On Sun, Jul 15, 2001 at 01:11:44AM +0200, Alexander Langer wrote:
> Thus spake Brooks Davis ([EMAIL PROTECTED]):
>
> People. This is a good example why style fixes shouldn't happen
> with content fixes. I can't figure out what changed in this file:
>
> > RCS file: /home/ncvs/src/sys/modules/Makefile,v
> > retrieving revision 1.190
> > diff -u -r1.190 Makefile
> > --- sys/modules/Makefile 2001/07/08 04:17:26 1.190
> > +++ sys/modules/Makefile 2001/07/13 23:10:49
dd pointed this out too. I just forgot that I had patched that. I'll
commit it seperatly. (The actual change is adding if_faith.)
> > diff -u -r1.6 if_faith.c
> > --- sys/net/if_faith.c 2001/07/05 14:42:54 1.6
> > +++ sys/net/if_faith.c 2001/07/14 01:21:02
>
> I'm just curious: Why don't you need to include sys/module.h?
> You don't include any other file that itself includes it, but
> module.h defines DECLARE_MODULE and friends, which you are using.
> I'm a little bit confused :-)
I'm not sure. Things worked, so I didn't add more includes.
> > +#include <machine/bus.h> /* XXX: Shouldn't really be required! */
>
> Why do you include it then? :-)
Because sys/rman.h bogusly contains refrences to bus_space_tag_t and
bus_space_handle_t. The comment is actually brain's since that code is
pretty much copyed from the cloning code for tun(4).
Thanks,
Brooks
--
Any statement of the form "X is the one, true Y" is FALSE.
PGP fingerprint 655D 519C 26A7 82E7 2529 9BF0 5D8E 8BE9 F238 1AD4
PGP signature
