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: <20190306201413.14153-1-tycho@tycho.ws> <20190306201413.14153-2-tycho@tycho.ws>
 <CAGXu5jKaK-mTq=Qez7E_fYysz916_xEN7Houj-y8gdmNGkMMVw@mail.gmail.com>
In-Reply-To: <CAGXu5jKaK-mTq=Qez7E_fYysz916_xEN7Houj-y8gdmNGkMMVw@mail.gmail.com>
From:   Kees Cook <keescook@chromium.org>
Date:   Tue, 23 Apr 2019 16:31:45 -0700
Message-ID: <CAGXu5jKM1k0roNJEqtns2fe-yY_2G_qhvBHids-ZWimL-ddWsw@mail.gmail.com>
Subject: Re: [PATCH 2/2] seccomp: disallow NEW_LISTENER and TSYNC flags
To:     Tycho Andersen <tycho@tycho.ws>, James Morris <jmorris@namei.org>
Cc:     LKML <linux-kernel@vger.kernel.org>,
        "# 3.4.x" <stable@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Tue, Apr 23, 2019 at 3:09 PM Kees Cook <keescook@chromium.org> wrote:
>
> On Wed, Mar 6, 2019 at 12:14 PM Tycho Andersen <tycho@tycho.ws> wrote:
> >
> > As the comment notes, the return codes for TSYNC and NEW_LISTENER conflict,
> > because they both return positive values, one in the case of success and
> > one in the case of error. So, let's disallow both of these flags together.
> >
> > While this is technically a userspace break, all the users I know of are
> > still waiting on me to land this feature in libseccomp, so I think it'll be
> > safe. Also, at present my use case doesn't require TSYNC at all, so this
> > isn't a big deal to disallow. If someone wanted to support this, a path
> > forward would be to add a new flag like
> > TSYNC_AND_LISTENER_YES_I_UNDERSTAND_THAT_TSYNC_WILL_JUST_RETURN_EAGAIN, but
> > the use cases are so different I don't see it really happening.
> >
> > Finally, it's worth noting that this does actually fix a UAF issue: at the end
> > of seccomp_set_mode_filter(), we have:
> >
> >         if (flags & SECCOMP_FILTER_FLAG_NEW_LISTENER) {
> >                 if (ret < 0) {
> >                         listener_f->private_data = NULL;
> >                         fput(listener_f);
> >                         put_unused_fd(listener);
> >                 } else {
> >                         fd_install(listener, listener_f);
> >                         ret = listener;
> >                 }
> >         }
> > out_free:
> >         seccomp_filter_free(prepared);
> >
> > But if ret > 0 because TSYNC raced, we'll install the listener fd and then free
> > the filter out from underneath it, causing a UAF when the task closes it or
> > dies. This patch also switches the condition to be simply if (ret), so that
> > if someone does add the flag mentioned above, they won't have to remember
> > to fix this too.
> >
> > Signed-off-by: Tycho Andersen <tycho@tycho.ws>
> > Fixes: 6a21cc50f0c7 ("seccomp: add a return code to trap to userspace")
> > CC: stable@vger.kernel.org # v5.0+
>
> Thanks! Sorry I missed this. James, can you take this for Linus's
> fixes for v5.1? (Or should I send a pull request to you?)
>
> Acked-by: Kees Cook <keescook@chromium.org>
>
> Let's also add:
>
> Reported-by: syzbot+b562969adb2e04af3442@syzkaller.appspotmail.com
>
> > ---
> >  kernel/seccomp.c | 17 +++++++++++++++--
> >  1 file changed, 15 insertions(+), 2 deletions(-)
> >
> > diff --git a/kernel/seccomp.c b/kernel/seccomp.c
> > index d0d355ded2f4..79bada51091b 100644
> > --- a/kernel/seccomp.c
> > +++ b/kernel/seccomp.c
> > @@ -500,7 +500,10 @@ seccomp_prepare_user_filter(const char __user *user_filter)
> >   *
> >   * Caller must be holding current->sighand->siglock lock.
> >   *
> > - * Returns 0 on success, -ve on error.
> > + * Returns 0 on success, -ve on error, or
> > + *   - in TSYNC mode: the pid of a thread which was either not in the correct
> > + *     seccomp mode or did not have an ancestral seccomp filter
> > + *   - in NEW_LISTENER mode: the fd of the new listener
> >   */
> >  static long seccomp_attach_filter(unsigned int flags,
> >                                   struct seccomp_filter *filter)
> > @@ -1256,6 +1259,16 @@ static long seccomp_set_mode_filter(unsigned int flags,
> >         if (flags & ~SECCOMP_FILTER_FLAG_MASK)
> >                 return -EINVAL;
> >
> > +       /*
> > +        * In the successful case, NEW_LISTENER returns the new listener fd.
> > +        * But in the failure case, TSYNC returns the thread that died. If you
> > +        * combine these two flags, there's no way to tell whether something
> > +        * succeded or failed. So, let's disallow this combination.
>
> also a tiny typo: succeeded
>
> > +        */
> > +       if ((flags & SECCOMP_FILTER_FLAG_TSYNC) &&
> > +           (flags && SECCOMP_FILTER_FLAG_NEW_LISTENER))

also a typo: && should be &

> > +               return -EINVAL;
> > +
> >         /* Prepare the new filter before holding any locks. */
> >         prepared = seccomp_prepare_user_filter(filter);
> >         if (IS_ERR(prepared))
> > @@ -1302,7 +1315,7 @@ static long seccomp_set_mode_filter(unsigned int flags,
> >                 mutex_unlock(&current->signal->cred_guard_mutex);
> >  out_put_fd:
> >         if (flags & SECCOMP_FILTER_FLAG_NEW_LISTENER) {
> > -               if (ret < 0) {
> > +               if (ret) {
> >                         listener_f->private_data = NULL;
> >                         fput(listener_f);
> >                         put_unused_fd(listener);
> > --
> > 2.19.1
> >
>
> -Kees
>
> --
> Kees Cook


-- 
Kees Cook