f2fs: flush data when enabling checkpoint back

During checkpoint=disable period, f2fs bypasses all the synchronous IOs such as
sync and fsync. So, when enabling it back, we must flush all of them in order
to keep the data persistent. Otherwise, suddern power-cut right after enabling
checkpoint will cause data loss.

Fixes: 4354994 ("f2fs: checkpoint disabling")
Cc: stable@vger.kernel.org
Reviewed-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
[dwoo08.lee: backport from mainline commit b0ff4fe746fd]
Signed-off-by: Dongwoo Lee <dwoo08.lee@samsung.com>
Change-Id: Ic0810d99c5a3938e191461bd66e7862a4a68e12b

diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c
index be58d77..dacb3bc 100644 (file)
--- a/fs/f2fs/super.c
+++ b/fs/f2fs/super.c
@@ -1465,6 +1465,9 @@ static int f2fs_disable_checkpoint(struct f2fs_sb_info *sbi)
 
 static void f2fs_enable_checkpoint(struct f2fs_sb_info *sbi)
 {
+       /* we should flush all the data to keep data consistency */
+       sync_inodes_sb(sbi->sb);
+
        mutex_lock(&sbi->gc_mutex);
        f2fs_dirty_to_prefree(sbi);