Netdev List
 help / color / mirror / Atom feed
From: Fernando Fernandez Mancera <fmancera@suse.de>
To: nicolas.dichtel@6wind.com, Paolo Abeni <pabeni@redhat.com>,
	Jakub Kicinski <kuba@kernel.org>
Cc: netdev@vger.kernel.org, linux-kselftest@vger.kernel.org,
	horms@kernel.org, edumazet@google.com, dsahern@kernel.org,
	davem@davemloft.net
Subject: Re: [PATCH 3/3 net-next v4] selftests: net: add test for IPv4 devconf netlink notifications
Date: Tue, 19 May 2026 23:27:17 +0200	[thread overview]
Message-ID: <f2cd87f0-e173-4af7-9bc8-3a1f38c58c06@suse.de> (raw)
In-Reply-To: <b8b89071-6f17-4f6a-8a06-101056f780fd@6wind.com>

On 5/19/26 3:45 PM, Nicolas Dichtel wrote:
> Le 19/05/2026 à 09:17, Paolo Abeni a écrit :
>> On 5/18/26 11:23 PM, Jakub Kicinski wrote:
>>> On Sat, 16 May 2026 09:02:27 +0200 Fernando Fernandez Mancera wrote:
>>>>>> Oh sorry, I should have clarified it. ip netconf only shows/query them
>>>>>> but it cannot set them AFAICS.
>>>>>>
>>>>>> Usage: ip netconf show [ dev STRING ]
>>>>>
>>>>> Oh, fun. anyway, you gotta find a better way. Or just give up on the
>>>>> selftest, the YNL hack is too ugly to live
>>>>>    
>>>>
>>>> Fair, let me see what I can do. Otherwise, Paolo as you requested the
>>>> selftest, would it be acceptable for you to drop it given the situation?
>>>> I looking for alternatives anyway :-)
>>>
>>> Maybe extend iproute2 ?
>>
>> Ideally I think extending iproute2 would be the better solution.
>> @Fernando: perhaps you could update the self-test accordingly, adding an
>> explicit check for the iproute subcommand. If `ip netconf set` is not
>> available skip the test, so that iproute2 patches could land later.
>>
>> Does the above look feasible?
> FWIW, netconf != devconf.
> 
> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/include/uapi/linux/ip.h#n156
> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/include/uapi/linux/netconf.h
> 
> The devconf API exposes all sysctl of an interface via RTM_NEWLINK /
> IFLA_AF_SPEC / IFLA_INET_CONF. There is no event when a sysctl changes.
> These sysctl can be set via RTM_SETLINK.
> 
> The netconf API (RTM_NEWNETCONF) is a notification mechanism used to notify
> users when a sysctl changes. It handles sysctl from interfaces but also from the
> 'all' and the 'default' entries. RTM_GETNETCONF is also supported, but not
> RTM_SETNETCONF.
> 
> 'ip netconf' uses the netconf API.
> 


I think we can reword this to.. "extend iproute2 to support setting 
devconf values"? If the naming or implementation on iproute2 is not 
clear now, we can merge this without the selftest and after the iproute2 
implementation merge the selftest.

> Regards,
> Nicolas
> 


  reply	other threads:[~2026-05-19 21:27 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-05-10  8:15 [PATCH 1/3 net-next v4] ipv4: centralize devconf sysctl handling Fernando Fernandez Mancera
2026-05-10  8:15 ` [PATCH 2/3 net-next v4] ipv4: handle devconf post-set actions on netlink updates Fernando Fernandez Mancera
2026-05-14  2:30   ` Jakub Kicinski
2026-05-14  9:16     ` Fernando Fernandez Mancera
2026-05-10  8:15 ` [PATCH 3/3 net-next v4] selftests: net: add test for IPv4 devconf netlink notifications Fernando Fernandez Mancera
2026-05-14  2:35   ` Jakub Kicinski
2026-05-14  9:42     ` Fernando Fernandez Mancera
2026-05-14 23:24       ` Jakub Kicinski
2026-05-15  8:02         ` Fernando Fernandez Mancera
2026-05-16  1:16           ` Jakub Kicinski
2026-05-16  7:02             ` Fernando Fernandez Mancera
2026-05-18 21:23               ` Jakub Kicinski
2026-05-19  7:17                 ` Paolo Abeni
2026-05-19  7:24                   ` Fernando Fernandez Mancera
2026-05-19 13:45                   ` Nicolas Dichtel
2026-05-19 21:27                     ` Fernando Fernandez Mancera [this message]
2026-05-20  7:22                       ` Nicolas Dichtel

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=f2cd87f0-e173-4af7-9bc8-3a1f38c58c06@suse.de \
    --to=fmancera@suse.de \
    --cc=davem@davemloft.net \
    --cc=dsahern@kernel.org \
    --cc=edumazet@google.com \
    --cc=horms@kernel.org \
    --cc=kuba@kernel.org \
    --cc=linux-kselftest@vger.kernel.org \
    --cc=netdev@vger.kernel.org \
    --cc=nicolas.dichtel@6wind.com \
    --cc=pabeni@redhat.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox