f2fs: do not activate auto_recovery for fallocated i_size

commit 26787236b36660baf4d136281d40b5bb33a570ec upstream.

If a file needs to keep its i_size by fallocate, we need to turn off auto
recovery during roll-forward recovery.

This will resolve the below scenario.

1. xfs_io -f /mnt/f2fs/file -c "pwrite 0 4096" -c "fsync"
2. xfs_io -f /mnt/f2fs/file -c "falloc -k 4096 4096" -c "fsync"
3. md5sum /mnt/f2fs/file;
4. godown /mnt/f2fs/
5. umount /mnt/f2fs/
6. mount -t f2fs /dev/sdx /mnt/f2fs
7. md5sum /mnt/f2fs/file

Reported-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>

Author: Jaegeuk Kim

fs/f2fs/f2fs.h

@@ -449,6 +449,7 @@ struct f2fs_map_blocks {
 #define FADVISE_LOST_PINO_BIT	0x02
 #define FADVISE_ENCRYPT_BIT	0x04
 #define FADVISE_ENC_NAME_BIT	0x08
+#define FADVISE_KEEP_SIZE_BIT	0x10
 
 #define file_is_cold(inode)	is_file(inode, FADVISE_COLD_BIT)
 #define file_wrong_pino(inode)	is_file(inode, FADVISE_LOST_PINO_BIT)
@@ -461,6 +462,8 @@ struct f2fs_map_blocks {
 #define file_clear_encrypt(inode) clear_file(inode, FADVISE_ENCRYPT_BIT)
 #define file_enc_name(inode)	is_file(inode, FADVISE_ENC_NAME_BIT)
 #define file_set_enc_name(inode) set_file(inode, FADVISE_ENC_NAME_BIT)
+#define file_keep_isize(inode)	is_file(inode, FADVISE_KEEP_SIZE_BIT)
+#define file_set_keep_isize(inode) set_file(inode, FADVISE_KEEP_SIZE_BIT)
 
 #define DEF_DIR_LEVEL		0
 
@@ -1763,23 +1766,6 @@ static inline void f2fs_i_size_write(struct inode *inode, loff_t i_size)
 		set_inode_flag(inode, FI_AUTO_RECOVER);
 }
 
-static inline bool f2fs_skip_inode_update(struct inode *inode, int dsync)
-{
-	if (dsync) {
-		struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
-		bool ret;
-
-		spin_lock(&sbi->inode_lock[DIRTY_META]);
-		ret = list_empty(&F2FS_I(inode)->gdirty_list);
-		spin_unlock(&sbi->inode_lock[DIRTY_META]);
-		return ret;
-	}
-	if (!is_inode_flag_set(inode, FI_AUTO_RECOVER) ||
-			i_size_read(inode) & PAGE_MASK)
-		return false;
-	return F2FS_I(inode)->last_disk_size == i_size_read(inode);
-}
-
 static inline void f2fs_i_depth_write(struct inode *inode, unsigned int depth)
 {
 	F2FS_I(inode)->i_current_depth = depth;
@@ -1932,6 +1918,24 @@ static inline void clear_file(struct inode *inode, int type)
 	f2fs_mark_inode_dirty_sync(inode, true);
 }
 
+static inline bool f2fs_skip_inode_update(struct inode *inode, int dsync)
+{
+	if (dsync) {
+		struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
+		bool ret;
+
+		spin_lock(&sbi->inode_lock[DIRTY_META]);
+		ret = list_empty(&F2FS_I(inode)->gdirty_list);
+		spin_unlock(&sbi->inode_lock[DIRTY_META]);
+		return ret;
+	}
+	if (!is_inode_flag_set(inode, FI_AUTO_RECOVER) ||
+			file_keep_isize(inode) ||
+			i_size_read(inode) & PAGE_MASK)
+		return false;
+	return F2FS_I(inode)->last_disk_size == i_size_read(inode);
+}
+
 static inline int f2fs_readonly(struct super_block *sb)
 {
 	return sb->s_flags & MS_RDONLY;

fs/f2fs/file.c

@@ -1403,6 +1403,8 @@ static long f2fs_fallocate(struct file *file, int mode,
 	if (!ret) {
 		inode->i_mtime = inode->i_ctime = current_time(inode);
 		f2fs_mark_inode_dirty_sync(inode, false);
+		if (mode & FALLOC_FL_KEEP_SIZE)
+			file_set_keep_isize(inode);
 		f2fs_update_time(F2FS_I_SB(inode), REQ_TIME);
 	}
 

fs/f2fs/recovery.c

@@ -187,6 +187,8 @@ static void recover_inode(struct inode *inode, struct page *page)
 	inode->i_ctime.tv_nsec = le32_to_cpu(raw->i_ctime_nsec);
 	inode->i_mtime.tv_nsec = le32_to_cpu(raw->i_mtime_nsec);
 
+	F2FS_I(inode)->i_advise = raw->i_advise;
+
 	if (file_enc_name(inode))
 		name = "<encrypted>";
 	else
@@ -425,7 +427,8 @@ static int do_recover_data(struct f2fs_sb_info *sbi, struct inode *inode,
 			continue;
 		}
 
-		if (i_size_read(inode) <= (start << PAGE_SHIFT))
+		if (!file_keep_isize(inode) &&
+				(i_size_read(inode) <= (start << PAGE_SHIFT)))
 			f2fs_i_size_write(inode, (start + 1) << PAGE_SHIFT);
 
 		/*
@@ -478,8 +481,10 @@ static int do_recover_data(struct f2fs_sb_info *sbi, struct inode *inode,
 	f2fs_put_dnode(&dn);
 out:
 	f2fs_msg(sbi->sb, KERN_NOTICE,
-		"recover_data: ino = %lx, recovered = %d blocks, err = %d",
-		inode->i_ino, recovered, err);
+		"recover_data: ino = %lx (i_size: %s) recovered = %d, err = %d",
+		inode->i_ino,
+		file_keep_isize(inode) ? "keep" : "recover",
+		recovered, err);
 	return err;
 }