Re: BUG:af_packet fails to TX TSO frames

[Anton Ivanov <anton.ivanov@cambridgegreys.com>]

On 10/12/17 18:25, Willem de Bruijn wrote:
> On Thu, Oct 12, 2017 at 12:30 PM, Willem de Bruijn
> <willemdebruijn.kernel@gmail.com> wrote:
>> On Thu, Oct 12, 2017 at 11:44 AM, Anton Ivanov
>> <anton.ivanov@cambridgegreys.com> wrote:
>>> Found it.
>>>
>>> Two bugs canceling each other.
>>> The bind sequence in:  psock_txring_vnet.c is wrong.
>>>
>>> It does the following addr.sll_protocol =    htons(ETH_P_IP);
>>> before calling bind.
>>>
>>> If you set addr.sll_protocol to ETH_P_ALL where it should have been in the
>>> first place the test program blows up with -ENOBUFS
>> There is no such requirement that the socket should bind to ETH_P_ALL.

There is no requirement to bind to ETH_P_IP either and most code 
examples going back more than 10 years to the days of TCP Illustrated 
use ALL.

I just did the v6 test - if you bind with ETH_P_IP GSO on TCPv6 is 
broken and returns NOBUF and vice versa.

>>
>>> I think what is happening is that this value is taken into account when
>>> looking at "what should I use to segment it with" in skb_mac_gso_segment
>>> which is invoked at the end of the verification chain which starts in
>>> packet_direct_xmit in af_packet.c
>> packet_snd sets skb->protocol based on the protocol that the packet
>> socket is bound to. Binding to ETH_P_IP is the right choice here.
> To avoid having to open multiple sockets for different protocols,
> sockaddr_ll can also be passed in the msg_name argument on
> each call.

Does not work for vnet headers - it honors what you bound with. I tried 
to bind with ETH_ALL and pass ETH_P_IP as an arg and it ENOBUF-ed

>

-- 
Anton R. Ivanov

Cambridge Greys Limited, England and Wales company No 10273661
http://www.cambridgegreys.com/