[PATCH] ext4: fix off-by-one error when writing back pages before dio read

[PATCH] ext4: fix off-by-one error when writing back pages before dio read

[Eric Biggers]

From: Eric Biggers <ebiggers@google.com>

The 'lend' argument of filemap_write_and_wait_range() is inclusive, so
we need to subtract 1 from pos + count.

Note that 'count' is guaranteed to be nonzero since
ext4_file_read_iter() returns early when given a 0 count.

Fixes: 16c54688592c ("ext4: Allow parallel DIO reads")
Signed-off-by: Eric Biggers <ebiggers@google.com>
---
 fs/ext4/inode.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 7c6e715b4d2e..3080ea82dd6d 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -3710,7 +3710,7 @@ static ssize_t ext4_direct_IO_read(struct kiocb *iocb, struct iov_iter *iter)
 	 */
 	inode_lock_shared(inode);
 	ret = filemap_write_and_wait_range(mapping, iocb->ki_pos,
-					   iocb->ki_pos + count);
+					   iocb->ki_pos + count - 1);
 	if (ret)
 		goto out_unlock;
 	ret = __blockdev_direct_IO(iocb, inode, inode->i_sb->s_bdev,
-- 
2.13.0.219.gdb65acc882-goog

Re: [PATCH] ext4: fix off-by-one error when writing back pages before dio read

[Jan Kara]

On Mon 22-05-17 18:34:41, Eric Biggers wrote:
> From: Eric Biggers <ebiggers@google.com>
> 
> The 'lend' argument of filemap_write_and_wait_range() is inclusive, so
> we need to subtract 1 from pos + count.
> 
> Note that 'count' is guaranteed to be nonzero since
> ext4_file_read_iter() returns early when given a 0 count.
> 
> Fixes: 16c54688592c ("ext4: Allow parallel DIO reads")
> Signed-off-by: Eric Biggers <ebiggers@google.com>

Yeah. You can add:

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

								Honza

> ---
>  fs/ext4/inode.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
> index 7c6e715b4d2e..3080ea82dd6d 100644
> --- a/fs/ext4/inode.c
> +++ b/fs/ext4/inode.c
> @@ -3710,7 +3710,7 @@ static ssize_t ext4_direct_IO_read(struct kiocb *iocb, struct iov_iter *iter)
>  	 */
>  	inode_lock_shared(inode);
>  	ret = filemap_write_and_wait_range(mapping, iocb->ki_pos,
> -					   iocb->ki_pos + count);
> +					   iocb->ki_pos + count - 1);
>  	if (ret)
>  		goto out_unlock;
>  	ret = __blockdev_direct_IO(iocb, inode, inode->i_sb->s_bdev,
> -- 
> 2.13.0.219.gdb65acc882-goog
> 
-- 
Jan Kara <jack@suse.com>
SUSE Labs, CR

Re: [PATCH] ext4: fix off-by-one error when writing back pages before dio read

[Theodore Ts'o]

On Tue, May 23, 2017 at 09:53:05AM +0200, Jan Kara wrote:
> On Mon 22-05-17 18:34:41, Eric Biggers wrote:
> > From: Eric Biggers <ebiggers@google.com>
> > 
> > The 'lend' argument of filemap_write_and_wait_range() is inclusive, so
> > we need to subtract 1 from pos + count.
> > 
> > Note that 'count' is guaranteed to be nonzero since
> > ext4_file_read_iter() returns early when given a 0 count.
> > 
> > Fixes: 16c54688592c ("ext4: Allow parallel DIO reads")
> > Signed-off-by: Eric Biggers <ebiggers@google.com>
> 
> Yeah. You can add:
> 
> Reviewed-by: Jan Kara <jack@suse.cz>

Thanks, applied.

					- Ted