netdev - Re: [PATCH net] ip6_tunnel: set inner ipproto before ip6_tnl

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-Id: <E9B679B3-58CA-4BB7-A9B9-1A28A6148D75@yandex-team.ru>
Date:   Fri, 30 Oct 2020 15:54:19 +0300
From:   Alexander Ovechkin <ovov@...dex-team.ru>
To:     Vadim Fedorenko <vfedorenko@...ek.ru>
Cc:     Willem de Bruijn <willemdebruijn.kernel@...il.com>,
        Network Development <netdev@...r.kernel.org>,
        Tom Herbert <tom@...bertland.com>
Subject: Re: [PATCH net] ip6_tunnel: set inner ipproto before ip6_tnl_encap.

On 30 Oct 2020, at 14:01, Vadim Fedorenko <vfedorenko@...ek.ru> wrote:
> Could not reproduce the bug. Could you please provide a test scenario?

It can be reproduced if your net device doesn’t support udp tunnel segmentation (i.e its features do not have SKB_GSO_UDP_TUNNEL).
If you try to send packet larger than the MTU fou6-only tunnel (without any other encap) it will be dropped, because of invalid skb->inner_ipproto (that will be equal to IPPROTO_UDP — outer protocol, instead of IPPROTO_IPV6).
skb->inner_ipproto is used here:
https://git.kernel.org/pub/scm/linux/kernel/git/netdev/net.git/tree/net/ipv4/udp_offload.c?id=07e0887302450a62f51dba72df6afb5fabb23d1c#n168