Re: [PATCH 04/30] ipcns: add create_ipc_ns()

[Alexey Dobriyan <adobriyan@gmail.com>]

On Fri, Apr 10, 2009 at 11:01:49AM +0200, Ingo Molnar wrote:
> 
> * Alexey Dobriyan <adobriyan@gmail.com> wrote:
> 
> > --- a/include/linux/ipc_namespace.h
> > +++ b/include/linux/ipc_namespace.h
> > @@ -97,6 +97,7 @@ static inline int mq_init_ns(struct ipc_namespace *ns) { return 0; }
> >  
> >  #if defined(CONFIG_IPC_NS)
> >  extern void free_ipc_ns(struct ipc_namespace *ns);
> > +struct ipc_namespace *create_ipc_ns(void);
> >  extern struct ipc_namespace *copy_ipcs(unsigned long flags,
> >  				       struct ipc_namespace *ns);
> >  extern void free_ipcs(struct ipc_namespace *ns, struct ipc_ids *ids,
> 
> Hm, doesnt the existing, correct pattern strike your eyes out:
> 
>   extern foo1();
>   extern foo2();
>   extern foo3();
> 
> and then you add a new method in this inconsistent way:
> 
>   extern foo1();
>   foox();
>   extern foo2();
>   extern foo3();
> 
> Instead of continuing the existing pattern via:
> 
>   extern foo1();
>   extern foox();
>   extern foo2();
>   extern foo3();
> 
> ?
> 
> I think we need a new checkpatch warning for such things. It might 
> be a small detail in the big picture, but a thousand small details 
> create a big mess easily so we have to try to get all the small 
> details right, all the time - that is the only way to create a 
> better kernel in the end.

We (OK, I don't care anyway) should remove extern in front of prototypes.

They slow down compilation, they make easier to overcome 80-column limit resulting in
multiline prototypes which are ugly.