Re: linux-next: build failure after merge of the final tree (net-next tree related)

[Andrew Morton <akpm@linux-foundation.org>]

On Wed, 23 Nov 2011 07:09:37 -0500
Neil Horman <nhorman@tuxdriver.com> wrote:

> On Wed, Nov 23, 2011 at 03:00:04PM +1100, Stephen Rothwell wrote:
> > Hi all,
> > 
> > After merging the final tree, today's linux-next build (powerpc allnoconfig)
> > failed like this:
> > 
> > In file included from include/linux/netdevice.h:53:0,
> >                  from include/linux/icmpv6.h:173,
> >                  from include/linux/ipv6.h:220,
> >                  from include/net/ipv6.h:16,
> >                  from include/linux/sunrpc/clnt.h:26,
> >                  from include/linux/nfs_fs.h:50,
> >                  from init/do_mounts.c:20:
> > include/net/netprio_cgroup.h:23:29: error: field 'css' has incomplete type
>
> ...
>
> FYI, I've got a more appropriate fix building right now getting posted for the
> net-next tree.  I'll cc you on it.

This header looks to be pretty screwed up:

- Extraneous newline after "struct cgroup_netprio_state"

- Weird special-casing of the CONFIG_CGROUPS &&
  !CONFIG_NETPRIO_CGROUP case looks suspicious.

- Unnecessary use of IS_ENABLED - jsut use #ifdef CONFIG_NETPRIO_CGROUP

- Adds code stubs specifically for the CONFIG_CGROUPS &&
  !CONFIG_NETPRIO_CGROUP case.  Doesn't add any for the !CONFIG_CGROUPS
  case.  Seems wrong.

- Uses empty macros for the sock_update_netprioidx() and
  skb_update_prio() which can cause build errors.  Should be changed to
  plain old typechecked inline C functions.  If that causes build
  errors then something else is screwed up.

- Adds a stub for skb_update_prio() but there's no non-stub version
  of it.

also

- doesn't compile


Needs a thorough redo.  Please take a look at the other cgroup
subsystem headers.  The basic pattern is

#ifdef CONFIG_FOO_CGROUP
<definitions and declarations>
#else
<any needed stubs go here>
#endif

And that's it.