Re: [PATCH v6] af_packet: Handle outgoing VLAN packets without hardware offloading

[Willem de Bruijn <willemdebruijn.kernel@gmail.com>]

Chengen Du wrote:
> Hi Willem,
> 
> On Tue, Jun 11, 2024 at 7:18 AM Willem de Bruijn
> <willemdebruijn.kernel@gmail.com> wrote:
> >
> > Chengen Du wrote:
> > > Hi Willem,
> > >
> > > I'm sorry, but I would like to confirm the issue further.
> > >
> > > On Mon, Jun 10, 2024 at 4:19 AM Willem de Bruijn
> > > <willemdebruijn.kernel@gmail.com> wrote:
> > > >
> > > > Chengen Du wrote:
> > > > > The issue initially stems from libpcap. The ethertype will be overwritten
> > > > > as the VLAN TPID if the network interface lacks hardware VLAN offloading.
> > > > > In the outbound packet path, if hardware VLAN offloading is unavailable,
> > > > > the VLAN tag is inserted into the payload but then cleared from the sk_buff
> > > > > struct. Consequently, this can lead to a false negative when checking for
> > > > > the presence of a VLAN tag, causing the packet sniffing outcome to lack
> > > > > VLAN tag information (i.e., TCI-TPID). As a result, the packet capturing
> > > > > tool may be unable to parse packets as expected.
> > > > >
> > > > > The TCI-TPID is missing because the prb_fill_vlan_info() function does not
> > > > > modify the tp_vlan_tci/tp_vlan_tpid values, as the information is in the
> > > > > payload and not in the sk_buff struct. The skb_vlan_tag_present() function
> > > > > only checks vlan_all in the sk_buff struct. In cooked mode, the L2 header
> > > > > is stripped, preventing the packet capturing tool from determining the
> > > > > correct TCI-TPID value. Additionally, the protocol in SLL is incorrect,
> > > > > which means the packet capturing tool cannot parse the L3 header correctly.
> > > > >
> > > > > Link: https://github.com/the-tcpdump-group/libpcap/issues/1105
> > > > > Link: https://lore.kernel.org/netdev/20240520070348.26725-1-chengen.du@canonical.com/T/#u
> > > > > Fixes: 393e52e33c6c ("packet: deliver VLAN TCI to userspace")
> > > > > Cc: stable@vger.kernel.org
> > > > > Signed-off-by: Chengen Du <chengen.du@canonical.com>
> > > >
> > > > Overall, solid.
> > > >
> > > > > ---
> > > > >  net/packet/af_packet.c | 57 ++++++++++++++++++++++++++++++++++++++++--
> > > > >  1 file changed, 55 insertions(+), 2 deletions(-)
> > > > >
> > > > > diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c
> > > > > index ea3ebc160e25..8cffbe1f912d 100644
> > > > > --- a/net/packet/af_packet.c
> > > > > +++ b/net/packet/af_packet.c
> > > > > @@ -538,6 +538,43 @@ static void *packet_current_frame(struct packet_sock *po,
> > > > >       return packet_lookup_frame(po, rb, rb->head, status);
> > > > >  }
> > > > >
> > > > > +static u16 vlan_get_tci(struct sk_buff *skb)
> > > > > +{
> > > > > +     struct vlan_hdr vhdr, *vh;
> > > > > +     u8 *skb_orig_data = skb->data;
> > > > > +     int skb_orig_len = skb->len;
> > > > > +
> > > > > +     skb_push(skb, skb->data - skb_mac_header(skb));
> > > > > +     vh = skb_header_pointer(skb, ETH_HLEN, sizeof(vhdr), &vhdr);
> > > >
> > > > Don't harcode Ethernet.
> > > >
> > > > According to documentation VLANs are used with other link layers.
> > > >
> > > > More importantly, in practice PF_PACKET allows inserting this
> > > > skb->protocol on any device.
> > > >
> > > > We don't use link layer specific constants anywhere in the packet
> > > > socket code for this reason. But instead dev->hard_header_len.
> > > >
> > > > One caveat there is variable length link layer headers, where
> > > > dev->min_header_len != dev->hard_header_len. Will just have to fail
> > > > on those.
> > >
> > > Thank you for pointing out this error. I would like to confirm if I
> > > need to use dev->hard_header_len to get the correct header length and
> > > return zero if dev->min_header_len != dev->hard_header_len to handle
> > > variable-length link layer headers. Is there something I
> > > misunderstand, or are there other aspects I need to consider further?
> >
> > That's right.
> >
> > The min_header_len != hard_header_len check is annoying and may seem
> > pedantic. But it's the only way to trust that the next header starts
> > at hard_header_len.
> 
> Thank you for your advice.
> I have implemented the modification, but I found that the
> (min_header_len != hard_header_len) check results in unexpected
> behavior in the following test scenario:
>     ip link add link ens18 ens18.24 type vlan proto 802.1ad id 24
>     ip link add link ens18.24 ens18.24.25 type vlan proto 802.1Q id 25
>     ifconfig ens18.24 1.0.24.1/24
>     ifconfig ens18.24.25 1.0.25.1/24
>     ping -n 1.0.25.3 > /dev/null 2>&1 &
>     tcpdump -nn -i any -y LINUX_SLL -Q out not tcp and not udp
> 
> While receiving a packet from ens18.24.25 (802.1Q), the min_header_len
> and hard_header_len are 14 and 18, respectively.
> This check results in the TCI being 0 instead of 25.
> Should we skip this check to display the correct value, or is there
> another check that can achieve the same purpose?

Interesting. Glad you found this.

Makes sense, as VLAN devices have

    vlandev->hard_header_len = dev->hard_header_len + VLAN_HLEN;

Does

    if (min_header_len && min_header_len != hard_header_len)

resolve it?

Few devices actually set min_header_len. Initially, only Ethernet in
ether_setup() and loopback. It was introduced for validation in
dev_validate_header, and a min_header_len of 0 just skips some basic
validation.

As long as VLAN devices do not initialize min_header_len (e.g., by
inheriting it from the physical device and incorrectly setting it to
ETH_HLEN), then this should be fine.