On Fri, Mar 9, 2018 at 5:08 AM, Jiri Benc <jb...@redhat.com> wrote: > On Fri, 9 Mar 2018 04:53:07 -0800, Eric Dumazet wrote: >> Unless you bring it up ;) > > Often, you need some options to be specified, thus you end up creating > another interface anyway. > >> Compatibility problems, mostly. >> Some users might depend on existing behavior. > > How is this a compatibility problem, when the user has to opt in? > Without the default value, the kernel behaves as before. I'm sorry, > I don't see the problem, what am I missing?
How do you opt in exactly ? Once it is too late and gre0 is already there ? ip_gre can be statically built in vmlinux Are you suggesting to add a module parameter ? I wont do it but feel free to send a patch and wait for David flames :) > >> You and me would not care of breaking our setups, but maybe not >> unaware people out there. > > Sure, that's why this is an option and it defaults to off. > >> Since init_ns is created at boot time before the sysctl can be >> changed, we rather should not change the default behavior for init_ns. > > This could be a problem for built in drivers. With modules, I'm > perfectly happy doing this in initrd :-) > > To behave consistently in all cases, we might consider adding a boot > option for this, too, though. Feel free to send a patch. And do not forget we have many types of tunnels. So do not only patch gre.
