This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
[Bug libfortran/15235] [gfortran] libgfortran doesn't build on Solaris 10/x86
- From: "ro at techfak dot uni-bielefeld dot de" <gcc-bugzilla at gcc dot gnu dot org>
- To: gcc-bugs at gcc dot gnu dot org
- Date: 14 May 2004 16:16:49 -0000
- Subject: [Bug libfortran/15235] [gfortran] libgfortran doesn't build on Solaris 10/x86
- References: <20040430200445.15235.ro@techfak.uni-bielefeld.de>
- Reply-to: gcc-bugzilla at gcc dot gnu dot org
------- Additional Comments From ro at techfak dot uni-bielefeld dot de 2004-05-14 16:16 -------
Subject: Re: [gfortran] libgfortran doesn't build on Solaris 10/x86
tobi at gcc dot gnu dot org writes:
> Is offset_t a macro on Solaris? Then we could maybe replace the typedef by
> #ifndef offset_t
> #define offset_t off_t
> #endif
> ?
No, it's a typedef in <sys/types.h>, which clashes with the libgfortran.h
definition:
typedef longlong_t offset_t;
vs.
typedef off_t offset_t;
> Does Solaris have off_t? Maybe it would be the smartest to just use off_t
> instead of offset_t.
Certainly: off_t is in POSIX.1, and I currently see no reason to invent
anything g95 specific. If there is, the type should be called gfc_offset_t
or something like this to avoid the clashes.
> libstdc++ handles this by defining __off_t in system-dependent headers and then
> using this. We could also copy this.
>
> Maybe an autoconf test is in place here?
Depends on what is desired here.
Rainer
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15235