[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CANn89iJGdn2J-UwK9ux+m9r8mRhAND_t2kU6mLCs=RszBhCyRA@mail.gmail.com>
Date: Mon, 1 Sep 2025 09:03:03 -0700
From: Eric Dumazet <edumazet@...gle.com>
To: F6BVP <f6bvp@...e.fr>
Cc: David Ranch <linux-hams@...nnet.net>, Paolo Abeni <pabeni@...hat.com>,
Dan Carpenter <dan.carpenter@...aro.org>, linux-hams@...r.kernel.org,
netdev <netdev@...r.kernel.org>, Dan Cross <crossd@...il.com>,
Folkert van Heusden <folkert@...heusden.com>, Florian Westphal <fw@...len.de>
Subject: Re: [ROSE] [AX25] 6.15.10 long term stable kernel oops
On Mon, Sep 1, 2025 at 8:59 AM F6BVP <f6bvp@...e.fr> wrote:
>
> Radioamateur have contributed to Linux since the begining.
>
> If a protocole should be removed from Linux kernel as soon as a commit
> breaks it, Linux itself would be t be abandonned.
>
> AX25 is not responsible of a kernel Oops due to a commit in dev.c
>
> Like David KI6ZHD mentionned, many hams are still experimenting using
> packet radio.
>
> Not mentioning that a large number of pico satellites from universities
> all around the world are using AX25 for TM/TC !
>
Keep calm, I am just saying that the bisection pointed to a fine commit,
but it took a _lot_ of time to root-cause the issue.
And the bug is in ax25, not in Paolo patch.
Please test the fix, and thank me for actually working on a fix, while
I have more urgent work on my plate.
> Bernard Pidoux
> F6BVP /AI7BG
> Founder president AMSAT-France
> President Dimension Parabole
> http://radiotelescope-lavillette.fr
>
>
> Le 01/09/2025 à 16:43, David Ranch a écrit :
> >
> > Hello Eric, Everyone,
> >
> >>> At some point we will have to remove ax25, this has been quite broken
> >>> for a long time.
> >
> > I can appreciate that the code implementing AX.25 in the kernel is very
> > old but say it needs to be removed will impact a lot of people. There
> > is a very active community around AX.25 packet radio today and Linux's
> > native implementation still offers features and functions that aren't
> > implemented anywhere else. There are also some large / popular projects
> > that are dependent on it for their connectivity via libax25, etc. I
> > continue to hope someone will be willing to step forward and write a
> > modernized version of this stack (and netrom and rose too) so we can
> > continue to run things natively on Linux.
> >
> > --David
> > KI6ZHD
>
Powered by blists - more mailing lists