Re: [PATCH 2/2] fs: return EOPNOTSUPP from file_setattr/file_getattr syscalls

["Darrick J. Wong" <djwong@kernel.org>]

On Wed, Oct 08, 2025 at 02:44:18PM +0200, Andrey Albershteyn wrote:
> These syscalls call to vfs_fileattr_get/set functions which return
> ENOIOCTLCMD if filesystem doesn't support setting file attribute on an
> inode. For syscalls EOPNOTSUPP would be more appropriate return error.
> 
> Signed-off-by: Andrey Albershteyn <aalbersh@kernel.org>
> ---
>  fs/file_attr.c | 4 ++++
>  1 file changed, 4 insertions(+)
> 
> diff --git a/fs/file_attr.c b/fs/file_attr.c
> index 460b2dd21a85..5e3e2aba97b5 100644
> --- a/fs/file_attr.c
> +++ b/fs/file_attr.c
> @@ -416,6 +416,8 @@ SYSCALL_DEFINE5(file_getattr, int, dfd, const char __user *, filename,
>  	}
>  
>  	error = vfs_fileattr_get(filepath.dentry, &fa);
> +	if (error == -ENOIOCTLCMD)

Hrm.  Back in 6.17, XFS would return ENOTTY if you called ->fileattr_get
on a special file:

int
xfs_fileattr_get(
	struct dentry		*dentry,
	struct file_kattr	*fa)
{
	struct xfs_inode	*ip = XFS_I(d_inode(dentry));

	if (d_is_special(dentry))
		return -ENOTTY;
	...
}

Given that there are other fileattr_[gs]et implementations out there
that might return ENOTTY (e.g. fuse servers and other externally
maintained filesystems), I think both syscall functions need to check
for that as well:

	if (error == -ENOIOCTLCMD || error == -ENOTTY)
		return -EOPNOTSUPP;

--D

> +		error = -EOPNOTSUPP;
>  	if (error)
>  		return error;
>  
> @@ -483,6 +485,8 @@ SYSCALL_DEFINE5(file_setattr, int, dfd, const char __user *, filename,
>  	if (!error) {
>  		error = vfs_fileattr_set(mnt_idmap(filepath.mnt),
>  					 filepath.dentry, &fa);
> +		if (error == -ENOIOCTLCMD)
> +			error = -EOPNOTSUPP;
>  		mnt_drop_write(filepath.mnt);
>  	}
>  
> 
> -- 
> 2.51.0
> 
>