Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
MIME-Version: 1.0
References: <CAFSh4UwMr7t+R9mWUCjdecadJL6=_7jdgagAQK6Y1Yj0+Eu0sg@mail.gmail.com>
 <CAFSh4UwAmR+sdfbdyxHRDnDr8r+TXxo2bvWtY3gmLAJekWc3Sw@mail.gmail.com>
 <CAFSh4Uwsj5GfPRUe+oT8h=DBxHppqbE-zsDV8-J5rTK3-xyZFQ@mail.gmail.com>
 <CAADnVQ+tfm-k1Pz3bGm9oVJzayMgg=prenqhqrPfm3QnaCqL7Q@mail.gmail.com> <CANn89iL87E65sYSP0JTa8_WmKsOySM1NQqxg0Ot8+ggZ73F+vg@mail.gmail.com>
In-Reply-To: <CANn89iL87E65sYSP0JTa8_WmKsOySM1NQqxg0Ot8+ggZ73F+vg@mail.gmail.com>
From:   Tom Cook <tom.k.cook@gmail.com>
Date:   Fri, 15 Jan 2021 22:45:59 +0000
Message-ID: <CAFSh4UzG=WpnLr5ZYCK16t3M22HV9gf-UuvM4Un+8sQBm2uSVw@mail.gmail.com>
Subject: Re: cBPF socket filters failing - inexplicably?
To:     Eric Dumazet <edumazet@google.com>
Cc:     Alexei Starovoitov <alexei.starovoitov@gmail.com>,
        bpf <bpf@vger.kernel.org>,
        Network Development <netdev@vger.kernel.org>,
        LKML <linux-kernel@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Fri, Jan 15, 2021 at 2:16 PM Eric Dumazet <edumazet@google.com> wrote:
[snip]
> > My wild guess is that as soon as socket got created:
> > socket(PF_PACKET, SOCK_RAW, htons(ETH_P_ALL));
> > the packets were already queued to it.
> > So later setsockopt() is too late to filter.
> >
> > Eric, thoughts?
>
> Exactly, this is what happens.

I understand.  Thanks for the explanation.

> I do not know how tcpdump and other programs deal with this.
>
> Maybe by setting a small buffer size, or draining the queue.

libpcap has its own cBPF implementation which it applies after it
receives the packets from the queue.

Thanks again,
Tom Cook