netdev - Re: [PATCH bpf-next 03/12] net: Add IFLA_XDP_EGRESS for XDP programs in the egress path

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 for Android: free password hash cracker in your pocket

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

Message-ID: <5290f2dd-ade6-ce0f-bf01-0872d4cfe14f@gmail.com>
Date:   Sun, 26 Jan 2020 09:38:30 -0700
From:   David Ahern <dsahern@...il.com>
To:     Jesper Dangaard Brouer <jbrouer@...hat.com>
Cc:     Toke Høiland-Jørgensen <toke@...hat.com>,
        Jakub Kicinski <kuba@...nel.org>,
        David Ahern <dsahern@...nel.org>, netdev@...r.kernel.org,
        prashantbhole.linux@...il.com, jasowang@...hat.com,
        davem@...emloft.net, mst@...hat.com, toshiaki.makita1@...il.com,
        daniel@...earbox.net, john.fastabend@...il.com, ast@...nel.org,
        kafai@...com, songliubraving@...com, yhs@...com, andriin@...com,
        David Ahern <dahern@...italocean.com>
Subject: Re: [PATCH bpf-next 03/12] net: Add IFLA_XDP_EGRESS for XDP programs
 in the egress path

On 1/26/20 5:49 AM, Jesper Dangaard Brouer wrote:
>> This has
>> been discussed before as a need (e.g, Toke's reference above), and I am
>> trying to get this initial support done.
>>
>> I very much wanted to avoid copy-paste-modify for the entire XDP API for
>> this. For the most part XDP means ebpf at the NIC driver / hardware
>> level (obviously with the exception of generic mode). The goal is
>> tempered with the need for the verifier to reject rx entries in the
>> xdp_md context. Hence the reason for use of an attach_type - existing
>> infrastructure to test and reject the accesses.
>>
>> That said, Martin's comment throws a wrench in the goal: if the existing
>> code does not enforce expected_attach_type then that option can not be
>> used in which case I guess I have to go with a new program type
>> (BPF_PROG_TYPE_XDP_EGRESS) which takes a new context (xdp_egress_md),
>> has different return codes, etc.
> 
> Taking about return codes.  Does XDP the return codes make sense for
> this EGRESS hook? (if thinking about this being egress on the real NIC).
> 
> E.g. XDP_REDIRECT would have to be supported, which is interesting, but
> also have implications (like looping packets).
> 
> E.g. what is the semantics/action of XDP_TX return code?

This has been discussed. XDP_TX in the EGRESS path could arguably be
equal to XDP_PASS.

> 
> E.g. I'm considering adding a XDP_CONGESTED return code that can cause
> backpressure towards qdisc layer.
> 
> Also think about that if this EGRESS hook uses standard prog type for
> XDP (BPF_PROG_TYPE_XDP), then we need to convert xdp_frame to xdp_buff
> (and also convert SKBs to xdp_buff).

Why? What about the patch set requires that change to be done to have
support for EGRESS path?

> 
> Are we sure that reusing the same bpf prog type is the right choice?
> 

Martin's comment about existing checking on the expected attach type is
the only reason I have seen so far to not have the same program type.

Looking at the helpers for use in XDP programs do you believe any of
those should not be allowed with EGRESS programs? Do you have any reason
to think that existing XDP capabilities should be prohibited or
different for EGRESS? As mentioned earlier the attach type can be used
to have the verifier handle small context differences (and restrict
helpers if needed).