diff mbox series

[v2] ext4: Fix i_disksize exceeding i_size problem in paritally written case

Message ID 20230321013721.89818-1-chengzhihao1@huawei.com
State New
Headers
Series [v2] ext4: Fix i_disksize exceeding i_size problem in paritally written case |

Commit Message

Zhihao Cheng March 21, 2023, 1:37 a.m. UTC 
  Following process makes i_disksize exceed i_size:

generic_perform_write
 copied = iov_iter_copy_from_user_atomic(len) // copied < len
 ext4_da_write_end
 | ext4_update_i_disksize
 |  new_i_size = pos + copied;
 |  WRITE_ONCE(EXT4_I(inode)->i_disksize, newsize) // update i_disksize
 | generic_write_end
 |  copied = block_write_end(copied, len) // copied = 0
 |   if (unlikely(copied < len))
 |    if (!PageUptodate(page))
 |     copied = 0;
 |  if (pos + copied > inode->i_size) // return false
 if (unlikely(copied == 0))
  goto again;
 if (unlikely(iov_iter_fault_in_readable(i, bytes))) {
  status = -EFAULT;
  break;
 }

We get i_disksize greater than i_size here, which could trigger WARNING
check 'i_size_read(inode) < EXT4_I(inode)->i_disksize' while doing dio:

ext4_dio_write_iter
 iomap_dio_rw
  __iomap_dio_rw // return err, length is not aligned to 512
 ext4_handle_inode_extension
  WARN_ON_ONCE(i_size_read(inode) < EXT4_I(inode)->i_disksize) // Oops

 WARNING: CPU: 2 PID: 2609 at fs/ext4/file.c:319
 CPU: 2 PID: 2609 Comm: aa Not tainted 6.3.0-rc2
 RIP: 0010:ext4_file_write_iter+0xbc7
 Call Trace:
  vfs_write+0x3b1
  ksys_write+0x77
  do_syscall_64+0x39

Fix it by updating 'copied' value before updating i_disksize just like
ext4_write_inline_data_end() does.

Fetch a reproducer in [Link].

Link: https://bugzilla.kernel.org/show_bug.cgi?id=217209
Fixes: 64769240bd07 ("ext4: Add delayed allocation support in data=writeback mode")
Signed-off-by: Zhihao Cheng <chengzhihao1@huawei.com>
---
 v1->v2: Check and update 'copied' value before updating i_disksize rather
	 than opencoding generic_write_end().
 fs/ext4/inode.c | 3 +++
 1 file changed, 3 insertions(+)

Comments

Jan Kara March 22, 2023, 9:18 a.m. UTC | #1 
On Tue 21-03-23 09:37:21, Zhihao Cheng wrote:
> Following process makes i_disksize exceed i_size:
> 
> generic_perform_write
>  copied = iov_iter_copy_from_user_atomic(len) // copied < len
>  ext4_da_write_end
>  | ext4_update_i_disksize
>  |  new_i_size = pos + copied;
>  |  WRITE_ONCE(EXT4_I(inode)->i_disksize, newsize) // update i_disksize
>  | generic_write_end
>  |  copied = block_write_end(copied, len) // copied = 0
>  |   if (unlikely(copied < len))
>  |    if (!PageUptodate(page))
>  |     copied = 0;
>  |  if (pos + copied > inode->i_size) // return false
>  if (unlikely(copied == 0))
>   goto again;
>  if (unlikely(iov_iter_fault_in_readable(i, bytes))) {
>   status = -EFAULT;
>   break;
>  }
> 
> We get i_disksize greater than i_size here, which could trigger WARNING
> check 'i_size_read(inode) < EXT4_I(inode)->i_disksize' while doing dio:
> 
> ext4_dio_write_iter
>  iomap_dio_rw
>   __iomap_dio_rw // return err, length is not aligned to 512
>  ext4_handle_inode_extension
>   WARN_ON_ONCE(i_size_read(inode) < EXT4_I(inode)->i_disksize) // Oops
> 
>  WARNING: CPU: 2 PID: 2609 at fs/ext4/file.c:319
>  CPU: 2 PID: 2609 Comm: aa Not tainted 6.3.0-rc2
>  RIP: 0010:ext4_file_write_iter+0xbc7
>  Call Trace:
>   vfs_write+0x3b1
>   ksys_write+0x77
>   do_syscall_64+0x39
> 
> Fix it by updating 'copied' value before updating i_disksize just like
> ext4_write_inline_data_end() does.
> 
> Fetch a reproducer in [Link].
> 
> Link: https://bugzilla.kernel.org/show_bug.cgi?id=217209
> Fixes: 64769240bd07 ("ext4: Add delayed allocation support in data=writeback mode")
> Signed-off-by: Zhihao Cheng <chengzhihao1@huawei.com>

Looks good to me. Feel free to add:

Reviewed-by: Jan Kara <jack@suse.cz>

								Honza

> ---
>  v1->v2: Check and update 'copied' value before updating i_disksize rather
> 	 than opencoding generic_write_end().
>  fs/ext4/inode.c | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
> index bf0b7dea4900..41ba1c432844 100644
> --- a/fs/ext4/inode.c
> +++ b/fs/ext4/inode.c
> @@ -3148,6 +3148,9 @@ static int ext4_da_write_end(struct file *file,
>  	    ext4_has_inline_data(inode))
>  		return ext4_write_inline_data_end(inode, pos, len, copied, page);
>  
> +	if (unlikely(copied < len) && !PageUptodate(page))
> +		copied = 0;
> +
>  	start = pos & (PAGE_SIZE - 1);
>  	end = start + copied - 1;
>  
> -- 
> 2.31.1
>
Theodore Ts'o April 30, 2023, 5:59 p.m. UTC | #2 
On Tue, 21 Mar 2023 09:37:21 +0800, Zhihao Cheng wrote:
> Following process makes i_disksize exceed i_size:
> 
> generic_perform_write
>  copied = iov_iter_copy_from_user_atomic(len) // copied < len
>  ext4_da_write_end
>  | ext4_update_i_disksize
>  |  new_i_size = pos + copied;
>  |  WRITE_ONCE(EXT4_I(inode)->i_disksize, newsize) // update i_disksize
>  | generic_write_end
>  |  copied = block_write_end(copied, len) // copied = 0
>  |   if (unlikely(copied < len))
>  |    if (!PageUptodate(page))
>  |     copied = 0;
>  |  if (pos + copied > inode->i_size) // return false
>  if (unlikely(copied == 0))
>   goto again;
>  if (unlikely(iov_iter_fault_in_readable(i, bytes))) {
>   status = -EFAULT;
>   break;
>  }
> 
> [...]

Applied, thanks!

[1/1] ext4: Fix i_disksize exceeding i_size problem in paritally written case
      commit: 1dedde690303c05ef732b7c5c8356fdf60a4ade3

Best regards,
diff mbox series

Patch

diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index bf0b7dea4900..41ba1c432844 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -3148,6 +3148,9 @@  static int ext4_da_write_end(struct file *file,
 	    ext4_has_inline_data(inode))
 		return ext4_write_inline_data_end(inode, pos, len, copied, page);
 
+	if (unlikely(copied < len) && !PageUptodate(page))
+		copied = 0;
+
 	start = pos & (PAGE_SIZE - 1);
 	end = start + copied - 1;