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;
From:   Hans Schultz <schultz.hans@gmail.com>
To:     Ido Schimmel <idosch@idosch.org>,
        Hans Schultz <schultz.hans@gmail.com>
Cc:     davem@davemloft.net, kuba@kernel.org, netdev@vger.kernel.org,
        Andrew Lunn <andrew@lunn.ch>,
        Vivien Didelot <vivien.didelot@gmail.com>,
        Florian Fainelli <f.fainelli@gmail.com>,
        Vladimir Oltean <olteanv@gmail.com>,
        Jiri Pirko <jiri@resnulli.us>,
        Ivan Vecera <ivecera@redhat.com>,
        Roopa Prabhu <roopa@nvidia.com>,
        Nikolay Aleksandrov <razor@blackwall.org>,
        Daniel Borkmann <daniel@iogearbox.net>,
        Ido Schimmel <idosch@nvidia.com>, linux-kernel@vger.kernel.org,
        bridge@lists.linux-foundation.org
Subject: Re: [PATCH net-next 0/3] Extend locked port feature with FDB locked
 flag (MAC-Auth/MAB)
In-Reply-To: <Yi9kTh6XZu3OiCz0@shredder>
References: <20220310142320.611738-1-schultz.hans+netdev@gmail.com>
 <Yi9kTh6XZu3OiCz0@shredder>
Date:   Tue, 15 Mar 2022 09:59:49 +0100
Message-ID: <86ee33h9q2.fsf@gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On m=C3=A5n, mar 14, 2022 at 17:50, Ido Schimmel <idosch@idosch.org> wrote:
> On Thu, Mar 10, 2022 at 03:23:17PM +0100, Hans Schultz wrote:
>> This patch set extends the locked port feature for devices
>> that are behind a locked port, but do not have the ability to
>> authorize themselves as a supplicant using IEEE 802.1X.
>> Such devices can be printers, meters or anything related to
>> fixed installations. Instead of 802.1X authorization, devices
>> can get access based on their MAC addresses being whitelisted.
>>=20
>> For an authorization daemon to detect that a device is trying
>> to get access through a locked port, the bridge will add the
>> MAC address of the device to the FDB with a locked flag to it.
>> Thus the authorization daemon can catch the FDB add event and
>> check if the MAC address is in the whitelist and if so replace
>> the FDB entry without the locked flag enabled, and thus open
>> the port for the device.
>>=20
>> This feature is known as MAC-Auth or MAC Authentication Bypass
>> (MAB) in Cisco terminology, where the full MAB concept involves
>> additional Cisco infrastructure for authorization. There is no
>> real authentication process, as the MAC address of the device
>> is the only input the authorization daemon, in the general
>> case, has to base the decision if to unlock the port or not.
>>=20
>> With this patch set, an implementation of the offloaded case is
>> supplied for the mv88e6xxx driver. When a packet ingresses on
>> a locked port, an ATU miss violation event will occur. When
>
> When do you get an ATU miss violation? In case there is no FDB entry for
> the SA or also when there is an FDB entry, but it points to a different
> port? I see that the bridge will only create a "locked" FDB entry in
> case there is no existing entry, but it will not transition an existing
> entry to "locked" state. I guess ATU miss refers to an actual miss and
> not mismatch.
>

On a locked port, I get ATU miss violations when there is no FDB entry
for the SA, while if there is an entry but it is not assigned to the
port, then I get an ATU member violation (which I have now masked on
locked ports to limit unwanted interrupts).

So it seems to me that my 'ATU miss' corresponds to your MISS and my
'ATU member' corresponds to your MISMATCH. Since I inject an entry with
destination port vector (DPV) zero I get member violations after the
first miss violation.

> The HW I work with doesn't have the ability to generate such
> notifications, but it can trap packets on MISS (no entry) or MISMATCH
> (exists, but with different port). I believe that in order to support
> this feature we need to inject MISS-ed packets to the Rx path so that
> eventually the bridge itself will create the "locked" entry as opposed
> to notifying the bridge about the entry as in your case.
>

This seems to me to be the way forward in your case. What kind or family
of chips is your HW based on?

>> handling such ATU miss violation interrupts, the MAC address of
>> the device is added to the FDB with a zero destination port
>> vector (DPV) and the MAC address is communicated through the
>> switchdev layer to the bridge, so that a FDB entry with the
>> locked flag enabled can be added.
>>=20
>> Hans Schultz (3):
>>   net: bridge: add fdb flag to extent locked port feature
>>   net: switchdev: add support for offloading of fdb locked flag
>>   net: dsa: mv88e6xxx: mac-auth/MAB implementation
>
> Please extend tools/testing/selftests/net/forwarding/bridge_locked_port.sh
> with new test cases for this code.
>

Shall do.

>>=20
>>  drivers/net/dsa/mv88e6xxx/Makefile            |  1 +
>>  drivers/net/dsa/mv88e6xxx/chip.c              | 10 +--
>>  drivers/net/dsa/mv88e6xxx/chip.h              |  5 ++
>>  drivers/net/dsa/mv88e6xxx/global1.h           |  1 +
>>  drivers/net/dsa/mv88e6xxx/global1_atu.c       | 29 +++++++-
>>  .../net/dsa/mv88e6xxx/mv88e6xxx_switchdev.c   | 67 +++++++++++++++++++
>>  .../net/dsa/mv88e6xxx/mv88e6xxx_switchdev.h   | 20 ++++++
>>  drivers/net/dsa/mv88e6xxx/port.c              | 11 +++
>>  drivers/net/dsa/mv88e6xxx/port.h              |  1 +
>>  include/net/switchdev.h                       |  3 +-
>>  include/uapi/linux/neighbour.h                |  1 +
>>  net/bridge/br.c                               |  3 +-
>>  net/bridge/br_fdb.c                           | 13 +++-
>>  net/bridge/br_input.c                         | 11 ++-
>>  net/bridge/br_private.h                       |  5 +-
>>  15 files changed, 167 insertions(+), 14 deletions(-)
>>  create mode 100644 drivers/net/dsa/mv88e6xxx/mv88e6xxx_switchdev.c
>>  create mode 100644 drivers/net/dsa/mv88e6xxx/mv88e6xxx_switchdev.h
>>=20
>> --=20
>> 2.30.2
>>=20