[PATCH 2/2] ext4: Fix mmap data corruption when blocksize < pagesize

[Jan Kara]

On Tue 23-09-14 17:03:23, Jan Kara wrote:
> Use block_create_hole() when hole is being created in a file so that
> ->page_mkwrite() will get called for the partial tail page if it is
> mmaped (see the first patch in the series for details).
  Just out of curiosity I did a change similar to this one for ext4 to XFS
and indeed it fixed generic/030 test failures for XFS with blocksize 1k.

								Honza

PS: I forgot to CC xfs list in the original posting. You can find the VFS
    patch e.g. at http://www.spinics.net/lists/linux-mm/msg78976.html

> Signed-off-by: Jan Kara <jack at suse.cz>
> ---
>  fs/ext4/inode.c | 6 +++++-
>  1 file changed, 5 insertions(+), 1 deletion(-)
> 
> diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
> index 3aa26e9117c4..fdcb007c2c9e 100644
> --- a/fs/ext4/inode.c
> +++ b/fs/ext4/inode.c
> @@ -4536,8 +4536,12 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr)
>  				ext4_orphan_del(NULL, inode);
>  				goto err_out;
>  			}
> -		} else
> +		} else {
> +			loff_t old_size = inode->i_size;
> +
>  			i_size_write(inode, attr->ia_size);
> +			block_create_hole(inode, old_size, inode->i_size);
> +		}
>  
>  		/*
>  		 * Blocks are going to be removed from the inode. Wait
> -- 
> 1.8.1.4
> 
-- 
Jan Kara <jack at suse.cz>
SUSE Labs, CR