netdev - Re: [PATCH net-next 3/4] bpf: BPF for lightweight tunnel encapsulation

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]

Date:   Mon, 31 Oct 2016 16:06:40 +0100
From:   Thomas Graf <tgraf@...g.ch>
To:     Tom Herbert <tom@...bertland.com>
Cc:     "David S. Miller" <davem@...emloft.net>,
        Alexei Starovoitov <alexei.starovoitov@...il.com>,
        Daniel Borkmann <daniel@...earbox.net>,
        Linux Kernel Network Developers <netdev@...r.kernel.org>,
        roopa <roopa@...ulusnetworks.com>
Subject: Re: [PATCH net-next 3/4] bpf: BPF for lightweight tunnel
 encapsulation

On 10/31/16 at 07:17am, Tom Herbert wrote:
> On Mon, Oct 31, 2016 at 5:59 AM, Thomas Graf <tgraf@...g.ch> wrote:
> > Noticed while implementing this: How does ILA ensure that dst_output()
> > is not invoked in a circular manner?
> >
> > dstA->output() -> dstB->otuput() -> dstA->output() -> ...
> 
> It doesn't. We'll need to add a check for that. Maybe the rule should
> be that an skbuff is only allowed to hit one LWT route?

I'll add a per cpu variable to do a recursion limit for dst_output()
which callers to possibly recurse can use. ILA can use that as well.

> Another scenario to consider: Suppose someone is doing protocol
> translation like in RFC7915. This is one of operations we'd need with
> ILA or GRE to implement an IPv4 overlay network over IPv6. Would this
> be allowed/supported in LWT BPF?

In lwtunnel_xmit() yes, input and output would not support this right
now. It will need some logic as the orig_input and orig_output would
obviously be expecting the same protocol. This is the reason why the
xmit prog type currently has a wider set of allowed helpers.