Commit c42d28ce authored by Chao Yu's avatar Chao Yu Committed by Jaegeuk Kim
Browse files

f2fs: fix potential data inconsistence of checkpoint 



Previously, we changed lock from cp_rwsem to node_change, it solved
the deadlock issue which was caused by below race condition:

Thread A			Thread B
- f2fs_setattr
 - f2fs_lock_op  -- read_lock
 - dquot_transfer
  - __dquot_transfer
   - dquot_acquire
    - commit_dqblk
     - f2fs_quota_write
      - f2fs_write_begin
       - f2fs_write_failed
				- write_checkpoint
				 - block_operations
				  - f2fs_lock_all  -- write_lock
        - f2fs_truncate_blocks
         - f2fs_lock_op  -- read_lock

But it breaks the sematics of cp_rwsem, in other callers like:
- f2fs_file_write_iter -> f2fs_write_begin -> f2fs_write_failed
- f2fs_direct_IO -> f2fs_write_failed

We allow to truncate dnode w/o cp_rwsem held, result in incorrect sit
bitmap update, which can cause further data corruption.

So this patch reverts previous fix implementation, and try to fix
deadlock by skipping calling f2fs_truncate_blocks() in f2fs_write_failed()
only for quota file, and keep the preallocated data/node in the tail of
quota file, we can expecte that the preallocated space can be used to
store quota info latter soon.

Fixes: af033b2a ("f2fs: guarantee journalled quota data by checkpoint")
Signed-off-by: default avatarGao Xiang <gaoxiang25@huawei.com>
Signed-off-by: default avatarSheng Yong <shengyong1@huawei.com>
Signed-off-by: default avatarChao Yu <yuchao0@huawei.com>
Signed-off-by: default avatarJaegeuk Kim <jaegeuk@kernel.org>
parent 025cdb16

fs/f2fs/data.c

+2 −1
Original line number Diff line number Diff line
@@ -2312,7 +2312,8 @@ static void f2fs_write_failed(struct address_space *mapping, loff_t to) 
		down_write(&F2FS_I(inode)->i_mmap_sem);



		truncate_pagecache(inode, i_size);

		f2fs_truncate_blocks(inode, i_size, true, true);

		if (!IS_NOQUOTA(inode))

			f2fs_truncate_blocks(inode, i_size, true);



		up_write(&F2FS_I(inode)->i_mmap_sem);

		up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);

fs/f2fs/f2fs.h

+1 −2
Original line number Diff line number Diff line
@@ -2843,8 +2843,7 @@ static inline bool is_valid_data_blkaddr(struct f2fs_sb_info *sbi, 
 */

int f2fs_sync_file(struct file *file, loff_t start, loff_t end, int datasync);

void f2fs_truncate_data_blocks(struct dnode_of_data *dn);

int f2fs_truncate_blocks(struct inode *inode, u64 from, bool lock,

							bool buf_write);

int f2fs_truncate_blocks(struct inode *inode, u64 from, bool lock);

int f2fs_truncate(struct inode *inode);

int f2fs_getattr(const struct path *path, struct kstat *stat,

			u32 request_mask, unsigned int flags);

fs/f2fs/file.c

+6 −8
Original line number Diff line number Diff line
@@ -589,8 +589,7 @@ truncate_out: 
	return 0;

}



int f2fs_truncate_blocks(struct inode *inode, u64 from, bool lock,

							bool buf_write)

int f2fs_truncate_blocks(struct inode *inode, u64 from, bool lock)

{

	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);

	struct dnode_of_data dn;
@@ -598,7 +597,6 @@ int f2fs_truncate_blocks(struct inode *inode, u64 from, bool lock, 
	int count = 0, err = 0;

	struct page *ipage;

	bool truncate_page = false;

	int flag = buf_write ? F2FS_GET_BLOCK_PRE_AIO : F2FS_GET_BLOCK_PRE_DIO;



	trace_f2fs_truncate_blocks_enter(inode, from);
@@ -608,7 +606,7 @@ int f2fs_truncate_blocks(struct inode *inode, u64 from, bool lock, 
		goto free_partial;



	if (lock)

		__do_map_lock(sbi, flag, true);

		f2fs_lock_op(sbi);



	ipage = f2fs_get_node_page(sbi, inode->i_ino);

	if (IS_ERR(ipage)) {
@@ -646,7 +644,7 @@ free_next: 
	err = f2fs_truncate_inode_blocks(inode, free_from);

out:

	if (lock)

		__do_map_lock(sbi, flag, false);

		f2fs_unlock_op(sbi);

free_partial:

	/* lastly zero out the first data page */

	if (!err)
@@ -681,7 +679,7 @@ int f2fs_truncate(struct inode *inode) 
			return err;

	}



	err = f2fs_truncate_blocks(inode, i_size_read(inode), true, false);

	err = f2fs_truncate_blocks(inode, i_size_read(inode), true);

	if (err)

		return err;
@@ -1262,7 +1260,7 @@ static int f2fs_collapse_range(struct inode *inode, loff_t offset, loff_t len) 
	new_size = i_size_read(inode) - len;

	truncate_pagecache(inode, new_size);



	ret = f2fs_truncate_blocks(inode, new_size, true, false);

	ret = f2fs_truncate_blocks(inode, new_size, true);

	up_write(&F2FS_I(inode)->i_mmap_sem);

	if (!ret)

		f2fs_i_size_write(inode, new_size);
@@ -1447,7 +1445,7 @@ static int f2fs_insert_range(struct inode *inode, loff_t offset, loff_t len) 
	f2fs_balance_fs(sbi, true);



	down_write(&F2FS_I(inode)->i_mmap_sem);

	ret = f2fs_truncate_blocks(inode, i_size_read(inode), true, false);

	ret = f2fs_truncate_blocks(inode, i_size_read(inode), true);

	up_write(&F2FS_I(inode)->i_mmap_sem);

	if (ret)

		return ret;

fs/f2fs/inline.c

+2 −2
Original line number Diff line number Diff line
@@ -298,7 +298,7 @@ process_inline: 
		clear_inode_flag(inode, FI_INLINE_DATA);

		f2fs_put_page(ipage, 1);

	} else if (ri && (ri->i_inline & F2FS_INLINE_DATA)) {

		if (f2fs_truncate_blocks(inode, 0, false, false))

		if (f2fs_truncate_blocks(inode, 0, false))

			return false;

		goto process_inline;

	}
@@ -470,7 +470,7 @@ static int f2fs_add_inline_entries(struct inode *dir, void *inline_dentry) 
	return 0;

punch_dentry_pages:

	truncate_inode_pages(&dir->i_data, 0);

	f2fs_truncate_blocks(dir, 0, false, false);

	f2fs_truncate_blocks(dir, 0, false);

	f2fs_remove_dirty_inode(dir);

	return err;

}

CRA Git | Maintained and supported by SUSTech CRA and CCSE