On Thu, 2019-07-25 at 11:54 +0200, David Disseldorp wrote: > Hi, > > On Thu, 25 Jul 2019 10:37:40 +0100, Luis Henriques wrote: > > > Jeff Layton <jlayton@xxxxxxxxxx> writes: > > > > (CC'ing David) > > > > > Most filesystems that provide virtual xattrs (e.g. CIFS) don't display > > > them via listxattr(). Ceph does, and that causes some of the tests in > > > xfstests to fail. > > > > > > Have cephfs stop listing vxattrs in listxattr. Userspace can always > > > query them directly when the name is known. > > > > I don't see a problem with this, unless we already have applications > > relying on this behaviour. The first thing that came to my mind was > > samba, but I guess David can probably confirm whether this is true or > > not. > > > > If we're unable to stop listing there xattrs, the other option for > > fixing the xfstests that _may_ be acceptable by the maintainers is to > > use an output filter (lots of examples in common/filter). Yeah, I rolled a half-assed xfstests patch that did this, and HCH gave it a NAK. He's probably right though, and fixing it in ceph.ko is a better approach I think. > > Samba currently only makes use of the ceph.snap.btime vxattr. It doesn't > depend on it appearing in the listxattr(), so removal would be fine. Not > sure about other applications though. > > Cheers, David Ok, thanks guys. I'll go ahead and push this into the ceph/testing branch and see if teuthology complains. Thanks, -- Jeff Layton <jlayton@xxxxxxxxxx>
