netdev - Re: [Patch bpf] udp: validate checksum in udp_read

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] [day] [month] [year] [list]

Message-ID: <61670e88c86ee_4d1c0208a3@john-XPS-13-9370.notmuch>
Date:   Wed, 13 Oct 2021 09:51:20 -0700
From:   John Fastabend <john.fastabend@...il.com>
To:     Cong Wang <xiyou.wangcong@...il.com>,
        John Fastabend <john.fastabend@...il.com>
Cc:     Linux Kernel Network Developers <netdev@...r.kernel.org>,
        bpf <bpf@...r.kernel.org>, Cong Wang <cong.wang@...edance.com>,
        Daniel Borkmann <daniel@...earbox.net>,
        Lorenz Bauer <lmb@...udflare.com>,
        Jakub Sitnicki <jakub@...udflare.com>
Subject: Re: [Patch bpf] udp: validate checksum in udp_read_sock()

Cong Wang wrote:
> On Mon, Oct 11, 2021 at 10:06 PM John Fastabend
> <john.fastabend@...il.com> wrote:
> >
> > Cong Wang wrote:
> > > From: Cong Wang <cong.wang@...edance.com>
> > >
> > > It turns out the skb's in sock receive queue could have
> > > bad checksums, as both ->poll() and ->recvmsg() validate
> > > checksums. We have to do the same for ->read_sock() path
> > > too before they are redirected in sockmap.
> > >
> > > Fixes: d7f571188ecf ("udp: Implement ->read_sock() for sockmap")
> > > Reported-by: Daniel Borkmann <daniel@...earbox.net>
> > > Reported-by: John Fastabend <john.fastabend@...il.com>
> > > Cc: Lorenz Bauer <lmb@...udflare.com>
> > > Cc: Jakub Sitnicki <jakub@...udflare.com>
> > > Signed-off-by: Cong Wang <cong.wang@...edance.com>
> > > ---
> > >  net/ipv4/udp.c | 11 +++++++++++
> > >  1 file changed, 11 insertions(+)
> > >
> > > diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c
> > > index 8536b2a7210b..0ae8ab5e05b4 100644
> > > --- a/net/ipv4/udp.c
> > > +++ b/net/ipv4/udp.c
> > > @@ -1808,6 +1808,17 @@ int udp_read_sock(struct sock *sk, read_descriptor_t *desc,
> > >               skb = skb_recv_udp(sk, 0, 1, &err);
> > >               if (!skb)
> > >                       return err;
> > > +
> > > +             if (udp_lib_checksum_complete(skb)) {
> > > +                     __UDP_INC_STATS(sock_net(sk), UDP_MIB_CSUMERRORS,
> > > +                                     IS_UDPLITE(sk));
> > > +                     __UDP_INC_STATS(sock_net(sk), UDP_MIB_INERRORS,
> > > +                                     IS_UDPLITE(sk));
> > > +                     atomic_inc(&sk->sk_drops);
> > > +                     kfree_skb(skb);
> >
> > We could use sock_drop() here? Otherwise looks good thanks.
> 
> sock_drop() is in include/linux/skmsg.h, I think we need to move it
> to sock.h before using it here in net/ipv4/udp.c, right?

Yes it would be necessary. Lets not do it here otherwise backports
will be ugly.

Acked-by: John Fastabend <john.fastabend@...il.com>

> 
> And there are other similar patterns which can be replaced with
> sock_drop(), so we can do the replacement for all in a separate
> patch.
> 
> Thanks.