open_by_handle_at.2: Clarifications needed due to NFS reexport
commit896d2825b04b7e04c6f00ace78e1c41dc1f6b458
authorNeilBrown <neilb@suse.com>
Tue, 21 Nov 2017 21:15:13 +0000 (22 08:15 +1100)
committerMichael Kerrisk <mtk.manpages@gmail.com>
Thu, 23 Nov 2017 08:18:37 +0000 (23 09:18 +0100)
tree7e25d071c9e734c4e6016f1d2b989d72f924864e
parenta8d5f567ba2f111cf63d87ab30fa1c8788a6ca41
open_by_handle_at.2: Clarifications needed due to NFS reexport

The recent addition of NFS re-export and the possibility of using
name_to_handle_at() on an NFS filesystem raises issues with
name_to_handle_at() which have not been properly documented.

Getting the file handle for an untriggered automount point is
arguably meaningless and in certainly not supported by NFS.
name_to_handle_at() will return -EOVERFLOW even though the
requested "handle_bytes" is large enough.  This is an unfortunate
overloading of the error code, but is manageable.

So clarify this and also note that the mount_id is returned when
EOVERFLOW is reported.

Thought: it would be nice if mount_id were returned in the
EOPNOTSUPP case too.  I guess it is too late to fix that (?).

Link: https://github.com/systemd/systemd/issues/7082
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>
man2/open_by_handle_at.2