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: <20220512112347.18717-1-andrew@daynix.com> <CACGkMEvH1yE0CZYdstAK32DkEucejNO+V7PEAZD_641+rp2aKA@mail.gmail.com>
In-Reply-To: <CACGkMEvH1yE0CZYdstAK32DkEucejNO+V7PEAZD_641+rp2aKA@mail.gmail.com>
From:   Andrew Melnichenko <andrew@daynix.com>
Date:   Tue, 24 May 2022 14:06:46 +0300
Message-ID: <CABcq3pFJcsoj+dDf6tirT_hfTB6rj9+f6KNFafwg+usqYwTdDA@mail.gmail.com>
Subject: Re: [RFC PATCH v2 0/5] TUN/VirtioNet USO features support.
To:     Jason Wang <jasowang@redhat.com>
Cc:     davem <davem@davemloft.net>, Eric Dumazet <edumazet@google.com>,
        Jakub Kicinski <kuba@kernel.org>,
        Paolo Abeni <pabeni@redhat.com>, mst <mst@redhat.com>,
        netdev <netdev@vger.kernel.org>,
        linux-kernel <linux-kernel@vger.kernel.org>,
        virtualization <virtualization@lists.linux-foundation.org>,
        Yan Vugenfirer <yan@daynix.com>,
        Yuri Benditovich <yuri.benditovich@daynix.com>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

Hi all,

The issue is that host segments packets between guests on the same host.
Tests show that it happens because SKB_GSO_DODGY skb offload in
virtio_net_hdr_from_skb().
To do segmentation you need to remove SKB_GSO_DODGY or add SKB_GSO_PARTIAL
The solution with DODGY/PARTIAL offload looks like a dirty hack, so
for now, I've lived it as it is for further investigation.


On Tue, May 17, 2022 at 9:32 AM Jason Wang <jasowang@redhat.com> wrote:
>
> On Thu, May 12, 2022 at 7:33 PM Andrew Melnychenko <andrew@daynix.com> wrote:
> >
> > Added new offloads for TUN devices TUN_F_USO4 and TUN_F_USO6.
> > Technically they enable NETIF_F_GSO_UDP_L4
> > (and only if USO4 & USO6 are set simultaneously).
> > It allows to transmission of large UDP packets.
> >
> > Different features USO4 and USO6 are required for qemu where Windows guests can
> > enable disable USO receives for IPv4 and IPv6 separately.
> > On the other side, Linux can't really differentiate USO4 and USO6, for now.
> > For now, to enable USO for TUN it requires enabling USO4 and USO6 together.
> > In the future, there would be a mechanism to control UDP_L4 GSO separately.
> >
> > Test it WIP Qemu https://github.com/daynix/qemu/tree/Dev_USOv2
> >
> > New types for VirtioNet already on mailing:
> > https://lists.oasis-open.org/archives/virtio-comment/202110/msg00010.html
> >
> > Also, there is a known issue with transmitting packages between two guests.
>
> Could you explain this more? It looks like a bug. (Or any pointer to
> the discussion)
>
> Thanks
>
> > Without hacks with skb's GSO - packages are still segmented on the host's postrouting.
> >
> > Andrew (5):
> >   uapi/linux/if_tun.h: Added new offload types for USO4/6.
> >   driver/net/tun: Added features for USO.
> >   uapi/linux/virtio_net.h: Added USO types.
> >   linux/virtio_net.h: Support USO offload in vnet header.
> >   drivers/net/virtio_net.c: Added USO support.
> >
> >  drivers/net/tap.c               | 10 ++++++++--
> >  drivers/net/tun.c               |  8 +++++++-
> >  drivers/net/virtio_net.c        | 19 +++++++++++++++----
> >  include/linux/virtio_net.h      |  9 +++++++++
> >  include/uapi/linux/if_tun.h     |  2 ++
> >  include/uapi/linux/virtio_net.h |  4 ++++
> >  6 files changed, 45 insertions(+), 7 deletions(-)
> >
> > --
> > 2.35.1
> >
>