Re: [PATCH v2] ext4: fix race condition between ext4_write and ext4_convert_inline_data

[Jan Kara <jack@suse.cz>]

On Tue 26-04-22 22:06:58, Baokun Li wrote:
> Hulk Robot reported a BUG_ON:
>  ==================================================================
>  EXT4-fs error (device loop3): ext4_mb_generate_buddy:805: group 0,
>  block bitmap and bg descriptor inconsistent: 25 vs 31513 free clusters
>  kernel BUG at fs/ext4/ext4_jbd2.c:53!
>  invalid opcode: 0000 [#1] SMP KASAN PTI
>  CPU: 0 PID: 25371 Comm: syz-executor.3 Not tainted 5.10.0+ #1
>  RIP: 0010:ext4_put_nojournal fs/ext4/ext4_jbd2.c:53 [inline]
>  RIP: 0010:__ext4_journal_stop+0x10e/0x110 fs/ext4/ext4_jbd2.c:116
>  [...]
>  Call Trace:
>   ext4_write_inline_data_end+0x59a/0x730 fs/ext4/inline.c:795
>   generic_perform_write+0x279/0x3c0 mm/filemap.c:3344
>   ext4_buffered_write_iter+0x2e3/0x3d0 fs/ext4/file.c:270
>   ext4_file_write_iter+0x30a/0x11c0 fs/ext4/file.c:520
>   do_iter_readv_writev+0x339/0x3c0 fs/read_write.c:732
>   do_iter_write+0x107/0x430 fs/read_write.c:861
>   vfs_writev fs/read_write.c:934 [inline]
>   do_pwritev+0x1e5/0x380 fs/read_write.c:1031
>  [...]
>  ==================================================================
> 
> Above issue may happen as follows:
>            cpu1                     cpu2
> __________________________|__________________________
> do_pwritev
>  vfs_writev
>   do_iter_write
>    ext4_file_write_iter
>     ext4_buffered_write_iter
>      generic_perform_write
>       ext4_da_write_begin
>                            vfs_fallocate
>                             ext4_fallocate
>                              ext4_convert_inline_data
>                               ext4_convert_inline_data_nolock
>                                ext4_destroy_inline_data_nolock
>                                 clear EXT4_STATE_MAY_INLINE_DATA
>                                ext4_map_blocks
>                                 ext4_ext_map_blocks
>                                  ext4_mb_new_blocks
>                                   ext4_mb_regular_allocator
>                                    ext4_mb_good_group_nolock
>                                     ext4_mb_init_group
>                                      ext4_mb_init_cache
>                                       ext4_mb_generate_buddy  --> error
>        ext4_test_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA)
>                                 ext4_restore_inline_data
>                                  set EXT4_STATE_MAY_INLINE_DATA
>        ext4_block_write_begin
>       ext4_da_write_end
>        ext4_test_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA)
>        ext4_write_inline_data_end
>         handle=NULL
>         ext4_journal_stop(handle)
>          __ext4_journal_stop
>           ext4_put_nojournal(handle)
>            ref_cnt = (unsigned long)handle
>            BUG_ON(ref_cnt == 0)  ---> BUG_ON
> 
> The lock held by ext4_convert_inline_data is xattr_sem, but the lock
> held by generic_perform_write is i_rwsem. Therefore, the two locks can
> be concurrent. To solve above issue, we just add inode_lock in
> ext4_convert_inline_data.
> 
> Fixes: 0c8d414f163f ("ext4: let fallocate handle inline data correctly")
> Cc: stable@vger.kernel.org
> Reported-by: Hulk Robot <hulkci@huawei.com>
> Signed-off-by: Baokun Li <libaokun1@huawei.com>

Thanks for the report and the patch. I agree inode_lock is needed to
protect the conversion.  However you cannot lock it inside
ext4_convert_inline_data() because it is often called with inode_lock
already held (in fact in most of the other cases as far as I have checked).
So the right fix is to reorganize ext4_fallocate() a bit to do something
like:

	ext4_clear_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA);
	if (ext4_has_inline_data(inode)) {
		inode_lock(inode);
		ret = ext4_convert_inline_data(inode);
		inode_unlock(inode);
		if (ret)
			goto exit;
	}

and remove similar handling from ext4_punch_hole() (and make sure it uses
the new handling in ext4_fallocate()).

								Honza
-- 
Jan Kara <jack@suse.com>
SUSE Labs, CR