Commit 04f0b2ea authored by Qiuyang Sun's avatar Qiuyang Sun Committed by Jaegeuk Kim
Browse files

f2fs: ioctl for removing a range from F2FS 



This ioctl shrinks a given length (aligned to sections) from end of the
main area. Any cursegs and valid blocks will be moved out before
invalidating the range.

This feature can be used for adjusting partition sizes online.

History of the patch:

Sahitya Tummala:
 - Add this ioctl for f2fs_compat_ioctl() as well.
 - Fix debugfs status to reflect the online resize changes.
 - Fix potential race between online resize path and allocate new data
   block path or gc path.

Others:
 - Rename some identifiers.
 - Add some error handling branches.
 - Clear sbi->next_victim_seg[BG_GC/FG_GC] in shrinking range.
 - Implement this interface as ext4's, and change the parameter from shrunk
bytes to new block count of F2FS.
 - During resizing, force to empty sit_journal and forbid adding new
   entries to it, in order to avoid invalid segno in journal after resize.
 - Reduce sbi->user_block_count before resize starts.
 - Commit the updated superblock first, and then update in-memory metadata
   only when the former succeeds.
 - Target block count must align to sections.
 - Write checkpoint before and after committing the new superblock, w/o
CP_FSCK_FLAG respectively, so that the FS can be fixed by fsck even if
resize fails after the new superblock is committed.
 - In free_segment_range(), reduce granularity of gc_mutex.
 - Add protection on curseg migration.
 - Add freeze_bdev() and thaw_bdev() for resize fs.
 - Remove CUR_MAIN_SECS and use MAIN_SECS directly for allocation.
 - Recover super_block and FS metadata when resize fails.
 - No need to clear CP_FSCK_FLAG in update_ckpt_flags().
 - Clean up the sb and fs metadata update functions for resize_fs.

Geert Uytterhoeven:
 - Use div_u64*() for 64-bit divisions

Arnd Bergmann:
 - Not all architectures support get_user() with a 64-bit argument:
    ERROR: "__get_user_bad" [fs/f2fs/f2fs.ko] undefined!
    Use copy_from_user() here, this will always work.

Signed-off-by: default avatarQiuyang Sun <sunqiuyang@huawei.com>
Signed-off-by: default avatarChao Yu <yuchao0@huawei.com>
Signed-off-by: default avatarSahitya Tummala <stummala@codeaurora.org>
Signed-off-by: default avatarGeert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: default avatarArnd Bergmann <arnd@arndb.de>
Reviewed-by: default avatarChao Yu <yuchao0@huawei.com>
Signed-off-by: default avatarJaegeuk Kim <jaegeuk@kernel.org>
parent 5043a964

fs/f2fs/checkpoint.c

+2 −1
Original line number Diff line number Diff line
@@ -1302,7 +1302,8 @@ static void update_ckpt_flags(struct f2fs_sb_info *sbi, struct cp_control *cpc) 
	else

		__clear_ckpt_flags(ckpt, CP_ORPHAN_PRESENT_FLAG);



	if (is_sbi_flag_set(sbi, SBI_NEED_FSCK))

	if (is_sbi_flag_set(sbi, SBI_NEED_FSCK) ||

		is_sbi_flag_set(sbi, SBI_IS_RESIZEFS))

		__set_ckpt_flags(ckpt, CP_FSCK_FLAG);



	if (is_sbi_flag_set(sbi, SBI_CP_DISABLED))

fs/f2fs/debug.c

+7 −0
Original line number Diff line number Diff line
@@ -27,8 +27,15 @@ static DEFINE_MUTEX(f2fs_stat_mutex); 
static void update_general_status(struct f2fs_sb_info *sbi)

{

	struct f2fs_stat_info *si = F2FS_STAT(sbi);

	struct f2fs_super_block *raw_super = F2FS_RAW_SUPER(sbi);

	int i;



	/* these will be changed if online resize is done */

	si->main_area_segs = le32_to_cpu(raw_super->segment_count_main);

	si->main_area_sections = le32_to_cpu(raw_super->section_count);

	si->main_area_zones = si->main_area_sections /

				le32_to_cpu(raw_super->secs_per_zone);



	/* validation check of the segment numbers */

	si->hit_largest = atomic64_read(&sbi->read_hit_largest);

	si->hit_cached = atomic64_read(&sbi->read_hit_cached);

fs/f2fs/f2fs.h

+6 −0
Original line number Diff line number Diff line
@@ -415,6 +415,7 @@ static inline bool __has_cursum_space(struct f2fs_journal *journal, 
#define F2FS_IOC_SET_PIN_FILE		_IOW(F2FS_IOCTL_MAGIC, 13, __u32)

#define F2FS_IOC_GET_PIN_FILE		_IOR(F2FS_IOCTL_MAGIC, 14, __u32)

#define F2FS_IOC_PRECACHE_EXTENTS	_IO(F2FS_IOCTL_MAGIC, 15)

#define F2FS_IOC_RESIZE_FS		_IOW(F2FS_IOCTL_MAGIC, 16, __u64)



#define F2FS_IOC_SET_ENCRYPTION_POLICY	FS_IOC_SET_ENCRYPTION_POLICY

#define F2FS_IOC_GET_ENCRYPTION_POLICY	FS_IOC_GET_ENCRYPTION_POLICY
@@ -1116,6 +1117,7 @@ enum { 
	SBI_QUOTA_NEED_FLUSH,			/* need to flush quota info in CP */

	SBI_QUOTA_SKIP_FLUSH,			/* skip flushing quota in current CP */

	SBI_QUOTA_NEED_REPAIR,			/* quota file may be corrupted */

	SBI_IS_RESIZEFS,			/* resizefs is in process */

};



enum {
@@ -1236,6 +1238,7 @@ struct f2fs_sb_info { 
	unsigned int segs_per_sec;		/* segments per section */

	unsigned int secs_per_zone;		/* sections per zone */

	unsigned int total_sections;		/* total section count */

	struct mutex resize_mutex;		/* for resize exclusion */

	unsigned int total_node_count;		/* total node block count */

	unsigned int total_valid_node_count;	/* valid node block count */

	loff_t max_file_blocks;			/* max block index of file */
@@ -3058,6 +3061,8 @@ block_t f2fs_get_unusable_blocks(struct f2fs_sb_info *sbi); 
int f2fs_disable_cp_again(struct f2fs_sb_info *sbi, block_t unusable);

void f2fs_release_discard_addrs(struct f2fs_sb_info *sbi);

int f2fs_npages_for_summary_flush(struct f2fs_sb_info *sbi, bool for_ra);

void allocate_segment_for_resize(struct f2fs_sb_info *sbi, int type,

					unsigned int start, unsigned int end);

void f2fs_allocate_new_segments(struct f2fs_sb_info *sbi);

int f2fs_trim_fs(struct f2fs_sb_info *sbi, struct fstrim_range *range);

bool f2fs_exist_trim_candidates(struct f2fs_sb_info *sbi,
@@ -3196,6 +3201,7 @@ block_t f2fs_start_bidx_of_node(unsigned int node_ofs, struct inode *inode); 
int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, bool background,

			unsigned int segno);

void f2fs_build_gc_manager(struct f2fs_sb_info *sbi);

int f2fs_resize_fs(struct f2fs_sb_info *sbi, __u64 block_count);



/*

 * recovery.c

fs/f2fs/file.c

+24 −0
Original line number Diff line number Diff line
@@ -3059,6 +3059,27 @@ static int f2fs_ioc_precache_extents(struct file *filp, unsigned long arg) 
	return f2fs_precache_extents(file_inode(filp));

}



static int f2fs_ioc_resize_fs(struct file *filp, unsigned long arg)

{

	struct f2fs_sb_info *sbi = F2FS_I_SB(file_inode(filp));

	__u64 block_count;

	int ret;



	if (!capable(CAP_SYS_ADMIN))

		return -EPERM;



	if (f2fs_readonly(sbi->sb))

		return -EROFS;



	if (copy_from_user(&block_count, (void __user *)arg,

			   sizeof(block_count)))

		return -EFAULT;



	ret = f2fs_resize_fs(sbi, block_count);



	return ret;

}



long f2fs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)

{

	if (unlikely(f2fs_cp_error(F2FS_I_SB(file_inode(filp)))))
@@ -3115,6 +3136,8 @@ long f2fs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) 
		return f2fs_ioc_set_pin_file(filp, arg);

	case F2FS_IOC_PRECACHE_EXTENTS:

		return f2fs_ioc_precache_extents(filp, arg);

	case F2FS_IOC_RESIZE_FS:

		return f2fs_ioc_resize_fs(filp, arg);

	default:

		return -ENOTTY;

	}
@@ -3228,6 +3251,7 @@ long f2fs_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg) 
	case F2FS_IOC_GET_PIN_FILE:

	case F2FS_IOC_SET_PIN_FILE:

	case F2FS_IOC_PRECACHE_EXTENTS:

	case F2FS_IOC_RESIZE_FS:

		break;

	default:

		return -ENOIOCTLCMD;

fs/f2fs/gc.c

+180 −2
Original line number Diff line number Diff line
@@ -311,10 +311,11 @@ static int get_victim_by_default(struct f2fs_sb_info *sbi, 
	struct sit_info *sm = SIT_I(sbi);

	struct victim_sel_policy p;

	unsigned int secno, last_victim;

	unsigned int last_segment = MAIN_SEGS(sbi);

	unsigned int last_segment;

	unsigned int nsearched = 0;



	mutex_lock(&dirty_i->seglist_lock);

	last_segment = MAIN_SECS(sbi) * sbi->segs_per_sec;



	p.alloc_mode = alloc_mode;

	select_policy(sbi, gc_type, type, &p);
@@ -405,7 +406,8 @@ next: 
				sm->last_victim[p.gc_mode] = last_victim + 1;

			else

				sm->last_victim[p.gc_mode] = segno + 1;

			sm->last_victim[p.gc_mode] %= MAIN_SEGS(sbi);

			sm->last_victim[p.gc_mode] %=

				(MAIN_SECS(sbi) * sbi->segs_per_sec);

			break;

		}

	}
@@ -1361,3 +1363,179 @@ void f2fs_build_gc_manager(struct f2fs_sb_info *sbi) 
		SIT_I(sbi)->last_victim[ALLOC_NEXT] =

				GET_SEGNO(sbi, FDEV(0).end_blk) + 1;

}



static int free_segment_range(struct f2fs_sb_info *sbi, unsigned int start,

							unsigned int end)

{

	int type;

	unsigned int segno, next_inuse;

	int err = 0;



	/* Move out cursegs from the target range */

	for (type = CURSEG_HOT_DATA; type < NR_CURSEG_TYPE; type++)

		allocate_segment_for_resize(sbi, type, start, end);



	/* do GC to move out valid blocks in the range */

	for (segno = start; segno <= end; segno += sbi->segs_per_sec) {

		struct gc_inode_list gc_list = {

			.ilist = LIST_HEAD_INIT(gc_list.ilist),

			.iroot = RADIX_TREE_INIT(gc_list.iroot, GFP_NOFS),

		};



		mutex_lock(&sbi->gc_mutex);

		do_garbage_collect(sbi, segno, &gc_list, FG_GC);

		mutex_unlock(&sbi->gc_mutex);

		put_gc_inode(&gc_list);



		if (get_valid_blocks(sbi, segno, true))

			return -EAGAIN;

	}



	err = f2fs_sync_fs(sbi->sb, 1);

	if (err)

		return err;



	next_inuse = find_next_inuse(FREE_I(sbi), end + 1, start);

	if (next_inuse <= end) {

		f2fs_msg(sbi->sb, KERN_ERR,

			"segno %u should be free but still inuse!", next_inuse);

		f2fs_bug_on(sbi, 1);

	}

	return err;

}



static void update_sb_metadata(struct f2fs_sb_info *sbi, int secs)

{

	struct f2fs_super_block *raw_sb = F2FS_RAW_SUPER(sbi);

	int section_count = le32_to_cpu(raw_sb->section_count);

	int segment_count = le32_to_cpu(raw_sb->segment_count);

	int segment_count_main = le32_to_cpu(raw_sb->segment_count_main);

	long long block_count = le64_to_cpu(raw_sb->block_count);

	int segs = secs * sbi->segs_per_sec;



	raw_sb->section_count = cpu_to_le32(section_count + secs);

	raw_sb->segment_count = cpu_to_le32(segment_count + segs);

	raw_sb->segment_count_main = cpu_to_le32(segment_count_main + segs);

	raw_sb->block_count = cpu_to_le64(block_count +

					(long long)segs * sbi->blocks_per_seg);

}



static void update_fs_metadata(struct f2fs_sb_info *sbi, int secs)

{

	int segs = secs * sbi->segs_per_sec;

	long long user_block_count =

				le64_to_cpu(F2FS_CKPT(sbi)->user_block_count);



	SM_I(sbi)->segment_count = (int)SM_I(sbi)->segment_count + segs;

	MAIN_SEGS(sbi) = (int)MAIN_SEGS(sbi) + segs;

	FREE_I(sbi)->free_sections = (int)FREE_I(sbi)->free_sections + secs;

	FREE_I(sbi)->free_segments = (int)FREE_I(sbi)->free_segments + segs;

	F2FS_CKPT(sbi)->user_block_count = cpu_to_le64(user_block_count +

					(long long)segs * sbi->blocks_per_seg);

}



int f2fs_resize_fs(struct f2fs_sb_info *sbi, __u64 block_count)

{

	__u64 old_block_count, shrunk_blocks;

	unsigned int secs;

	int gc_mode, gc_type;

	int err = 0;

	__u32 rem;



	old_block_count = le64_to_cpu(F2FS_RAW_SUPER(sbi)->block_count);

	if (block_count > old_block_count)

		return -EINVAL;



	/* new fs size should align to section size */

	div_u64_rem(block_count, BLKS_PER_SEC(sbi), &rem);

	if (rem)

		return -EINVAL;



	if (block_count == old_block_count)

		return 0;



	if (is_sbi_flag_set(sbi, SBI_NEED_FSCK)) {

		f2fs_msg(sbi->sb, KERN_ERR,

			"Should run fsck to repair first.");

		return -EINVAL;

	}



	if (test_opt(sbi, DISABLE_CHECKPOINT)) {

		f2fs_msg(sbi->sb, KERN_ERR,

			"Checkpoint should be enabled.");

		return -EINVAL;

	}



	freeze_bdev(sbi->sb->s_bdev);



	shrunk_blocks = old_block_count - block_count;

	secs = div_u64(shrunk_blocks, BLKS_PER_SEC(sbi));

	spin_lock(&sbi->stat_lock);

	if (shrunk_blocks + valid_user_blocks(sbi) +

		sbi->current_reserved_blocks + sbi->unusable_block_count +

		F2FS_OPTION(sbi).root_reserved_blocks > sbi->user_block_count)

		err = -ENOSPC;

	else

		sbi->user_block_count -= shrunk_blocks;

	spin_unlock(&sbi->stat_lock);

	if (err) {

		thaw_bdev(sbi->sb->s_bdev, sbi->sb);

		return err;

	}



	mutex_lock(&sbi->resize_mutex);

	set_sbi_flag(sbi, SBI_IS_RESIZEFS);



	mutex_lock(&DIRTY_I(sbi)->seglist_lock);



	MAIN_SECS(sbi) -= secs;



	for (gc_mode = 0; gc_mode < MAX_GC_POLICY; gc_mode++)

		if (SIT_I(sbi)->last_victim[gc_mode] >=

					MAIN_SECS(sbi) * sbi->segs_per_sec)

			SIT_I(sbi)->last_victim[gc_mode] = 0;



	for (gc_type = BG_GC; gc_type <= FG_GC; gc_type++)

		if (sbi->next_victim_seg[gc_type] >=

					MAIN_SECS(sbi) * sbi->segs_per_sec)

			sbi->next_victim_seg[gc_type] = NULL_SEGNO;



	mutex_unlock(&DIRTY_I(sbi)->seglist_lock);



	err = free_segment_range(sbi, MAIN_SECS(sbi) * sbi->segs_per_sec,

			MAIN_SEGS(sbi) - 1);

	if (err)

		goto out;



	update_sb_metadata(sbi, -secs);



	err = f2fs_commit_super(sbi, false);

	if (err) {

		update_sb_metadata(sbi, secs);

		goto out;

	}



	update_fs_metadata(sbi, -secs);

	clear_sbi_flag(sbi, SBI_IS_RESIZEFS);

	err = f2fs_sync_fs(sbi->sb, 1);

	if (err) {

		update_fs_metadata(sbi, secs);

		update_sb_metadata(sbi, secs);

		f2fs_commit_super(sbi, false);

	}

out:

	if (err) {

		set_sbi_flag(sbi, SBI_NEED_FSCK);

		f2fs_msg(sbi->sb, KERN_ERR,

				"resize_fs failed, should run fsck to repair!");



		MAIN_SECS(sbi) += secs;

		spin_lock(&sbi->stat_lock);

		sbi->user_block_count += shrunk_blocks;

		spin_unlock(&sbi->stat_lock);

	}

	clear_sbi_flag(sbi, SBI_IS_RESIZEFS);

	mutex_unlock(&sbi->resize_mutex);

	thaw_bdev(sbi->sb->s_bdev, sbi->sb);

	return err;

}

CRA Git | Maintained and supported by SUSTech CRA and CCSE