This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
Re: 2.95.3 destruction in Linux header
- To: moshier at moshier dot ne dot mediaone dot net
- Subject: Re: 2.95.3 destruction in Linux header
- From: Geoff Keating <geoffk at geoffk dot org>
- Date: Sat, 24 Mar 2001 07:23:07 -0800
- CC: gcc-bugs at gcc dot gnu dot org
- References: <Pine.LNX.4.20.0103231854040.12867-100000@moshier.ne.mediaone.net>
- Reply-to: Geoff Keating <geoffk at redhat dot com>
> Date: Fri, 23 Mar 2001 19:21:21 -0500 (EST)
> From: Stephen L Moshier <moshier@mediaone.net>
> Reply-To: moshier@moshier.ne.mediaone.net
> cc: gcc-bugs@gcc.gnu.org
>
> >From Geoff Keating <geoffk@geoffk.org>:
> > /usr/include/linux/hpfs_fs_i.h:5: parse error before `ino_t'
> >
> > On my system, if you're building the linux kernel, it gets defined in
> > <linux/types.h>. Is it defined in your linux/types.h?
>
> No, it is not, on linux 2.2.14. I think this would get it,
> from /usr/include/sys/types.h:
> /usr/include/bits/types.h:typedef __u_long __ino_t;
That's still not 'ino_t'. Perhaps somewhere ino_t is defined to be
__ino_t.
> The code fragment was boiled down from a device driver, and
> apparently various things are left undefined by the system headers
> that the program includes.
> The effect seems to be that the program works with old gcc's but not
> with 2.95.3.
This seems very odd. Are both gcc versions really using the same headers?
--
- Geoffrey Keating <geoffk@geoffk.org>