On Fri, Jun 21, 2013 at 02:45:53PM -0300, Carlos Maiolino wrote:
> XFS removes sgid bits of subdirectories under a directory containing a default
> acl.
>
> When a default acl is set, it implies xfs to call xfs_setattr_nonsize() in its
> code path. Such function is shared among mkdir and chmod system calls, and
> does some checks unneeded by mkdir (calling inode_change_ok()). Such checks
> remove sgid bit from the inode after it has been granted.
>
> With this patch, we extend the meaning of XFS_ATTR_NOACL flag to avoid these
> checks when acls are being inherited (thanks hch).
>
> Also, xfs_setattr_mode, doesn't need to re-check for group id and capabilities
> permissions, this only implies in another try to remove sgid bit from the
> directories. Such check is already done either on inode_change_ok() or
> xfs_setattr_nonsize().
>
> Changelog:
>
> V2: Extends the meaning of XFS_ATTR_NOACL instead of wrap the tests into
> another
> function
>
> V3: Remove S_ISDIR check in xfs_setattr_nonsize() from the patch
>
> Signed-off-by: Carlos Maiolino <cmaiolino@xxxxxxxxxx>
> ---
> fs/xfs/xfs_iops.c | 20 ++++++++++----------
> 1 file changed, 10 insertions(+), 10 deletions(-)
>
> diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c
> index ca9ecaa..2e5aca8 100644
> --- a/fs/xfs/xfs_iops.c
> +++ b/fs/xfs/xfs_iops.c
> @@ -467,9 +467,6 @@ xfs_setattr_mode(
> ASSERT(tp);
> ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL));
>
> - if (!in_group_p(inode->i_gid) && !capable(CAP_FSETID))
> - mode &= ~S_ISGID;
> -
inode_change_ok has had the check for whether to clear S_ISGID since the
initial import of Linus's tree, and it is called when ATTR_MODE is set there,
just as in xfs_setattr_nonsize, and xfs_setattr_size. That aspect of this
looks ok to me.
> ip->i_d.di_mode &= S_IFMT;
> ip->i_d.di_mode |= mode & ~S_IFMT;
>
> @@ -495,15 +492,18 @@ xfs_setattr_nonsize(
>
> trace_xfs_setattr(ip);
>
> - if (mp->m_flags & XFS_MOUNT_RDONLY)
> - return XFS_ERROR(EROFS);
> + /* If acls are being inherited, we already have this checked */
> + if (!(flags & XFS_ATTR_NOACL)) {
> + if (mp->m_flags & XFS_MOUNT_RDONLY)
> + return XFS_ERROR(EROFS);
>
> - if (XFS_FORCED_SHUTDOWN(mp))
> - return XFS_ERROR(EIO);
> + if (XFS_FORCED_SHUTDOWN(mp))
> + return XFS_ERROR(EIO);
>
> - error = -inode_change_ok(inode, iattr);
> - if (error)
> - return XFS_ERROR(error);
> + error = -inode_change_ok(inode, iattr);
> + if (error)
> + return XFS_ERROR(error);
> + }
I'm not so sure about this change yet. Looks like the two relevant callers are:
.set - xattr_handler
xfs_xattr_acl_set
xfs_set_mode
xfs_setattr_nonsize(..., XFS_ATTR_NOACL);
and
xfs_vn_mknod
xfs_inherit_acl
xfs_set_mode
xfs_setattr_nonsize(..., XFS_ATTR_NOACL);
I suggest moving the forced shutdown and readonly checks outside of the
XFS_ATTR_NOACL conditional. I'm not seeing those checks in xfs_attr_acl_set or
xfs_vn_mknod and it won't hurt to be careful.
It also seems like inode_change_ok might have some other checks that are
necessary to determine whether it is ok to update the mode and ctime here. A
call to inode_owner_or_capable as is done in inode_change_ok would cover this
possibility.
Other than those two suggestions this looks pretty good to me.
Regards,
Ben
|