Date: Sun, 11 Dec 2016 08:15:01 +0100
From: Greg KH <gregkh@linuxfoundation.org>
To: Dan =?iso-8859-1?Q?L=FCdtke?= <mail@danrl.com>
Cc: Daniel Kahn Gillmor <dkg@fifthhorseman.net>, linux-mips@linux-mips.org,
        Netdev <netdev@vger.kernel.org>, LKML <linux-kernel@vger.kernel.org>,
        Hannes Frederic Sowa <hannes@stressinduktion.org>,
        WireGuard mailing list <wireguard@lists.zx2c4.com>
Subject: Re: Misalignment, MIPS, and ip_hdr(skb)->version
Message-ID: <20161211071501.GA32621@kroah.com>
References: <CAHmME9o_eCNXpVztOZKW55kpRtE+1KSEQTQOjUBVn68Y2+or2g@mail.gmail.com>
 <095cac5b-b757-6f4a-e699-8eedf9ed7221@stressinduktion.org>
 <87vauvhwdu.fsf@alice.fifthhorseman.net>
 <CE942916-BF45-44CC-A5F5-3838CF9C93BC@danrl.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
In-Reply-To: <CE942916-BF45-44CC-A5F5-3838CF9C93BC@danrl.com>
User-Agent: Mutt/1.7.2 (2016-11-26)
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1389
Lines: 31

On Sat, Dec 10, 2016 at 11:18:14PM +0100, Dan L?dtke wrote:
> 
> > On 8 Dec 2016, at 05:34, Daniel Kahn Gillmor <dkg@fifthhorseman.net> wrote:
> > 
> > On Wed 2016-12-07 19:30:34 -0500, Hannes Frederic Sowa wrote:
> >> Your custom protocol should be designed in a way you get an aligned ip
> >> header. Most protocols of the IETF follow this mantra and it is always
> >> possible to e.g. pad options so you end up on aligned boundaries for the
> >> next header.
> > 
> > fwiw, i'm not convinced that "most protocols of the IETF follow this
> > mantra".  we've had multiple discussions in different protocol groups
> > about shaving or bloating by a few bytes here or there in different
> > protocols, and i don't think anyone has brought up memory alignment as
> > an argument in any of the discussions i've followed.
> > 
> 
> If the trade-off is between 1 padding byte and 2 byte alignment versus
> 3 padding bytes and 4 byte alignment I would definitely opt for 3
> padding bytes. I know how that waste feels like to a protocol
> designer, but I think it is worth it. Maybe the padding/reserved will
> be useful some day for an additional feature.

Note, if you do do this (hint, I think it is a good idea), require that
these reserved/pad fields always set to 0 for now, so that no one puts
garbage in them and then if you later want to use them, it will be a
mess.

thanks,

greg k-h