[libvirt PATCH v2 19/56] src: conditionalize use of net/if.h

[Daniel P. Berrangé]

On Tue, Jan 28, 2020 at 05:42:47PM +0100, Pavel Hrdina wrote:
> On Tue, Jan 28, 2020 at 01:11:00PM +0000, Daniel P. Berrangé wrote:
> > The net/if.h is not portable so we must check for its
> > existance and avoid using it when missing. Some use
> > of net/if.h was redundant and could be removed.
> > 
> > Signed-off-by: Daniel P. Berrangé <berrange at redhat.com>
> > ---
> >  build-aux/syntax-check.mk  | 2 +-
> >  configure.ac               | 1 +
> >  src/util/virnetdev.c       | 1 -
> >  src/util/virnetdev.h       | 4 +++-
> >  src/util/virnetdevbridge.c | 4 +++-
> >  src/util/virnetdevip.c     | 4 +++-
> >  src/util/virnetdevtap.c    | 4 +++-
> >  7 files changed, 14 insertions(+), 6 deletions(-)
> 
> Another mess with ifdefs in our code.  The header net/if.h provides:
> 
>     struct ifreq{} which we guard with HAVE_STRUCT_IFREQ,
> 
>     if_nametoindex which we guard with HAVE_STRUCT_IFREQ or
>         with __linux__ && HAVE_LIBNL or with IFDATA_DRIVERNAME
> 
>     if_indextoname which we guard with HAVE_IF_INDEXTONAME
> 
> Commit <ccaf0beec84b3f55f5206a71e2f1b768cc58cdda> removed the
> HAVE_NET_IF_H guard for net/if.h in favor of the gnulib module which
> is essentially reverted by this patch.
> 
> I would like to cleanup this ifdef guard mess someday but that is out of
> scope of this patch series.  I just wanted to point it out.

Yeah, I find the #ifdefs in the virnetdev*.c files quite scary.

I think in many/most/all cases there we'd be better off just
doing  #ifdef __Linux__ / __freebsd__, instead of testing
against countless different probed symbols.

> Reviewed-by: Pavel Hrdina <phrdina at redhat.com>



Regards,
Daniel
-- 
|: https://berrange.com      -o-    https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org         -o-            https://fstop138.berrange.com :|
|: https://entangle-photo.org    -o-    https://www.instagram.com/dberrange :|