Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
MIME-Version: 1.0
References: <20230508163751.841-1-beaub@linux.microsoft.com>
 <CAADnVQLYL-ZaP_2vViaktw0G4UKkmpOK2q4ZXBa+f=M7cC25Rg@mail.gmail.com>
 <20230509130111.62d587f1@rorschach.local.home> <20230509163050.127d5123@rorschach.local.home>
 <20230515165707.hv65ekwp2djkjj5i@MacBook-Pro-8.local> <20230515192407.GA85@W11-BEAU-MD.localdomain>
 <20230517003628.aqqlvmzffj7fzzoj@MacBook-Pro-8.local> <20230516212658.2f5cc2c6@gandalf.local.home>
 <20230517165028.GA71@W11-BEAU-MD.localdomain> <CAADnVQK3-NBLSVRVsgArUEjqsuY2S_8mWsWmLEAtTzo+U49CKQ@mail.gmail.com>
 <20230518001916.GB254@W11-BEAU-MD.localdomain>
In-Reply-To: <20230518001916.GB254@W11-BEAU-MD.localdomain>
From:   Alexei Starovoitov <alexei.starovoitov@gmail.com>
Date:   Wed, 17 May 2023 17:56:34 -0700
Message-ID: <CAADnVQJwK3p1QyYEvAn9B86M4nkX69kuUvx2W0Yqwy0e=RSPPg@mail.gmail.com>
Subject: Re: [PATCH] tracing/user_events: Run BPF program if attached
To:     Beau Belgrave <beaub@linux.microsoft.com>
Cc:     Steven Rostedt <rostedt@goodmis.org>,
        Masami Hiramatsu <mhiramat@kernel.org>,
        LKML <linux-kernel@vger.kernel.org>,
        linux-trace-kernel@vger.kernel.org,
        Alexei Starovoitov <ast@kernel.org>,
        Daniel Borkmann <daniel@iogearbox.net>,
        Andrii Nakryiko <andrii@kernel.org>, bpf <bpf@vger.kernel.org>,
        David Vernet <void@manifault.com>,
        Linus Torvalds <torvalds@linux-foundation.org>,
        Dave Thaler <dthaler@microsoft.com>,
        Christian Brauner <brauner@kernel.org>,
        Christoph Hellwig <hch@infradead.org>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Wed, May 17, 2023 at 5:19=E2=80=AFPM Beau Belgrave <beaub@linux.microsof=
t.com> wrote:
>
> On Wed, May 17, 2023 at 05:10:47PM -0700, Alexei Starovoitov wrote:
> > On Wed, May 17, 2023 at 9:50=E2=80=AFAM Beau Belgrave <beaub@linux.micr=
osoft.com> wrote:
> > > >
> > > > >
> > > > > Looks like user events were designed with intention to be unprivi=
leged.
> > > > > When I looked at kernel/trace/trace_events_user.c I assumed root.
> > > > > I doubt other people reviewed it from security perspective.
> > > > >
> > > > > Recommending "chmod a+rw /sys/kernel/tracing/user_events_data" do=
esn't sound like a good idea.
> > > > >
> > > > > For example, I think the following is possible:
> > > > > fd =3D open("/sys/kernel/tracing/user_events_data")
> > > > > ioclt(fd, DIAG_IOCSDEL)
> > > > >   user_events_ioctl_del
> > > > >      delete_user_event(info->group, name);
> > > > >
> > > > > 'info' is different for every FD, but info->group is the same for=
 all users/processes/fds,
> > > > > because only one global init_group is created.
> > > > > So one user can unregister other user event by knowing 'name'.
> > > > > A security hole, no?
> >
> > ...
> >
> > > Regarding deleting events, only users that are given access can delet=
e
> > > events. They must know the event name, just like users with access to
> > > delete files must know a path (and have access to it). Since the
> > > write_index and other details are per-process, unless the user has
> > > access to either /sys/kernel/tracing/events/user_events/* or
> > > /sys/kernel/tracing/user_events_status, they do not know which names =
are
> > > being used.
> > >
> > > If that is not enough, we could require CAP_SYSADMIN to be able to
> > > delete events even when they have access to the file. Users can also
> > > apply SELinux policies per-file to achieve further isolation, if
> > > required.
> >
> > Whether /sys/kernel/tracing/user_events_status gets g+rw
> > or it gets a+rw (as your documentation recommends)
> > it is still a security issue.
> > The "event name" is trivial to find out by looking at the source code
> > of the target process or just "string target_binary".
>
> I guess, if they have access to the binary, etc.
> So they need both access to the binary and to the tracefs directory.
> We would not give them access like this in any normal setup other than a
> developer environment.
>
> > Restricting to cap_sysadmin is not the answer, since you want unpriv.
>
> We do not need unpriv to delete events, only to write and create events.
>
> We allow unregistering call-sites, which would still work unpriv with
> this requirement.
>
> > SElinux is not the answer either.
> > Since it's unpriv, different processes should not be able to mess with
> > user events of other processes.
>
> How is this different than uprobes if we give a user access to
> /sys/kernel/tracing/dynamic_events? Users can delete those as well. I
> don't see a difference here.

Because kprobe/uprobe are root only.
No sane person will do chmod a+rw /sys/kernel/tracing/uprobe_events.
It's just like chmod a+rw /etc/passwd

Whereas this is your recommended approach for user_events.

> In our production environments we are not giving out wide security to
> this file.

Fine by me. Keep it insecure and broken. Do not send bpf patches then.
I refuse to have bpf callable from such subsystems.
Somebody will inevitably blame bpf for the insecurity of user_events.