[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <CAK-6q+hvw-3EtSoQA-01w+RPCZ1mUDtPEyw-um9a6ocOzW+ovg@mail.gmail.com>
Date: Sun, 26 Jun 2022 22:31:29 -0400
From: Alexander Aring <aahringo@...hat.com>
To: Jakub Kicinski <kuba@...nel.org>
Cc: Stefan Schmidt <stefan@...enfreihafen.org>,
linux-wpan - ML <linux-wpan@...r.kernel.org>,
linux-bluetooth@...r.kernel.org,
Network Development <netdev@...r.kernel.org>
Subject: Re: 6lowpan netlink
Hi,
On Thu, Jun 16, 2022 at 12:34 PM Jakub Kicinski <kuba@...nel.org> wrote:
>
> On Thu, 16 Jun 2022 09:00:08 -0400 Alexander Aring wrote:
> > > > I want to spread around that I started to work on some overdue
> > > > implementation, a netlink 6lowpan configuration interface, because
> > > > rtnetlink is not enough... it's for configuring very specific 6lowpan
> > > > device settings.
> > >
> > > Great, looking forward to it!
> >
> > I would like to trigger a discussion about rtnetlink or generic. I can
> > put a nested rtnetlink for some device specific settings but then the
> > whole iproute2 (as it's currently is) would maintain a specific
> > 6lowpan setting which maybe the user never wants...
> > I think we should follow this way when there is a strict ipv6 device
> > specific setting e.g. l2 neighbor information in ipv6 ndisc.
>
> Unless you'll have to repeat attributes which are already present
> in rtnetlink in an obvious way genetlink > rtnetlink.
I think this is not the case, I do not implement a new protocol
family... I just need to manipulate some 6lowpan stack settings that
aren't part of the current ipv6 protocol familty rtnetlink operations.
Thanks.
- Alex
Powered by blists - more mailing lists