Re: IP_HDRINCL byte ordering

[Roy Marples <roy%marples.name@localhost>]

 ---- On Mon, 23 Jun 2025 17:08:07 +0100  Michael van Elst <mlelstv%serpens.de@localhost> wrote --- 
 > roy%marples.name@localhost (Roy Marples) writes:
 > 
 > >No other part of the system behaves like this (that I know of) and it's not documented in ip(4).
 > 
 > https://litux.nl/mirror/unixnetworkprogramming/0131411551_ch28lev1sec3.html
 > 
 > raw sockets aren't portable, and the IP_HDRINCL is even less. With IPv6
 > not even supporting the option, I'd would rather keep the historical
 > behaviour.
 
And yet we claim compatibility with the Linux layer for IP_HDRINCL?
If we don't then surely we should return ENOTSUP instead?

I'm not suggesting we change the historical behaviour, that's why I'm adding a new knob which maintains the status quo and allows for Linux compat.
ie, we have both behaviours.

Roy