Re: [PATCH 16/16] vfs: return -EOPNOTSUPP for fscaps from vfs_*xattr()

On Thu, Nov 30, 2023 at 08:10:15AM +0200, Amir Goldstein wrote:

On Wed, Nov 29, 2023 at 11:51 PM Seth Forshee (DigitalOcean)
[off-list ref] wrote:

quoted

Now that the new vfs-level interfaces are fully supported and all code
has been converted to use them, stop permitting use of the top-level vfs
xattr interfaces for capabilities xattrs. Unlike with ACLs we still need
to be able to work with fscaps xattrs using lower-level interfaces in a
handful of places, so only use of the top-level xattr interfaces is
restricted.

Can you explain why?
Is there an inherent difference between ACLs and fscaps in that respect
or is it just a matter of more work that needs to be done?

There are a number of differences. ACLs have caching, require additional
permission checks, and require a lot of filesystem-specific handling.
fscaps are simpler by comparison, and most filesystems can rely on a
common implementation that just converts to/from raw disk xattrs.

So at minimum I think the lowest level interfaces,
__vfs_{get,set,remove}xattr(), need to continue to allow fscaps, and
that's where ACL xattrs are blocked. Allowing some of the others to
still work with them is a matter of convenience (e.g. using
vfs_getxattr_alloc()) and trying to reduce code duplication. But as you
pointed out I did miss at least duplicating fsnotify_xattr(), so I'm
going to have another look at how I implemented these.

quoted

Signed-off-by: Seth Forshee (DigitalOcean) <sforshee@kernel.org>
---
 fs/xattr.c | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/fs/xattr.c b/fs/xattr.c
index 372644b15457..4b779779ad8c 100644
--- a/fs/xattr.c
+++ b/fs/xattr.c

@@ -540,6 +540,9 @@ vfs_setxattr(struct mnt_idmap *idmap, struct dentry *dentry,
        const void  *orig_value = value;
        int error;

+       if (!strcmp(name, XATTR_NAME_CAPS))
+               return -EOPNOTSUPP;
+

It this is really not expected, then it should be an assert and
please use an inline helper like is_posix_acl_xattr():

if (WARN_ON_ONCE(is_fscaps_xattr(name)))

Ack, makes sense.