netdev - Re: [PATCH bpf-next v3 4/6] selftests/bpf: refactor xdp_context

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: <a05ba835-b109-46c9-ac3c-ddf8a3e48fcf@hetzner-cloud.de>
Date: Wed, 26 Feb 2025 18:39:06 +0100
From: Marcus Wichelmann <marcus.wichelmann@...zner-cloud.de>
To: Willem de Bruijn <willemdebruijn.kernel@...il.com>,
 netdev@...r.kernel.org, linux-kernel@...r.kernel.org, bpf@...r.kernel.org,
 linux-kselftest@...r.kernel.org
Cc: jasowang@...hat.com, andrew+netdev@...n.ch, davem@...emloft.net,
 edumazet@...gle.com, kuba@...nel.org, pabeni@...hat.com, andrii@...nel.org,
 eddyz87@...il.com, mykolal@...com, ast@...nel.org, daniel@...earbox.net,
 martin.lau@...ux.dev, song@...nel.org, yonghong.song@...ux.dev,
 john.fastabend@...il.com, kpsingh@...nel.org, sdf@...ichev.me,
 haoluo@...gle.com, jolsa@...nel.org, shuah@...nel.org, hawk@...nel.org
Subject: Re: [PATCH bpf-next v3 4/6] selftests/bpf: refactor
 xdp_context_functional test and bpf program

Am 25.02.25 um 19:23 schrieb Willem de Bruijn:
> Marcus Wichelmann wrote:
>> Am 24.02.25 um 16:29 schrieb Marcus Wichelmann:
>>> [...]
>>> +	/* By default, Linux sends IPv6 multicast listener reports which
>>> +	 * interfere with this test. Set the IFF_NOARP flag to ensure
>>> +	 * silence on the interface.
>>> +	 */
>>> +	SYS(close, "ip link set dev " RX_NAME " arp off");
>>>    	SYS(close, "ip link set dev " RX_NAME " up");
>>
>> Hm, setting the NOARP flag seems to have not been sufficient to fix the flaky
>> test:
>> https://github.com/kernel-patches/bpf/actions/runs/13507111620/job/37739614229
>>
>> I was not able to reproduce it locally or with my own CI runs unfortunately, but
>> I'll try something else in the next patch version which should definitely stop
>> IPv6 multicast listener report packets from messing with the tests.
> 
> You probably want to pass nodad to any ip -6 addr add.
> 
> This is a common option you'll find in tools/testing/selftests/net.
> 
> RFC 3810 section 5.2.13 says
> 
> "
>     For stateless autoconfiguration, as defined in [RFC2462], a node is
>     required to join several IPv6 multicast groups, in order to perform
>     Duplicate Address Detection (DAD).
> "

I'm not explicitly adding an IPv6 address, but this rather comes from the IPv6
link-local address being automatically assigned.
This can be avoided by setting addrgenmode=none, either using "ip link" or by
setting the corresponding sysctl.

But I think I'll add a filter for all non test related packets to the XDP program
instead.
I like this solution more, as it should be easier to grasp, very reliable and also
solves this for all tests in just one place.

Thanks!