Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
MIME-Version: 1.0
References: <20181009103752.21482-1-laurent@vivier.eu> <20181009103752.21482-2-laurent@vivier.eu>
 <f14ce318-5a1a-06f3-2e90-91220f39e74b@virtuozzo.com> <bc11ca64-8616-f06a-c1fa-c76511e60257@vivier.eu>
 <CAG48ez0pAxC+yetPfM+XyS8kfqA4hw1hxjKG55gbF-4qcNWUiA@mail.gmail.com> <795fea1a-97bd-87f5-6a14-6d47a4e42f74@vivier.eu>
In-Reply-To: <795fea1a-97bd-87f5-6a14-6d47a4e42f74@vivier.eu>
From:   Jann Horn <jannh@google.com>
Date:   Tue, 9 Oct 2018 19:01:56 +0200
Message-ID: <CAG48ez2HHv5VuW3GXk0ExxLqW-qzq8yuwSQ=8Z=-h7iNH0gizg@mail.gmail.com>
Subject: Re: [RFC v5 1/1] ns: add binfmt_misc to the user namespace
To:     Laurent Vivier <laurent@vivier.eu>
Cc:     ktkhai@virtuozzo.com, kernel list <linux-kernel@vger.kernel.org>,
        "Eric W. Biederman" <ebiederm@xmission.com>, dima@arista.com,
        Linux API <linux-api@vger.kernel.org>,
        James Bottomley <James.Bottomley@hansenpartnership.com>,
        Al Viro <viro@zeniv.linux.org.uk>,
        linux-fsdevel@vger.kernel.org, Andrei Vagin <avagin@gmail.com>,
        containers@lists.linux-foundation.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Tue, Oct 9, 2018 at 6:57 PM Laurent Vivier <laurent@vivier.eu> wrote:
> Le 09/10/2018 =C3=A0 18:53, Jann Horn a =C3=A9crit :
> > On Tue, Oct 9, 2018 at 6:45 PM Laurent Vivier <laurent@vivier.eu> wrote=
:
> >> Le 09/10/2018 =C3=A0 18:15, Kirill Tkhai a =C3=A9crit :
> >>> On 09.10.2018 13:37, Laurent Vivier wrote:
> >>>> This patch allows to have a different binfmt_misc configuration
> >>>> for each new user namespace. By default, the binfmt_misc configurati=
on
> >>>> is the one of the previous level, but if the binfmt_misc filesystem =
is
> >>>> mounted in the new namespace a new empty binfmt instance is created =
and
> >>>> used in this namespace.
> >>>>
> >>>> For instance, using "unshare" we can start a chroot of an another
> >>>> architecture and configure the binfmt_misc interpreter without being=
 root
> >>>> to run the binaries in this chroot.
> >>>>
> >>>> Signed-off-by: Laurent Vivier <laurent@vivier.eu>
> >>>> ---
> >>>>  fs/binfmt_misc.c               | 106 ++++++++++++++++++++++++------=
---
> >>>>  include/linux/user_namespace.h |  13 ++++
> >>>>  kernel/user.c                  |  13 ++++
> >>>>  kernel/user_namespace.c        |   3 +
> >>>>  4 files changed, 107 insertions(+), 28 deletions(-)
> >>>>
> >>>> diff --git a/fs/binfmt_misc.c b/fs/binfmt_misc.c
> >>>> index aa4a7a23ff99..1e0029d097d9 100644
> >>>> --- a/fs/binfmt_misc.c
> >>>> +++ b/fs/binfmt_misc.c
> >> ...
> >>>> @@ -80,18 +74,32 @@ static int entry_count;
> >>>>   */
> >>>>  #define MAX_REGISTER_LENGTH 1920
> >>>>
> >>>> +static struct binfmt_namespace *binfmt_ns(struct user_namespace *ns=
)
> >>>> +{
> >>>> +    struct binfmt_namespace *b_ns;
> >>>> +
> >>>> +    while (ns) {
> >>>> +            b_ns =3D READ_ONCE(ns->binfmt_ns);
> >>>> +            if (b_ns)
> >>>> +                    return b_ns;
> >>>> +            ns =3D ns->parent;
> >>>> +    }
> >>>> +    WARN_ON_ONCE(1);
> >>>> +    return NULL;
> >>>> +}
> >>>> +
> >> ...
> >>>> @@ -823,12 +847,34 @@ static const struct super_operations s_ops =3D=
 {
> >>>>  static int bm_fill_super(struct super_block *sb, void *data, int si=
lent)
> >>>>  {
> >>>>      int err;
> >>>> +    struct user_namespace *ns =3D sb->s_user_ns;
> >>>>      static const struct tree_descr bm_files[] =3D {
> >>>>              [2] =3D {"status", &bm_status_operations, S_IWUSR|S_IRU=
GO},
> >>>>              [3] =3D {"register", &bm_register_operations, S_IWUSR},
> >>>>              /* last one */ {""}
> >>>>      };
> >>>>
> >>>> +    /* create a new binfmt namespace
> >>>> +     * if we are not in the first user namespace
> >>>> +     * but the binfmt namespace is the first one
> >>>> +     */
> >>>> +    if (READ_ONCE(ns->binfmt_ns) =3D=3D NULL) {
> >>>> +            struct binfmt_namespace *new_ns;
> >>>> +
> >>>> +            new_ns =3D kmalloc(sizeof(struct binfmt_namespace),
> >>>> +                             GFP_KERNEL);
> >>>> +            if (new_ns =3D=3D NULL)
> >>>> +                    return -ENOMEM;
> >>>> +            INIT_LIST_HEAD(&new_ns->entries);
> >>>> +            new_ns->enabled =3D 1;
> >>>> +            rwlock_init(&new_ns->entries_lock);
> >>>> +            new_ns->bm_mnt =3D NULL;
> >>>> +            new_ns->entry_count =3D 0;
> >>>> +            /* ensure new_ns is completely initialized before shari=
ng it */
> >>>> +            smp_wmb();
> >>>
> >>> (I haven't dived into patch logic, here just small barrier remark fro=
m quick sight).
> >>> smp_wmb() has no sense without paired smp_rmb() on the read side. Pos=
sible,
> >>> you want something like below in read hunk:
> >>>
> >>> +             b_ns =3D READ_ONCE(ns->binfmt_ns);
> >>> +             if (b_ns) {
> >>> +                     smp_rmb();
> >>> +                     return b_ns;
> >>> +             }
> >>>
> >>>
> >>
> >> The write barrier is here to ensure the structure is fully written
> >> before we set the pointer.
> >>
> >> I don't understand how read barrier can change something at this level=
,
> >> IMHO the couple WRITE_ONCE()/READ_ONCE() should be enough to ensure we
> >> have correctly initialized the pointer and the structure when we read
> >> the pointer back.
> >>
> >> I think the pointer itself is the "barrier" to access the memory
> >> modified before.
> >
> > Things don't work that way on alpha, but that's why READ_ONCE()
> > includes an smp_read_barrier_depends():
> >
> > #define __READ_ONCE(x, check)                                          =
 \
> > ({                                                                     =
 \
> >         union { typeof(x) __val; char __c[1]; } __u;                   =
 \
> >         if (check)                                                     =
 \
> >                 __read_once_size(&(x), __u.__c, sizeof(x));            =
 \
> >         else                                                           =
 \
> >                 __read_once_size_nocheck(&(x), __u.__c, sizeof(x));    =
 \
> >         smp_read_barrier_depends(); /* Enforce dependency ordering from=
 x */ \
> >         __u.__val;                                                     =
 \
> > })
> > #define READ_ONCE(x) __READ_ONCE(x, 1)
> >
>
> So my questions are:
>
> - do we need a smp_wmb() barrier if we use READ_ONCE() and WRITE_ONCE()?

Yes.

> - if we need an smp_wmb() barrier, do we need an smp_rmb() barrier as
> the data we want to "protect" are behind an access to the pointer?

No. You only need an smp_read_barrier_depends() to access things
behind the pointer you're reading (documented in a big comment block
in arch/alpha/include/asm/barrier.h); and that barrier is implied by
READ_ONCE(), so you don't have to write it yourself.