[bytor@logicsouth.com] asm/types.h, netatalk/at.h [Was Re: 2.1.105 sound]

[Andreas Jaeger <aj@arthur.rhein-neckar.de>]

Hi,

I fear that bytor is right.  <asm/types.h> defines umode_t as unsigned 
short - and glibc 2.1 includes this header in a number of places:
$ grep "asm/types.h" net*/*.h
net/ppp_defs.h:#include <asm/types.h>
netatalk/at.h:#include <asm/types.h>
netinet/icmp.h:#include <asm/types.h>
netinet/if_fddi.h:#include <asm/types.h>
netinet/if_tr.h:#include <asm/types.h>
netinet/igmp.h:#include <asm/types.h>

What shall we do?  Provide our own version of <asm/types.h>?  Patch
<asm/types.h> and try to get it into Linux? ...

Andreas
-- 
 Andreas Jaeger   aj@arthur.rhein-neckar.de    jaeger@informatik.uni-kl.de
  for pgp-key finger ajaeger@alma.student.uni-kl.de

---

• To: linux-kernel@vger.rutgers.edu
• Subject: asm/types.h, netatalk/at.h [Was Re: 2.1.105 sound]
• From: bytor@logicsouth.com
• Date: Mon, 8 Jun 1998 19:57:33 -0400 (EWT)

On Mon, 8 Jun 1998, Alan Cox wrote:

> > Also, I noticed what may be a problem in <asm/types.h> and at least glibc
> > 2.1  (haven't tested 2.0). There is a type conflict between the glibc and
> > kernel declarations of umode_t. If its not a glibc 2.0 problem, a little
> > preventative medicine may be in order to avoid excess confusion at a later
> > date when 2.1 is final.
> 
> You shouldnt be mixing kernel and glibc headers for one thing.

Most certainly not, however this one either slipped by the glibc
maintainers or is necessary... netatalk/at.h includes <asm/types.h>, which
generates the conflict...

       +----------------------+----------------------------------------+
       | bytor@tucka.nmhu.edu | UNIX _is_ user-friendly. It's just not |
       |     By-Tor@EfNet     | ignorant-friendly and idiot-friendly.  |
       +----------------------+----------------------------------------+


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu

---