Skip site navigation (1)Skip section navigation (2)
Date:      Sat, 12 Jun 2010 17:15:49 +0300
From:      Kostik Belousov <kostikbel@gmail.com>
To:        Mikolaj Golub <to.my.trociny@gmail.com>
Cc:        Leon Me??ner <l.messner@physik.tu-berlin.de>, freebsd-stable@freebsd.org
Subject:   Re: freeBSD nullfs together with nfs and "silly rename"
Message-ID:  <20100612141549.GM13238@deviant.kiev.zoral.com.ua>
In-Reply-To: <8639wsk4t1.fsf@kopusha.home.net>
References:  <20100606144443.GA50876@emmi.physik-pool.tu-berlin.de> <8639wsk4t1.fsf@kopusha.home.net>

next in thread | previous in thread | raw e-mail | index | archive | help

--Cqq5NadOW2RfLMJ/
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Sat, Jun 12, 2010 at 11:56:10AM +0300, Mikolaj Golub wrote:
>=20
> On Sun, 6 Jun 2010 16:44:43 +0200 Leon Me??ner wrote:
>=20
>  LM> Hi,
>  LM> I hope this is not the wrong list to ask. Didn't get any answers on
>  LM> -questions.
>=20
>  LM> When you try to do the following inside a nullfs mounted directory,
>  LM> where the nullfs origin is itself mounted via nfs you get an error:
>=20
>  LM> # foo=20
>  LM> # tail -f foo&=20
>  LM> # rm -f foo=20
>  LM> tail: foo: Stale NFS file handle
>  LM> # fg
>=20
>  LM> This is really a problem when running services inside jails and using
>  LM> NFS as storage. As of [2] it looks like this problem is known for a
>  LM> while. On a normal NFS mount this does not happen as "silly renaming"
>  LM> [1] works there (producing nasty little .nfsXXXX files).
>=20
> nfs_sillyrename() is called when vnode's usecount is more then 1. It is
> expected that unlink() syscall increases vnode's usecount in namei() and =
if
> the file has been already opened usecount will be more then 1.
>=20
> But with nullfs layer present the reference counts are held by the upper =
node,
> not the lower (nfs) one, so when unlink() is called it increases usecount=
 of
> the upper vnode, not nfs vnode and nfs_sillyrename() is never called.
>=20
> The strightforward solution looks like to implement null_remove() that wi=
ll
> increase lower vnode's refcount before calling null_bypass() and then
> decrement it after the call. See the attached patch (it works for me on b=
oth
> 8-STABLE and CURRENT).

The upper vnode holds a reference to the lower vnode, as you noted.
Now, with your patch, I believe that _all_ calls to the nfs_remove()
are happen with refcount > 1.

--Cqq5NadOW2RfLMJ/
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (FreeBSD)

iEYEARECAAYFAkwTlpUACgkQC3+MBN1Mb4gkRwCg4QvbVcDxkzdEgeMmtKUpgDP4
F2AAnj+DyLLiPVvIe+7/bjdyKb87oy1f
=Jmsd
-----END PGP SIGNATURE-----

--Cqq5NadOW2RfLMJ/--



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20100612141549.GM13238>