On Thu, Jan 05, 2023 at 11:09:57AM +0100, Theo Buehler wrote:
> On Thu, Jan 05, 2023 at 11:03:04AM +0100, Claudio Jeker wrote:
> > gcc4 does not really support C99 initalizers. It works most of the time
> > but fails for more complex structs. Just fall back to memset() here.
>
> deraadt used { {0} } in kr_send_dependon(). Apparently that works.
> I really don't understand why we can't use a 24 years old standard.
It is very annoying. Would be nice if our gcc would allow it since c99
initalizers are a good thing. Not sure I like the { { 0 } } fix. Another
option is to init a specific value e.g.
struct bgpd_addr gateway = { .aid = 0 };
but again this is just an ugly workaround.
We could silence the warning in gcc but I'm not a big fan of that since
struct bgpd_addr gateway = { 1 };
should get us a warning.
gcc fixed this in https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64709
but the code was already quite different to what we have.
As a third option one could use the C23
struct bgpd_addr gateway = { };
This seems to work with our gcc and llvm.
> > --
> > :wq Claudio
> >
> > Index: kroute.c
> > ===================================================================
> > RCS file: /cvs/src/usr.sbin/bgpd/kroute.c,v
> > retrieving revision 1.303
> > diff -u -p -r1.303 kroute.c
> > --- kroute.c 28 Dec 2022 21:30:16 -0000 1.303
> > +++ kroute.c 5 Jan 2023 09:55:39 -0000
> > @@ -2159,10 +2159,12 @@ kroute6_validate(struct kroute6 *kr)
> > int
> > knexthop_true_nexthop(struct ktable *kt, struct kroute_full *kf)
> > {
> > - struct bgpd_addr gateway = { 0 };
> > + struct bgpd_addr gateway;
> > struct knexthop *kn;
> > struct kroute *kr;
> > struct kroute6 *kr6;
> > +
> > + memset(&gateway, 0, sizeof(gateway));
> >
> > /*
> > * Ignore the nexthop for VPN routes. The gateway is forced
> >
>
--
:wq Claudio
