netdev - Re: [PATCH] net/ipv4/ip_gre.c net/ipv6/ip6_gre.c: ip and gre header are recorded twice

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: <20220728081701.191a405b@kernel.org>
Date:   Thu, 28 Jul 2022 08:17:01 -0700
From:   Jakub Kicinski <kuba@...nel.org>
To:     Aleksey Shumnik <ashumnik9@...il.com>
Cc:     Ido Schimmel <idosch@...sch.org>, netdev@...r.kernel.org,
        David Ahern <dsahern@...il.com>, kuznet@....inr.ac.ru,
        xeb@...l.ru
Subject: Re: [PATCH] net/ipv4/ip_gre.c net/ipv6/ip6_gre.c: ip and gre header
 are recorded twice

On Thu, 28 Jul 2022 16:54:01 +0300 Aleksey Shumnik wrote:
> On Fri, Jul 8, 2022 at 2:23 AM Jakub Kicinski <kuba@...nel.org> wrote:
> > On Thu, 7 Jul 2022 19:41:23 +0300 Aleksey Shumnik wrote:
> >
> > Yeah, I've added the neigh entries (although the v6 addresses had to
> > be massaged a little for ip neigh to take them, the commands from the
> > email don't work cause iproute2 doesn't support :: in lladdr, AFAICT).
> >
> > What I've seen in tracing was that I hit:
> >
> > ip6gre_tunnel_xmit() -> ip6_tnl_xmit_ctl() -> ip6_tnl_get_cap()
> >
> > that returns IP6_TNL_F_CAP_PER_PACKET
> >
> > so back to ip6gre_tunnel_xmit() -> goto tx_err -> error, drop
> >
> > packet never leaves the interface.  
> 
> I skipped this check so that the packets wouldn't drop.
> I compared the implementations of ip_gre.c and ip6_gre.c and I
> concluded that in ip6_tnl_xmit_ctl() instead of tunnel params
> (&ip6_tnl->parms.laddr and &ip6_tnl->parms.raddr) it is better to use
> skb network header (ipv6_hdr(skb)->saddr and ipv6_hdr(skb)->daddr).
> It is illogical to use the tunnel parameters, because if we have an
> NBMA connection, the addresses will not be set in the tunnel
> parameters and packets will always drop on ip6_tnl_xmit_ctl().
> 
> > Hm, so you did get v6 to repro? Not sure what I'm doing wrong, I'm
> > trying to repro with a net namespace over veth but that can't be it...  
> 
> Yes, just skip ip6_tnl_xmit_ctl().

Mm. Having to remove checks for packets to pass thru makes it seem like 
a lot less of a bug.