public inbox for netdev@vger.kernel.org
 help / color / mirror / Atom feed
From: "Pali Rohár" <pali@kernel.org>
To: Martin Olivier <martin.olivier@live.fr>,
	Qingfang Deng <dqfext@gmail.com>
Cc: netdev@vger.kernel.org, linux-ppp@vger.kernel.org,
	andrew+netdev@lunn.ch, davem@davemloft.net, edumazet@google.com,
	kuba@kernel.org, pabeni@redhat.com, linux-kernel@vger.kernel.org,
	Paul Mackerras <paulus@ozlabs.org>,
	Guillaume Nault <gnault@redhat.com>
Subject: Re: [PATCH net-next v2] ppp: add IFLA_PPP_UNIT netlink attribute
Date: Fri, 3 Apr 2026 16:58:55 +0200	[thread overview]
Message-ID: <20260403145855.oyhsqunzy7e2ptsa@pali> (raw)
In-Reply-To: <20260318100738.vcnp7kntuzavyula@pali>

On Wednesday 18 March 2026 11:07:38 Pali Rohár wrote:
> On Wednesday 18 March 2026 09:59:29 Qingfang Deng wrote:
> > On Sat, 14 Mar 2026 02:14:29 +0100, Martin Olivier wrote:
> > > Currently, the PPP rtnetlink API allows creating a new network interface
> > > with a custom ifname, but it lacks the ability to specify a custom PPP
> > > unit id.
> > > 
> > > Setting a specific unit id is currently only possible with the
> > > PPPIOCNEWUNIT ioctl. If a user-space program also requires a custom
> > > interface name, it must create the interface first with PPPIOCNEWUNIT
> > > and then rename it.
> > > 
> > > Resolve this by introducing the IFLA_PPP_UNIT netlink attribute. This
> > > allows user-space programs to atomically request both a custom ifname
> > > and a specific PPP unit id during the RTM_NEWLINK creation process,
> > > eliminating the post-creation renaming for this use case.
> > > 
> > > Signed-off-by: Martin Olivier <martin.olivier@live.fr>
> > > ---
> > > Changes in v2:
> > > - use nl policy to set IFLA_PPP_UNIT min allowed value instead of a manual check in ppp_nl_validate()
> > > - use of nla_get_s32_default() to collect IFLA_PPP_UNIT value
> > > Link to v1: https://lore.kernel.org/netdev/PAWP192MB2411A5E7D3BE1B55E155A92F9747A@PAWP192MB2411.EURP192.PROD.OUTLOOK.COM/
> > 
> > The patch itself looks good to me, but I would like to check the
> > userspace changes too. Please create a pull request at
> > https://github.com/ppp-project/ppp/pulls
> > 
> > +Cc: Paul Mackerras, Guillaume Nault, Pali Rohár
> > 
> > Regards,
> > Qingfang
> 
> Hello Martin, in past I have sent similar change:
> https://lore.kernel.org/linux-ppp/20210807163749.18316-1-pali@kernel.org/T/#u
> Look at the discussion, it can be useful to understand why the change
> was not accepted.

Anyway, if you want to continue on this patch, here are my points for review:

- I would suggest to use IFLA_PPP_UNIT_ID name for consistency. We
  already have IFLA_PPP_DEV_FD value where is FD suffix which is saying
  what it is. So for naming consistency is the suffix "_ID" useful.

- Allow to specify -1 as an valid input. Currently the
  NLA_POLICY_MIN(NLA_S32, 0) does not allow -1.
  -1 should work as a default value, like if the value is not specified
  at all.

- Add #define IFLA_PPP_UNIT_ID ifla_ppp_unit_id into the enum { ... } in
  if_link.h. This is for userspace compatibility, which can check fi
  the IFLA_PPP_UNIT_ID is provided or not (via #ifdef).

- Do not rewrite error to -EEXIST when calling the ppp_dev_configure()
  from the ppp_nl_newlink() with specified unit_id. ppp_nl_newlink() is
  a new code path which will allow to specify unit_id, and hence it is
  not useful to hide a real error and replace it by -EEXIST.

All of these points I have already handled in my version of the patch
which I sent and is available in the above link. So you can use it as an
inspiration how to easily handle them.

Pali

      parent reply	other threads:[~2026-04-03 14:59 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-03-14  1:14 [PATCH net-next v2] ppp: add IFLA_PPP_UNIT netlink attribute Martin Olivier
2026-03-17 23:14 ` Jakub Kicinski
2026-03-18  1:59 ` Qingfang Deng
2026-03-18  9:49   ` Guillaume Nault
2026-03-18 10:07   ` Pali Rohár
2026-03-18 10:11     ` Pali Rohár
2026-04-03 14:58     ` Pali Rohár [this message]

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=20260403145855.oyhsqunzy7e2ptsa@pali \
    --to=pali@kernel.org \
    --cc=andrew+netdev@lunn.ch \
    --cc=davem@davemloft.net \
    --cc=dqfext@gmail.com \
    --cc=edumazet@google.com \
    --cc=gnault@redhat.com \
    --cc=kuba@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-ppp@vger.kernel.org \
    --cc=martin.olivier@live.fr \
    --cc=netdev@vger.kernel.org \
    --cc=pabeni@redhat.com \
    --cc=paulus@ozlabs.org \
    /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