libnetfilter_queue: get link layer header

libnetfilter_queue: get link layer header

[Gregor Maier]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi,

I was wondering if it's somehow possible to aquire the complete link
level header in libnetfilter_queue. I can get the HW Protocol
(Ethertype) and the SRC MAC but it is also possible to get the DST MAC.

I know that for outgoing packets the information is only available after
the arp lookup (so in/after POSTROUTING) but for INPUT / FORWARD packets
the DST Mac is always available. Maybe it would be a solution to set the
DST MAC to all zeros if it's not yet known and then add a mechanism to
pass the DST Mac to the userspace app listeing on the queue.

cu
gregor
- --
Gregor Maier                                      Lehrstuhl Informatik 8
gregor@net.in.tum.de                              Tel: +49 89  289-18010
http://www.net.in.tum.de                                     TU Muenchen
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (Darwin)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFD+0mIdGiwgbikMYMRAgzZAKCjSvZq0gxe3ZtxX+CfnHX+cYG7hgCfXL40
QYqK7Jyt1gbbtmNKi1mP+YQ=
=QoqI
-----END PGP SIGNATURE-----

Re: libnetfilter_queue: get link layer header

[Patrick McHardy]

Gregor Maier wrote:
> I was wondering if it's somehow possible to aquire the complete link
> level header in libnetfilter_queue. I can get the HW Protocol
> (Ethertype) and the SRC MAC but it is also possible to get the DST MAC.
> 
> I know that for outgoing packets the information is only available after
> the arp lookup (so in/after POSTROUTING) but for INPUT / FORWARD packets
> the DST Mac is always available. Maybe it would be a solution to set the
> DST MAC to all zeros if it's not yet known and then add a mechanism to
> pass the DST Mac to the userspace app listeing on the queue.

I agree that the current implementation isn't perfect since it always
asks the driver to fill in its own address, even if the received packet
was a broadcast packet, instead of using the data from the packet.
I've added it to my TODO list to look into this.