On Mon, Aug 17, 2020 at 12:00 PM Randy Dunlap <rdun...@infradead.org> wrote: > > On 8/17/20 11:55 AM, Cong Wang wrote: > > On Mon, Aug 17, 2020 at 11:49 AM Randy Dunlap <rdun...@infradead.org> wrote: > >> > >> On 8/17/20 11:31 AM, Cong Wang wrote: > >>> On Sun, Aug 16, 2020 at 11:37 PM Xin Long <lucien....@gmail.com> wrote: > >>>> > >>>> On Mon, Aug 17, 2020 at 2:29 AM Cong Wang <xiyou.wangc...@gmail.com> > >>>> wrote: > >>>>> > >>>>> Or put it into struct ipv6_stub? > >>>> Hi Cong, > >>>> > >>>> That could be one way. We may do it when this new function becomes more > >>>> common. > >>>> By now, I think it's okay to make TIPC depend on IPV6 || IPV6=n. > >>> > >>> I am not a fan of IPV6=m, but disallowing it for one symbol seems > >>> too harsh. > >> > >> Hi, > >> > >> Maybe I'm not following you, but this doesn't disallow IPV6=m. > > > > Well, by "disallowing IPV6=m" I meant "disallowing IPV6=m when > > enabling TIPC" for sure... Sorry that it misleads you to believe > > completely disallowing IPV6=m globally. > > > >> > >> It just restricts how TIPC can be built, so that > >> TIPC=y and IPV6=m cannot happen together, which causes > >> a build error. > > > > It also disallows TIPC=m and IPV6=m, right? In short, it disalows > > IPV6=m when TIPC is enabled. And this is exactly what I complain, > > as it looks too harsh. > > I haven't tested that specifically, but that should work. > This patch won't prevent that from working.
Please give it a try. I do not see how it allows IPV6=m and TIPC=m but disallows IPV6=m and TIPC=y. > > We have loadable modules calling other loadable modules > all over the kernel. True, we rely on request_module(). But I do not see TIPC calls request_module() to request IPV6 module to load "ipv6_dev_find". Thanks.
