							F2FS_DEF_PROJID);
 
 	err = fscrypt_prepare_new_inode(dir, inode, &encrypt);
@@ -196,7 +197,7 @@ static inline void set_file_temperature(struct f2fs_sb_info *sbi, struct inode *
 	__u8 (*extlist)[F2FS_EXTENSION_LEN] = sbi->raw_super->extension_list;
 	int i, cold_count, hot_count;
 
-	down_read(&sbi->sb_lock);
+	f2fs_down_read(&sbi->sb_lock);
 
 	cold_count = le32_to_cpu(sbi->raw_super->extension_count);
 	hot_count = sbi->raw_super->hot_ext_count;
@@ -206,7 +207,7 @@ static inline void set_file_temperature(struct f2fs_sb_info *sbi, struct inode *
 			break;
 	}
 
-	up_read(&sbi->sb_lock);
+	f2fs_up_read(&sbi->sb_lock);
 
 	if (i == cold_count + hot_count)
 		return;
@@ -299,19 +300,19 @@ static void set_compress_inode(struct f2fs_sb_info *sbi, struct inode *inode,
 			(!ext_cnt && !noext_cnt))
 		return;
 
-	down_read(&sbi->sb_lock);
+	f2fs_down_read(&sbi->sb_lock);
 
 	cold_count = le32_to_cpu(sbi->raw_super->extension_count);
 	hot_count = sbi->raw_super->hot_ext_count;
 
 	for (i = cold_count; i < cold_count + hot_count; i++) {
 		if (is_extension_exist(name, extlist[i], false)) {
-			up_read(&sbi->sb_lock);
+			f2fs_up_read(&sbi->sb_lock);
 			return;
 		}
 	}
 
-	up_read(&sbi->sb_lock);
+	f2fs_up_read(&sbi->sb_lock);
 
 	for (i = 0; i < noext_cnt; i++) {
 		if (is_extension_exist(name, noext[i], false)) {
@@ -349,7 +350,7 @@ static int f2fs_create(struct user_namespace *mnt_userns, struct inode *dir,
 	if (err)
 		return err;
 
-	inode = f2fs_new_inode(dir, mode);
+	inode = f2fs_new_inode(mnt_userns, dir, mode);
 	if (IS_ERR(inode))
 		return PTR_ERR(inode);
 
@@ -679,7 +680,7 @@ static int f2fs_symlink(struct user_namespace *mnt_userns, struct inode *dir,
 	if (err)
 		return err;
 
-	inode = f2fs_new_inode(dir, S_IFLNK | S_IRWXUGO);
+	inode = f2fs_new_inode(mnt_userns, dir, S_IFLNK | S_IRWXUGO);
 	if (IS_ERR(inode))
 		return PTR_ERR(inode);
 
@@ -750,7 +751,7 @@ static int f2fs_mkdir(struct user_namespace *mnt_userns, struct inode *dir,
 	if (err)
 		return err;
 
-	inode = f2fs_new_inode(dir, S_IFDIR | mode);
+	inode = f2fs_new_inode(mnt_userns, dir, S_IFDIR | mode);
 	if (IS_ERR(inode))
 		return PTR_ERR(inode);
 
@@ -807,7 +808,7 @@ static int f2fs_mknod(struct user_namespace *mnt_userns, struct inode *dir,
 	if (err)
 		return err;
 
-	inode = f2fs_new_inode(dir, mode);
+	inode = f2fs_new_inode(mnt_userns, dir, mode);
 	if (IS_ERR(inode))
 		return PTR_ERR(inode);
 
@@ -834,8 +835,9 @@ static int f2fs_mknod(struct user_namespace *mnt_userns, struct inode *dir,
 	return err;
 }
 
-static int __f2fs_tmpfile(struct inode *dir, struct dentry *dentry,
-					umode_t mode, struct inode **whiteout)
+static int __f2fs_tmpfile(struct user_namespace *mnt_userns, struct inode *dir,
+					struct dentry *dentry, umode_t mode,
+					struct inode **whiteout)
 {
 	struct f2fs_sb_info *sbi = F2FS_I_SB(dir);
 	struct inode *inode;
@@ -845,7 +847,7 @@ static int __f2fs_tmpfile(struct inode *dir, struct dentry *dentry,
 	if (err)
 		return err;
 
-	inode = f2fs_new_inode(dir, mode);
+	inode = f2fs_new_inode(mnt_userns, dir, mode);
 	if (IS_ERR(inode))
 		return PTR_ERR(inode);
 
@@ -909,20 +911,22 @@ static int f2fs_tmpfile(struct user_namespace *mnt_userns, struct inode *dir,
 	if (!f2fs_is_checkpoint_ready(sbi))
 		return -ENOSPC;
 
-	return __f2fs_tmpfile(dir, dentry, mode, NULL);
+	return __f2fs_tmpfile(mnt_userns, dir, dentry, mode, NULL);
 }
 
-static int f2fs_create_whiteout(struct inode *dir, struct inode **whiteout)
+static int f2fs_create_whiteout(struct user_namespace *mnt_userns,
+				struct inode *dir, struct inode **whiteout)
 {
 	if (unlikely(f2fs_cp_error(F2FS_I_SB(dir))))
 		return -EIO;
 
-	return __f2fs_tmpfile(dir, NULL, S_IFCHR | WHITEOUT_MODE, whiteout);
+	return __f2fs_tmpfile(mnt_userns, dir, NULL,
+				S_IFCHR | WHITEOUT_MODE, whiteout);
 }
 
-static int f2fs_rename(struct inode *old_dir, struct dentry *old_dentry,
-			struct inode *new_dir, struct dentry *new_dentry,
-			unsigned int flags)
+static int f2fs_rename(struct user_namespace *mnt_userns, struct inode *old_dir,
+			struct dentry *old_dentry, struct inode *new_dir,
+			struct dentry *new_dentry, unsigned int flags)
 {
 	struct f2fs_sb_info *sbi = F2FS_I_SB(old_dir);
 	struct inode *old_inode = d_inode(old_dentry);
@@ -960,7 +964,7 @@ static int f2fs_rename(struct inode *old_dir, struct dentry *old_dentry,
 	}
 
 	if (flags & RENAME_WHITEOUT) {
-		err = f2fs_create_whiteout(old_dir, &whiteout);
+		err = f2fs_create_whiteout(mnt_userns, old_dir, &whiteout);
 		if (err)
 			return err;
 	}
@@ -1023,11 +1027,11 @@ static int f2fs_rename(struct inode *old_dir, struct dentry *old_dentry,
 		new_page = NULL;
 
 		new_inode->i_ctime = current_time(new_inode);
-		down_write(&F2FS_I(new_inode)->i_sem);
+		f2fs_down_write(&F2FS_I(new_inode)->i_sem);
 		if (old_dir_entry)
 			f2fs_i_links_write(new_inode, false);
 		f2fs_i_links_write(new_inode, false);
-		up_write(&F2FS_I(new_inode)->i_sem);
+		f2fs_up_write(&F2FS_I(new_inode)->i_sem);
 
 		if (!new_inode->i_nlink)
 			f2fs_add_orphan_inode(new_inode);
@@ -1048,13 +1052,13 @@ static int f2fs_rename(struct inode *old_dir, struct dentry *old_dentry,
 			f2fs_i_links_write(new_dir, true);
 	}
 
-	down_write(&F2FS_I(old_inode)->i_sem);
+	f2fs_down_write(&F2FS_I(old_inode)->i_sem);
 	if (!old_dir_entry || whiteout)
 		file_lost_pino(old_inode);
 	else
 		/* adjust dir's i_pino to pass fsck check */
 		f2fs_i_pino_write(old_inode, new_dir->i_ino);
-	up_write(&F2FS_I(old_inode)->i_sem);
+	f2fs_up_write(&F2FS_I(old_inode)->i_sem);
 
 	old_inode->i_ctime = current_time(old_inode);
 	f2fs_mark_inode_dirty_sync(old_inode, false);
@@ -1107,8 +1111,7 @@ static int f2fs_rename(struct inode *old_dir, struct dentry *old_dentry,
 out_old:
 	f2fs_put_page(old_page, 0);
 out:
-	if (whiteout)
-		iput(whiteout);
+	iput(whiteout);
 	return err;
 }
 
@@ -1214,38 +1217,38 @@ static int f2fs_cross_rename(struct inode *old_dir, struct dentry *old_dentry,
 	/* update directory entry info of old dir inode */
 	f2fs_set_link(old_dir, old_entry, old_page, new_inode);
 
-	down_write(&F2FS_I(old_inode)->i_sem);
+	f2fs_down_write(&F2FS_I(old_inode)->i_sem);
 	if (!old_dir_entry)
 		file_lost_pino(old_inode);
 	else
 		/* adjust dir's i_pino to pass fsck check */
 		f2fs_i_pino_write(old_inode, new_dir->i_ino);
-	up_write(&F2FS_I(old_inode)->i_sem);
+	f2fs_up_write(&F2FS_I(old_inode)->i_sem);
 
 	old_dir->i_ctime = current_time(old_dir);
 	if (old_nlink) {
-		down_write(&F2FS_I(old_dir)->i_sem);
+		f2fs_down_write(&F2FS_I(old_dir)->i_sem);
 		f2fs_i_links_write(old_dir, old_nlink > 0);
-		up_write(&F2FS_I(old_dir)->i_sem);
+		f2fs_up_write(&F2FS_I(old_dir)->i_sem);
 	}
 	f2fs_mark_inode_dirty_sync(old_dir, false);
 
 	/* update directory entry info of new dir inode */
 	f2fs_set_link(new_dir, new_entry, new_page, old_inode);
 
-	down_write(&F2FS_I(new_inode)->i_sem);
+	f2fs_down_write(&F2FS_I(new_inode)->i_sem);
 	if (!new_dir_entry)
 		file_lost_pino(new_inode);
 	else
 		/* adjust dir's i_pino to pass fsck check */
 		f2fs_i_pino_write(new_inode, old_dir->i_ino);
-	up_write(&F2FS_I(new_inode)->i_sem);
+	f2fs_up_write(&F2FS_I(new_inode)->i_sem);
 
 	new_dir->i_ctime = current_time(new_dir);
 	if (new_nlink) {
-		down_write(&F2FS_I(new_dir)->i_sem);
+		f2fs_down_write(&F2FS_I(new_dir)->i_sem);
 		f2fs_i_links_write(new_dir, new_nlink > 0);
-		up_write(&F2FS_I(new_dir)->i_sem);
+		f2fs_up_write(&F2FS_I(new_dir)->i_sem);
 	}
 	f2fs_mark_inode_dirty_sync(new_dir, false);
 
@@ -1300,7 +1303,8 @@ static int f2fs_rename2(struct user_namespace *mnt_userns,
 	 * VFS has already handled the new dentry existence case,
 	 * here, we just deal with "RENAME_NOREPLACE" as regular rename.
 	 */
-	return f2fs_rename(old_dir, old_dentry, new_dir, new_dentry, flags);
+	return f2fs_rename(mnt_userns, old_dir, old_dentry,
+					new_dir, new_dentry, flags);
 }
 
 static const char *f2fs_encrypted_get_link(struct dentry *dentry,

fs/f2fs/node.h

@@ -31,6 +31,9 @@
 /* control total # of nats */
 #define DEF_NAT_CACHE_THRESHOLD			100000
 
+/* control total # of node writes used for roll-fowrad recovery */
+#define DEF_RF_NODE_BLOCKS			0
+
 /* vector size for gang look-up from nat cache that consists of radix tree */
 #define NATVEC_SIZE	64
 #define SETVEC_SIZE	32

fs/f2fs/recovery.c

@@ -56,6 +56,10 @@ bool f2fs_space_for_roll_forward(struct f2fs_sb_info *sbi)
 
 	if (sbi->last_valid_block_count + nalloc > sbi->user_block_count)
 		return false;
+	if (NM_I(sbi)->max_rf_node_blocks &&
+		percpu_counter_sum_positive(&sbi->rf_node_block_count) >=
+						NM_I(sbi)->max_rf_node_blocks)
+		return false;
 	return true;
 }
 
@@ -343,6 +347,19 @@ static int recover_inode(struct inode *inode, struct page *page)
 	return 0;
 }
 
+static unsigned int adjust_por_ra_blocks(struct f2fs_sb_info *sbi,
+				unsigned int ra_blocks, unsigned int blkaddr,
+				unsigned int next_blkaddr)
+{
+	if (blkaddr + 1 == next_blkaddr)
+		ra_blocks = min_t(unsigned int, RECOVERY_MAX_RA_BLOCKS,
+							ra_blocks * 2);
+	else if (next_blkaddr % sbi->blocks_per_seg)
+		ra_blocks = max_t(unsigned int, RECOVERY_MIN_RA_BLOCKS,
+							ra_blocks / 2);
+	return ra_blocks;
+}
+
 static int find_fsync_dnodes(struct f2fs_sb_info *sbi, struct list_head *head,
 				bool check_only)
 {
@@ -350,6 +367,7 @@ static int find_fsync_dnodes(struct f2fs_sb_info *sbi, struct list_head *head,
 	struct page *page = NULL;
 	block_t blkaddr;
 	unsigned int loop_cnt = 0;
+	unsigned int ra_blocks = RECOVERY_MAX_RA_BLOCKS;
 	unsigned int free_blocks = MAIN_SEGS(sbi) * sbi->blocks_per_seg -
 						valid_user_blocks(sbi);
 	int err = 0;
@@ -424,11 +442,14 @@ static int find_fsync_dnodes(struct f2fs_sb_info *sbi, struct list_head *head,
 			break;
 		}
 
+		ra_blocks = adjust_por_ra_blocks(sbi, ra_blocks, blkaddr,
+						next_blkaddr_of_node(page));
+
 		/* check next segment */
 		blkaddr = next_blkaddr_of_node(page);
 		f2fs_put_page(page, 1);
 
-		f2fs_ra_meta_pages_cond(sbi, blkaddr);
+		f2fs_ra_meta_pages_cond(sbi, blkaddr, ra_blocks);
 	}
 	return err;
 }
@@ -704,6 +725,7 @@ static int recover_data(struct f2fs_sb_info *sbi, struct list_head *inode_list,
 	struct page *page = NULL;
 	int err = 0;
 	block_t blkaddr;
+	unsigned int ra_blocks = RECOVERY_MAX_RA_BLOCKS;
 
 	/* get node pages in the current segment */
 	curseg = CURSEG_I(sbi, CURSEG_WARM_NODE);
@@ -715,8 +737,6 @@ static int recover_data(struct f2fs_sb_info *sbi, struct list_head *inode_list,
 		if (!f2fs_is_valid_blkaddr(sbi, blkaddr, META_POR))
 			break;
 
-		f2fs_ra_meta_pages_cond(sbi, blkaddr);
-
 		page = f2fs_get_tmp_page(sbi, blkaddr);
 		if (IS_ERR(page)) {
 			err = PTR_ERR(page);
@@ -759,9 +779,14 @@ static int recover_data(struct f2fs_sb_info *sbi, struct list_head *inode_list,
 		if (entry->blkaddr == blkaddr)
 			list_move_tail(&entry->list, tmp_inode_list);
 next:
+		ra_blocks = adjust_por_ra_blocks(sbi, ra_blocks, blkaddr,
+						next_blkaddr_of_node(page));
+
 		/* check next segment */
 		blkaddr = next_blkaddr_of_node(page);
 		f2fs_put_page(page, 1);
+
+		f2fs_ra_meta_pages_cond(sbi, blkaddr, ra_blocks);
 	}
 	if (!err)
 		f2fs_allocate_new_segments(sbi);
@@ -796,7 +821,7 @@ int f2fs_recover_fsync_data(struct f2fs_sb_info *sbi, bool check_only)
 	INIT_LIST_HEAD(&dir_list);
 
 	/* prevent checkpoint */
-	down_write(&sbi->cp_global_sem);
+	f2fs_down_write(&sbi->cp_global_sem);
 
 	/* step #1: find fsynced inode numbers */
 	err = find_fsync_dnodes(sbi, &inode_list, check_only);
@@ -845,7 +870,7 @@ int f2fs_recover_fsync_data(struct f2fs_sb_info *sbi, bool check_only)
 	if (!err)
 		clear_sbi_flag(sbi, SBI_POR_DOING);
 
-	up_write(&sbi->cp_global_sem);
+	f2fs_up_write(&sbi->cp_global_sem);
 
 	/* let's drop all the directory inodes for clean checkpoint */
 	destroy_fsync_dnodes(&dir_list, err);

fs/f2fs/segment.c

@@ -471,7 +471,7 @@ int f2fs_commit_inmem_pages(struct inode *inode)
 
 	f2fs_balance_fs(sbi, true);
 
-	down_write(&fi->i_gc_rwsem[WRITE]);
+	f2fs_down_write(&fi->i_gc_rwsem[WRITE]);
 
 	f2fs_lock_op(sbi);
 	set_inode_flag(inode, FI_ATOMIC_COMMIT);
@@ -483,7 +483,7 @@ int f2fs_commit_inmem_pages(struct inode *inode)
 	clear_inode_flag(inode, FI_ATOMIC_COMMIT);
 
 	f2fs_unlock_op(sbi);
-	up_write(&fi->i_gc_rwsem[WRITE]);
+	f2fs_up_write(&fi->i_gc_rwsem[WRITE]);
 
 	return err;
 }
@@ -521,7 +521,7 @@ void f2fs_balance_fs(struct f2fs_sb_info *sbi, bool need)
 			io_schedule();
 			finish_wait(&sbi->gc_thread->fggc_wq, &wait);
 		} else {
-			down_write(&sbi->gc_lock);
+			f2fs_down_write(&sbi->gc_lock);
 			f2fs_gc(sbi, false, false, false, NULL_SEGNO);
 		}
 	}
@@ -529,7 +529,7 @@ void f2fs_balance_fs(struct f2fs_sb_info *sbi, bool need)
 
 static inline bool excess_dirty_threshold(struct f2fs_sb_info *sbi)
 {
-	int factor = rwsem_is_locked(&sbi->cp_rwsem) ? 3 : 2;
+	int factor = f2fs_rwsem_is_locked(&sbi->cp_rwsem) ? 3 : 2;
 	unsigned int dents = get_pages(sbi, F2FS_DIRTY_DENTS);
 	unsigned int qdata = get_pages(sbi, F2FS_DIRTY_QDATA);
 	unsigned int nodes = get_pages(sbi, F2FS_DIRTY_NODES);
@@ -570,7 +570,7 @@ void f2fs_balance_fs_bg(struct f2fs_sb_info *sbi, bool from_bg)
 
 	/* there is background inflight IO or foreground operation recently */
 	if (is_inflight_io(sbi, REQ_TIME) ||
-		(!f2fs_time_over(sbi, REQ_TIME) && rwsem_is_locked(&sbi->cp_rwsem)))
+		(!f2fs_time_over(sbi, REQ_TIME) && f2fs_rwsem_is_locked(&sbi->cp_rwsem)))
 		return;
 
 	/* exceed periodical checkpoint timeout threshold */
@@ -1156,14 +1156,14 @@ static void __init_discard_policy(struct f2fs_sb_info *sbi,
 	dpolicy->ordered = false;
 	dpolicy->granularity = granularity;
 
-	dpolicy->max_requests = DEF_MAX_DISCARD_REQUEST;
+	dpolicy->max_requests = dcc->max_discard_request;
 	dpolicy->io_aware_gran = MAX_PLIST_NUM;
 	dpolicy->timeout = false;
 
 	if (discard_type == DPOLICY_BG) {
-		dpolicy->min_interval = DEF_MIN_DISCARD_ISSUE_TIME;
-		dpolicy->mid_interval = DEF_MID_DISCARD_ISSUE_TIME;
-		dpolicy->max_interval = DEF_MAX_DISCARD_ISSUE_TIME;
+		dpolicy->min_interval = dcc->min_discard_issue_time;
+		dpolicy->mid_interval = dcc->mid_discard_issue_time;
+		dpolicy->max_interval = dcc->max_discard_issue_time;
 		dpolicy->io_aware = true;
 		dpolicy->sync = false;
 		dpolicy->ordered = true;
@@ -1171,12 +1171,12 @@ static void __init_discard_policy(struct f2fs_sb_info *sbi,
 			dpolicy->granularity = 1;
 			if (atomic_read(&dcc->discard_cmd_cnt))
 				dpolicy->max_interval =
-					DEF_MIN_DISCARD_ISSUE_TIME;
+					dcc->min_discard_issue_time;
 		}
 	} else if (discard_type == DPOLICY_FORCE) {
-		dpolicy->min_interval = DEF_MIN_DISCARD_ISSUE_TIME;
-		dpolicy->mid_interval = DEF_MID_DISCARD_ISSUE_TIME;
-		dpolicy->max_interval = DEF_MAX_DISCARD_ISSUE_TIME;
+		dpolicy->min_interval = dcc->min_discard_issue_time;
+		dpolicy->mid_interval = dcc->mid_discard_issue_time;
+		dpolicy->max_interval = dcc->max_discard_issue_time;
 		dpolicy->io_aware = false;
 	} else if (discard_type == DPOLICY_FSTRIM) {
 		dpolicy->io_aware = false;
@@ -1781,7 +1781,7 @@ static int issue_discard_thread(void *data)
 	struct discard_cmd_control *dcc = SM_I(sbi)->dcc_info;
 	wait_queue_head_t *q = &dcc->discard_wait_queue;
 	struct discard_policy dpolicy;
-	unsigned int wait_ms = DEF_MIN_DISCARD_ISSUE_TIME;
+	unsigned int wait_ms = dcc->min_discard_issue_time;
 	int issued;
 
 	set_freezable();
@@ -2180,6 +2180,10 @@ static int create_discard_cmd_control(struct f2fs_sb_info *sbi)
 	atomic_set(&dcc->discard_cmd_cnt, 0);
 	dcc->nr_discards = 0;
 	dcc->max_discards = MAIN_SEGS(sbi) << sbi->log_blocks_per_seg;
+	dcc->max_discard_request = DEF_MAX_DISCARD_REQUEST;
+	dcc->min_discard_issue_time = DEF_MIN_DISCARD_ISSUE_TIME;
+	dcc->mid_discard_issue_time = DEF_MID_DISCARD_ISSUE_TIME;
+	dcc->max_discard_issue_time = DEF_MAX_DISCARD_ISSUE_TIME;
 	dcc->undiscard_blks = 0;
 	dcc->next_pos = 0;
 	dcc->root = RB_ROOT_CACHED;
@@ -2821,7 +2825,7 @@ static void __f2fs_init_atgc_curseg(struct f2fs_sb_info *sbi)
 	if (!sbi->am.atgc_enabled)
 		return;
 
-	down_read(&SM_I(sbi)->curseg_lock);
+	f2fs_down_read(&SM_I(sbi)->curseg_lock);
 
 	mutex_lock(&curseg->curseg_mutex);
 	down_write(&SIT_I(sbi)->sentry_lock);
@@ -2831,7 +2835,7 @@ static void __f2fs_init_atgc_curseg(struct f2fs_sb_info *sbi)
 	up_write(&SIT_I(sbi)->sentry_lock);
 	mutex_unlock(&curseg->curseg_mutex);
 
-	up_read(&SM_I(sbi)->curseg_lock);
+	f2fs_up_read(&SM_I(sbi)->curseg_lock);
 
 }
 void f2fs_init_inmem_curseg(struct f2fs_sb_info *sbi)
@@ -2982,7 +2986,7 @@ void f2fs_allocate_segment_for_resize(struct f2fs_sb_info *sbi, int type,
 	struct curseg_info *curseg = CURSEG_I(sbi, type);
 	unsigned int segno;
 
-	down_read(&SM_I(sbi)->curseg_lock);
+	f2fs_down_read(&SM_I(sbi)->curseg_lock);
 	mutex_lock(&curseg->curseg_mutex);
 	down_write(&SIT_I(sbi)->sentry_lock);
 
@@ -3006,7 +3010,7 @@ void f2fs_allocate_segment_for_resize(struct f2fs_sb_info *sbi, int type,
 			    type, segno, curseg->segno);
 
 	mutex_unlock(&curseg->curseg_mutex);
-	up_read(&SM_I(sbi)->curseg_lock);
+	f2fs_up_read(&SM_I(sbi)->curseg_lock);
 }
 
 static void __allocate_new_segment(struct f2fs_sb_info *sbi, int type,
@@ -3038,23 +3042,23 @@ static void __allocate_new_section(struct f2fs_sb_info *sbi,
 
 void f2fs_allocate_new_section(struct f2fs_sb_info *sbi, int type, bool force)
 {
-	down_read(&SM_I(sbi)->curseg_lock);
+	f2fs_down_read(&SM_I(sbi)->curseg_lock);
 	down_write(&SIT_I(sbi)->sentry_lock);
 	__allocate_new_section(sbi, type, force);
 	up_write(&SIT_I(sbi)->sentry_lock);
-	up_read(&SM_I(sbi)->curseg_lock);
+	f2fs_up_read(&SM_I(sbi)->curseg_lock);
 }
 
 void f2fs_allocate_new_segments(struct f2fs_sb_info *sbi)
 {
 	int i;
 
-	down_read(&SM_I(sbi)->curseg_lock);
+	f2fs_down_read(&SM_I(sbi)->curseg_lock);
 	down_write(&SIT_I(sbi)->sentry_lock);
 	for (i = CURSEG_HOT_DATA; i <= CURSEG_COLD_DATA; i++)
 		__allocate_new_segment(sbi, i, false, false);
 	up_write(&SIT_I(sbi)->sentry_lock);
-	up_read(&SM_I(sbi)->curseg_lock);
+	f2fs_up_read(&SM_I(sbi)->curseg_lock);
 }
 
 static const struct segment_allocation default_salloc_ops = {
@@ -3192,9 +3196,9 @@ int f2fs_trim_fs(struct f2fs_sb_info *sbi, struct fstrim_range *range)
 	if (sbi->discard_blks == 0)
 		goto out;
 
-	down_write(&sbi->gc_lock);
+	f2fs_down_write(&sbi->gc_lock);
 	err = f2fs_write_checkpoint(sbi, &cpc);
-	up_write(&sbi->gc_lock);
+	f2fs_up_write(&sbi->gc_lock);
 	if (err)
 		goto out;
 
@@ -3431,7 +3435,7 @@ void f2fs_allocate_data_block(struct f2fs_sb_info *sbi, struct page *page,
 	bool from_gc = (type == CURSEG_ALL_DATA_ATGC);
 	struct seg_entry *se = NULL;
 
-	down_read(&SM_I(sbi)->curseg_lock);
+	f2fs_down_read(&SM_I(sbi)->curseg_lock);
 
 	mutex_lock(&curseg->curseg_mutex);
 	down_write(&sit_i->sentry_lock);
@@ -3514,7 +3518,7 @@ void f2fs_allocate_data_block(struct f2fs_sb_info *sbi, struct page *page,
 
 	mutex_unlock(&curseg->curseg_mutex);
 
-	up_read(&SM_I(sbi)->curseg_lock);
+	f2fs_up_read(&SM_I(sbi)->curseg_lock);
 }
 
 void f2fs_update_device_state(struct f2fs_sb_info *sbi, nid_t ino,
@@ -3550,7 +3554,7 @@ static void do_write_page(struct f2fs_summary *sum, struct f2fs_io_info *fio)
 	bool keep_order = (f2fs_lfs_mode(fio->sbi) && type == CURSEG_COLD_DATA);
 
 	if (keep_order)
-		down_read(&fio->sbi->io_order_lock);
+		f2fs_down_read(&fio->sbi->io_order_lock);
 reallocate:
 	f2fs_allocate_data_block(fio->sbi, fio->page, fio->old_blkaddr,
 			&fio->new_blkaddr, sum, type, fio);
@@ -3570,7 +3574,7 @@ static void do_write_page(struct f2fs_summary *sum, struct f2fs_io_info *fio)
 	f2fs_update_device_state(fio->sbi, fio->ino, fio->new_blkaddr, 1);
 
 	if (keep_order)
-		up_read(&fio->sbi->io_order_lock);
+		f2fs_up_read(&fio->sbi->io_order_lock);
 }
 
 void f2fs_do_write_meta_page(struct f2fs_sb_info *sbi, struct page *page,
@@ -3705,7 +3709,7 @@ void f2fs_do_replace_block(struct f2fs_sb_info *sbi, struct f2fs_summary *sum,
 	se = get_seg_entry(sbi, segno);
 	type = se->type;
 
-	down_write(&SM_I(sbi)->curseg_lock);
+	f2fs_down_write(&SM_I(sbi)->curseg_lock);
 
 	if (!recover_curseg) {
 		/* for recovery flow */
@@ -3774,7 +3778,7 @@ void f2fs_do_replace_block(struct f2fs_sb_info *sbi, struct f2fs_summary *sum,
 
 	up_write(&sit_i->sentry_lock);
 	mutex_unlock(&curseg->curseg_mutex);
-	up_write(&SM_I(sbi)->curseg_lock);
+	f2fs_up_write(&SM_I(sbi)->curseg_lock);
 }
 
 void f2fs_replace_block(struct f2fs_sb_info *sbi, struct dnode_of_data *dn,
@@ -4789,6 +4793,13 @@ static int sanity_check_curseg(struct f2fs_sb_info *sbi)
 
 		sanity_check_seg_type(sbi, curseg->seg_type);
 
+		if (curseg->alloc_type != LFS && curseg->alloc_type != SSR) {
+			f2fs_err(sbi,
+				 "Current segment has invalid alloc_type:%d",
+				 curseg->alloc_type);
+			return -EFSCORRUPTED;
+		}
+
 		if (f2fs_test_bit(blkofs, se->cur_valid_map))
 			goto out;
 
@@ -5258,7 +5269,7 @@ int f2fs_build_segment_manager(struct f2fs_sb_info *sbi)
 
 	INIT_LIST_HEAD(&sm_info->sit_entry_set);
 
-	init_rwsem(&sm_info->curseg_lock);
+	init_f2fs_rwsem(&sm_info->curseg_lock);
 
 	if (!f2fs_readonly(sbi->sb)) {
 		err = f2fs_create_flush_cmd_control(sbi);

fs/f2fs/segment.h

@@ -651,7 +651,9 @@ static inline int utilization(struct f2fs_sb_info *sbi)
  *                     pages over min_fsync_blocks. (=default option)
  * F2FS_IPU_ASYNC - do IPU given by asynchronous write requests.
  * F2FS_IPU_NOCACHE - disable IPU bio cache.
- * F2FS_IPUT_DISABLE - disable IPU. (=default option in LFS mode)
+ * F2FS_IPU_HONOR_OPU_WRITE - use OPU write prior to IPU write if inode has
+ *                            FI_OPU_WRITE flag.
+ * F2FS_IPU_DISABLE - disable IPU. (=default option in LFS mode)
  */
 #define DEF_MIN_IPU_UTIL	70
 #define DEF_MIN_FSYNC_BLOCKS	8
@@ -667,6 +669,7 @@ enum {
 	F2FS_IPU_FSYNC,
 	F2FS_IPU_ASYNC,
 	F2FS_IPU_NOCACHE,
+	F2FS_IPU_HONOR_OPU_WRITE,
 };
 
 static inline unsigned int curseg_segno(struct f2fs_sb_info *sbi,

fs/f2fs/super.c

@@ -1355,16 +1355,16 @@ static struct inode *f2fs_alloc_inode(struct super_block *sb)
 	/* Initialize f2fs-specific inode info */
 	atomic_set(&fi->dirty_pages, 0);
 	atomic_set(&fi->i_compr_blocks, 0);
-	init_rwsem(&fi->i_sem);
+	init_f2fs_rwsem(&fi->i_sem);
 	spin_lock_init(&fi->i_size_lock);
 	INIT_LIST_HEAD(&fi->dirty_list);
 	INIT_LIST_HEAD(&fi->gdirty_list);
 	INIT_LIST_HEAD(&fi->inmem_ilist);
 	INIT_LIST_HEAD(&fi->inmem_pages);
 	mutex_init(&fi->inmem_lock);
-	init_rwsem(&fi->i_gc_rwsem[READ]);
-	init_rwsem(&fi->i_gc_rwsem[WRITE]);
-	init_rwsem(&fi->i_xattr_sem);
+	init_f2fs_rwsem(&fi->i_gc_rwsem[READ]);
+	init_f2fs_rwsem(&fi->i_gc_rwsem[WRITE]);
+	init_f2fs_rwsem(&fi->i_xattr_sem);
 
 	/* Will be used by directory only */
 	fi->i_dir_level = F2FS_SB(sb)->dir_level;
@@ -1501,8 +1501,9 @@ static void f2fs_free_inode(struct inode *inode)
 
 static void destroy_percpu_info(struct f2fs_sb_info *sbi)
 {
-	percpu_counter_destroy(&sbi->alloc_valid_block_count);
 	percpu_counter_destroy(&sbi->total_valid_inode_count);
+	percpu_counter_destroy(&sbi->rf_node_block_count);
+	percpu_counter_destroy(&sbi->alloc_valid_block_count);
 }
 
 static void destroy_device_list(struct f2fs_sb_info *sbi)
@@ -1662,11 +1663,15 @@ static int f2fs_freeze(struct super_block *sb)
 	/* ensure no checkpoint required */
 	if (!llist_empty(&F2FS_SB(sb)->cprc_info.issue_list))
 		return -EINVAL;
+
+	/* to avoid deadlock on f2fs_evict_inode->SB_FREEZE_FS */
+	set_sbi_flag(F2FS_SB(sb), SBI_IS_FREEZING);
 	return 0;
 }
 
 static int f2fs_unfreeze(struct super_block *sb)
 {
+	clear_sbi_flag(F2FS_SB(sb), SBI_IS_FREEZING);
 	return 0;
 }
 
@@ -2075,6 +2080,7 @@ static int f2fs_disable_checkpoint(struct f2fs_sb_info *sbi)
 {
 	unsigned int s_flags = sbi->sb->s_flags;
 	struct cp_control cpc;
+	unsigned int gc_mode;
 	int err = 0;
 	int ret;
 	block_t unusable;
@@ -2087,8 +2093,11 @@ static int f2fs_disable_checkpoint(struct f2fs_sb_info *sbi)
 
 	f2fs_update_time(sbi, DISABLE_TIME);
 
+	gc_mode = sbi->gc_mode;
+	sbi->gc_mode = GC_URGENT_HIGH;
+
 	while (!f2fs_time_over(sbi, DISABLE_TIME)) {
-		down_write(&sbi->gc_lock);
+		f2fs_down_write(&sbi->gc_lock);
 		err = f2fs_gc(sbi, true, false, false, NULL_SEGNO);
 		if (err == -ENODATA) {
 			err = 0;
@@ -2110,7 +2119,7 @@ static int f2fs_disable_checkpoint(struct f2fs_sb_info *sbi)
 		goto restore_flag;
 	}
 
-	down_write(&sbi->gc_lock);
+	f2fs_down_write(&sbi->gc_lock);
 	cpc.reason = CP_PAUSE;
 	set_sbi_flag(sbi, SBI_CP_DISABLED);
 	err = f2fs_write_checkpoint(sbi, &cpc);
@@ -2122,8 +2131,9 @@ static int f2fs_disable_checkpoint(struct f2fs_sb_info *sbi)
 	spin_unlock(&sbi->stat_lock);
 
 out_unlock:
-	up_write(&sbi->gc_lock);
+	f2fs_up_write(&sbi->gc_lock);
 restore_flag:
+	sbi->gc_mode = gc_mode;
 	sbi->sb->s_flags = s_flags;	/* Restore SB_RDONLY status */
 	return err;
 }
@@ -2142,12 +2152,12 @@ static void f2fs_enable_checkpoint(struct f2fs_sb_info *sbi)
 	if (unlikely(retry < 0))
 		f2fs_warn(sbi, "checkpoint=enable has some unwritten data.");
 
-	down_write(&sbi->gc_lock);
+	f2fs_down_write(&sbi->gc_lock);
 	f2fs_dirty_to_prefree(sbi);
 
 	clear_sbi_flag(sbi, SBI_CP_DISABLED);
 	set_sbi_flag(sbi, SBI_IS_DIRTY);
-	up_write(&sbi->gc_lock);
+	f2fs_up_write(&sbi->gc_lock);
 
 	f2fs_sync_fs(sbi->sb, 1);
 }
@@ -2688,7 +2698,7 @@ int f2fs_quota_sync(struct super_block *sb, int type)
 	struct f2fs_sb_info *sbi = F2FS_SB(sb);
 	struct quota_info *dqopt = sb_dqopt(sb);
 	int cnt;
-	int ret;
+	int ret = 0;
 
 	/*
 	 * Now when everything is written we can discard the pagecache so
@@ -2699,26 +2709,26 @@ int f2fs_quota_sync(struct super_block *sb, int type)
 		if (type != -1 && cnt != type)
 			continue;
 
-		if (!sb_has_quota_active(sb, type))
-			return 0;
+		if (!sb_has_quota_active(sb, cnt))
+			continue;
 
 		inode_lock(dqopt->files[cnt]);
 
 		/*
 		 * do_quotactl
 		 *  f2fs_quota_sync
-		 *  down_read(quota_sem)
+		 *  f2fs_down_read(quota_sem)
 		 *  dquot_writeback_dquots()
 		 *  f2fs_dquot_commit
 		 *			      block_operation
-		 *			      down_read(quota_sem)
+		 *			      f2fs_down_read(quota_sem)
 		 */
 		f2fs_lock_op(sbi);
-		down_read(&sbi->quota_sem);
+		f2fs_down_read(&sbi->quota_sem);
 
 		ret = f2fs_quota_sync_file(sbi, cnt);
 
-		up_read(&sbi->quota_sem);
+		f2fs_up_read(&sbi->quota_sem);
 		f2fs_unlock_op(sbi);
 
 		inode_unlock(dqopt->files[cnt]);
@@ -2843,11 +2853,11 @@ static int f2fs_dquot_commit(struct dquot *dquot)
 	struct f2fs_sb_info *sbi = F2FS_SB(dquot->dq_sb);
 	int ret;
 
-	down_read_nested(&sbi->quota_sem, SINGLE_DEPTH_NESTING);
+	f2fs_down_read_nested(&sbi->quota_sem, SINGLE_DEPTH_NESTING);
 	ret = dquot_commit(dquot);
 	if (ret < 0)
 		set_sbi_flag(sbi, SBI_QUOTA_NEED_REPAIR);
-	up_read(&sbi->quota_sem);
+	f2fs_up_read(&sbi->quota_sem);
 	return ret;
 }
 
@@ -2856,11 +2866,11 @@ static int f2fs_dquot_acquire(struct dquot *dquot)
 	struct f2fs_sb_info *sbi = F2FS_SB(dquot->dq_sb);
 	int ret;
 
-	down_read(&sbi->quota_sem);
+	f2fs_down_read(&sbi->quota_sem);
 	ret = dquot_acquire(dquot);
 	if (ret < 0)
 		set_sbi_flag(sbi, SBI_QUOTA_NEED_REPAIR);
-	up_read(&sbi->quota_sem);
+	f2fs_up_read(&sbi->quota_sem);
 	return ret;
 }
 
@@ -3574,6 +3584,7 @@ static void init_sb_info(struct f2fs_sb_info *sbi)
 	F2FS_NODE_INO(sbi) = le32_to_cpu(raw_super->node_ino);
 	F2FS_META_INO(sbi) = le32_to_cpu(raw_super->meta_ino);
 	sbi->cur_victim_sec = NULL_SECNO;
+	sbi->gc_mode = GC_NORMAL;
 	sbi->next_victim_seg[BG_GC] = NULL_SEGNO;
 	sbi->next_victim_seg[FG_GC] = NULL_SEGNO;
 	sbi->max_victim_search = DEF_MAX_VICTIM_SEARCH;
@@ -3601,14 +3612,14 @@ static void init_sb_info(struct f2fs_sb_info *sbi)
 
 	INIT_LIST_HEAD(&sbi->s_list);
 	mutex_init(&sbi->umount_mutex);
-	init_rwsem(&sbi->io_order_lock);
+	init_f2fs_rwsem(&sbi->io_order_lock);
 	spin_lock_init(&sbi->cp_lock);
 
 	sbi->dirty_device = 0;
 	spin_lock_init(&sbi->dev_lock);
 
-	init_rwsem(&sbi->sb_lock);
-	init_rwsem(&sbi->pin_sem);
+	init_f2fs_rwsem(&sbi->sb_lock);
+	init_f2fs_rwsem(&sbi->pin_sem);
 }
 
 static int init_percpu_info(struct f2fs_sb_info *sbi)
@@ -3619,11 +3630,20 @@ static int init_percpu_info(struct f2fs_sb_info *sbi)
 	if (err)
 		return err;
 
+	err = percpu_counter_init(&sbi->rf_node_block_count, 0, GFP_KERNEL);
+	if (err)
+		goto err_valid_block;
+
 	err = percpu_counter_init(&sbi->total_valid_inode_count, 0,
 								GFP_KERNEL);
 	if (err)
-		percpu_counter_destroy(&sbi->alloc_valid_block_count);
+		goto err_node_block;
+	return 0;
 
+err_node_block:
+	percpu_counter_destroy(&sbi->rf_node_block_count);
+err_valid_block:
+	percpu_counter_destroy(&sbi->alloc_valid_block_count);
 	return err;
 }
 
@@ -3957,7 +3977,8 @@ static void f2fs_tuning_parameters(struct f2fs_sb_info *sbi)
 		F2FS_OPTION(sbi).alloc_mode = ALLOC_MODE_REUSE;
 		if (f2fs_block_unit_discard(sbi))
 			sm_i->dcc_info->discard_granularity = 1;
-		sm_i->ipu_policy = 1 << F2FS_IPU_FORCE;
+		sm_i->ipu_policy = 1 << F2FS_IPU_FORCE |
+					1 << F2FS_IPU_HONOR_OPU_WRITE;
 	}
 
 	sbi->readdir_ra = 1;
@@ -4067,11 +4088,11 @@ static int f2fs_fill_super(struct super_block *sb, void *data, int silent)
 
 	/* init f2fs-specific super block info */
 	sbi->valid_super_block = valid_super_block;
-	init_rwsem(&sbi->gc_lock);
+	init_f2fs_rwsem(&sbi->gc_lock);
 	mutex_init(&sbi->writepages);
-	init_rwsem(&sbi->cp_global_sem);
-	init_rwsem(&sbi->node_write);
-	init_rwsem(&sbi->node_change);
+	init_f2fs_rwsem(&sbi->cp_global_sem);
+	init_f2fs_rwsem(&sbi->node_write);
+	init_f2fs_rwsem(&sbi->node_change);
 
 	/* disallow all the data/node/meta page writes */
 	set_sbi_flag(sbi, SBI_POR_DOING);
@@ -4092,18 +4113,18 @@ static int f2fs_fill_super(struct super_block *sb, void *data, int silent)
 		}
 
 		for (j = HOT; j < n; j++) {
-			init_rwsem(&sbi->write_io[i][j].io_rwsem);
+			init_f2fs_rwsem(&sbi->write_io[i][j].io_rwsem);
 			sbi->write_io[i][j].sbi = sbi;
 			sbi->write_io[i][j].bio = NULL;
 			spin_lock_init(&sbi->write_io[i][j].io_lock);
 			INIT_LIST_HEAD(&sbi->write_io[i][j].io_list);
 			INIT_LIST_HEAD(&sbi->write_io[i][j].bio_list);
-			init_rwsem(&sbi->write_io[i][j].bio_list_lock);
+			init_f2fs_rwsem(&sbi->write_io[i][j].bio_list_lock);
 		}
 	}
 
-	init_rwsem(&sbi->cp_rwsem);
-	init_rwsem(&sbi->quota_sem);
+	init_f2fs_rwsem(&sbi->cp_rwsem);
+	init_f2fs_rwsem(&sbi->quota_sem);
 	init_waitqueue_head(&sbi->cp_wait);
 	init_sb_info(sbi);
 
@@ -4528,7 +4549,7 @@ static struct file_system_type f2fs_fs_type = {
 	.name		= "f2fs",
 	.mount		= f2fs_mount,
 	.kill_sb	= kill_f2fs_super,
-	.fs_flags	= FS_REQUIRES_DEV,
+	.fs_flags	= FS_REQUIRES_DEV | FS_ALLOW_IDMAP,
 };
 MODULE_ALIAS_FS("f2fs");
 

fs/f2fs/sysfs.c

@@ -41,6 +41,16 @@ enum {
 	ATGC_INFO,	/* struct atgc_management */
 };
 
+static const char *gc_mode_names[MAX_GC_MODE] = {
+	"GC_NORMAL",
+	"GC_IDLE_CB",
+	"GC_IDLE_GREEDY",
+	"GC_IDLE_AT",
+	"GC_URGENT_HIGH",
+	"GC_URGENT_LOW",
+	"GC_URGENT_MID"
+};
+
 struct f2fs_attr {
 	struct attribute attr;
 	ssize_t (*show)(struct f2fs_attr *, struct f2fs_sb_info *, char *);
@@ -316,8 +326,13 @@ static ssize_t f2fs_sbi_show(struct f2fs_attr *a,
 		return sysfs_emit(buf, "%u\n", sbi->compr_new_inode);
 #endif
 
+	if (!strcmp(a->attr.name, "gc_urgent"))
+		return sysfs_emit(buf, "%s\n",
+				gc_mode_names[sbi->gc_mode]);
+
 	if (!strcmp(a->attr.name, "gc_segment_mode"))
-		return sysfs_emit(buf, "%u\n", sbi->gc_segment_mode);
+		return sysfs_emit(buf, "%s\n",
+				gc_mode_names[sbi->gc_segment_mode]);
 
 	if (!strcmp(a->attr.name, "gc_reclaimed_segments")) {
 		return sysfs_emit(buf, "%u\n",
@@ -363,7 +378,7 @@ static ssize_t __sbi_store(struct f2fs_attr *a,
 		if (!strlen(name) || strlen(name) >= F2FS_EXTENSION_LEN)
 			return -EINVAL;
 
-		down_write(&sbi->sb_lock);
+		f2fs_down_write(&sbi->sb_lock);
 
 		ret = f2fs_update_extension_list(sbi, name, hot, set);
 		if (ret)
@@ -373,7 +388,7 @@ static ssize_t __sbi_store(struct f2fs_attr *a,
 		if (ret)
 			f2fs_update_extension_list(sbi, name, hot, !set);
 out:
-		up_write(&sbi->sb_lock);
+		f2fs_up_write(&sbi->sb_lock);
 		return ret ? ret : count;
 	}
 
@@ -468,6 +483,13 @@ static ssize_t __sbi_store(struct f2fs_attr *a,
 			}
 		} else if (t == 2) {
 			sbi->gc_mode = GC_URGENT_LOW;
+		} else if (t == 3) {
+			sbi->gc_mode = GC_URGENT_MID;
+			if (sbi->gc_thread) {
+				sbi->gc_thread->gc_wake = 1;
+				wake_up_interruptible_all(
+					&sbi->gc_thread->gc_wait_queue_head);
+			}
 		} else {
 			return -EINVAL;
 		}
@@ -481,7 +503,7 @@ static ssize_t __sbi_store(struct f2fs_attr *a,
 		} else if (t == GC_IDLE_AT) {
 			if (!sbi->am.atgc_enabled)
 				return -EINVAL;
-			sbi->gc_mode = GC_AT;
+			sbi->gc_mode = GC_IDLE_AT;
 		} else {
 			sbi->gc_mode = GC_NORMAL;
 		}
@@ -716,6 +738,10 @@ F2FS_RW_ATTR(F2FS_SBI, f2fs_sb_info, gc_idle, gc_mode);
 F2FS_RW_ATTR(F2FS_SBI, f2fs_sb_info, gc_urgent, gc_mode);
 F2FS_RW_ATTR(SM_INFO, f2fs_sm_info, reclaim_segments, rec_prefree_segments);
 F2FS_RW_ATTR(DCC_INFO, discard_cmd_control, max_small_discards, max_discards);
+F2FS_RW_ATTR(DCC_INFO, discard_cmd_control, max_discard_request, max_discard_request);
+F2FS_RW_ATTR(DCC_INFO, discard_cmd_control, min_discard_issue_time, min_discard_issue_time);
+F2FS_RW_ATTR(DCC_INFO, discard_cmd_control, mid_discard_issue_time, mid_discard_issue_time);
+F2FS_RW_ATTR(DCC_INFO, discard_cmd_control, max_discard_issue_time, max_discard_issue_time);
 F2FS_RW_ATTR(DCC_INFO, discard_cmd_control, discard_granularity, discard_granularity);
 F2FS_RW_ATTR(RESERVED_BLOCKS, f2fs_sb_info, reserved_blocks, reserved_blocks);
 F2FS_RW_ATTR(SM_INFO, f2fs_sm_info, batched_trim_sections, trim_sections);
@@ -728,6 +754,7 @@ F2FS_RW_ATTR(SM_INFO, f2fs_sm_info, min_ssr_sections, min_ssr_sections);
 F2FS_RW_ATTR(NM_INFO, f2fs_nm_info, ram_thresh, ram_thresh);
 F2FS_RW_ATTR(NM_INFO, f2fs_nm_info, ra_nid_pages, ra_nid_pages);
 F2FS_RW_ATTR(NM_INFO, f2fs_nm_info, dirty_nats_ratio, dirty_nats_ratio);
+F2FS_RW_ATTR(NM_INFO, f2fs_nm_info, max_roll_forward_node_blocks, max_rf_node_blocks);
 F2FS_RW_ATTR(F2FS_SBI, f2fs_sb_info, max_victim_search, max_victim_search);
 F2FS_RW_ATTR(F2FS_SBI, f2fs_sb_info, migration_granularity, migration_granularity);
 F2FS_RW_ATTR(F2FS_SBI, f2fs_sb_info, dir_level, dir_level);
@@ -832,6 +859,10 @@ static struct attribute *f2fs_attrs[] = {
 	ATTR_LIST(reclaim_segments),
 	ATTR_LIST(main_blkaddr),
 	ATTR_LIST(max_small_discards),
+	ATTR_LIST(max_discard_request),
+	ATTR_LIST(min_discard_issue_time),
+	ATTR_LIST(mid_discard_issue_time),
+	ATTR_LIST(max_discard_issue_time),
 	ATTR_LIST(discard_granularity),
 	ATTR_LIST(pending_discard),
 	ATTR_LIST(batched_trim_sections),
@@ -847,6 +878,7 @@ static struct attribute *f2fs_attrs[] = {
 	ATTR_LIST(ram_thresh),
 	ATTR_LIST(ra_nid_pages),
 	ATTR_LIST(dirty_nats_ratio),
+	ATTR_LIST(max_roll_forward_node_blocks),
 	ATTR_LIST(cp_interval),
 	ATTR_LIST(idle_interval),
 	ATTR_LIST(discard_idle_interval),

fs/f2fs/verity.c

@@ -208,7 +208,7 @@ static int f2fs_end_enable_verity(struct file *filp, const void *desc,
 	 * from re-instantiating cached pages we are truncating (since unlike
 	 * normal file accesses, garbage collection isn't limited by i_size).
 	 */
-	down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+	f2fs_down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
 	truncate_inode_pages(inode->i_mapping, inode->i_size);
 	err2 = f2fs_truncate(inode);
 	if (err2) {
@@ -216,7 +216,7 @@ static int f2fs_end_enable_verity(struct file *filp, const void *desc,
 			 err2);
 		set_sbi_flag(sbi, SBI_NEED_FSCK);
 	}
-	up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+	f2fs_up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
 	clear_inode_flag(inode, FI_VERITY_IN_PROGRESS);
 	return err ?: err2;
 }

fs/f2fs/xattr.c

@@ -525,10 +525,10 @@ int f2fs_getxattr(struct inode *inode, int index, const char *name,
 	if (len > F2FS_NAME_LEN)
 		return -ERANGE;
 
-	down_read(&F2FS_I(inode)->i_xattr_sem);
+	f2fs_down_read(&F2FS_I(inode)->i_xattr_sem);
 	error = lookup_all_xattrs(inode, ipage, index, len, name,
 				&entry, &base_addr, &base_size, &is_inline);
-	up_read(&F2FS_I(inode)->i_xattr_sem);
+	f2fs_up_read(&F2FS_I(inode)->i_xattr_sem);
 	if (error)
 		return error;
 
@@ -562,9 +562,9 @@ ssize_t f2fs_listxattr(struct dentry *dentry, char *buffer, size_t buffer_size)
 	int error;
 	size_t rest = buffer_size;
 
-	down_read(&F2FS_I(inode)->i_xattr_sem);
+	f2fs_down_read(&F2FS_I(inode)->i_xattr_sem);
 	error = read_all_xattrs(inode, NULL, &base_addr);
-	up_read(&F2FS_I(inode)->i_xattr_sem);
+	f2fs_up_read(&F2FS_I(inode)->i_xattr_sem);
 	if (error)
 		return error;
 
@@ -786,9 +786,9 @@ int f2fs_setxattr(struct inode *inode, int index, const char *name,
 	f2fs_balance_fs(sbi, true);
 
 	f2fs_lock_op(sbi);
-	down_write(&F2FS_I(inode)->i_xattr_sem);
+	f2fs_down_write(&F2FS_I(inode)->i_xattr_sem);
 	err = __f2fs_setxattr(inode, index, name, value, size, ipage, flags);
-	up_write(&F2FS_I(inode)->i_xattr_sem);
+	f2fs_up_write(&F2FS_I(inode)->i_xattr_sem);
 	f2fs_unlock_op(sbi);
 
 	f2fs_update_time(sbi, REQ_TIME);