Re: Problem building mc on SuSE 6.2
- From: Robert L Krawitz <rlk alum mit edu>
- to: gnome-devel-list gnome org
- Subject: Re: Problem building mc on SuSE 6.2
- Date: Wed, 29 Sep 1999 22:05:08 -0400
Date: Tue, 28 Sep 1999 15:51:09 -0500
From: Miguel de Icaza <miguel@gnu.org>
> This problem in subshell.c seems to be because the check for grantpt
> succeeds. grantpt is found in my libc.so.6 (glibc 2.1). The fact
> that it was found cached doesn't really matter; my system hasn't
> changed since I last built it (and indeed it does exist, so that test
> is at least in principle correct).
>
> #ifdef HAVE_GRANTPT
> # include <stropts.h> /* For I_PUSH */
> #else
> # include <grp.h> /* For the group struct & getgrnam() */
> #endif
>
> I'm not sure what the best fix is for this. Certainly the use of
> I_PUSH is not conditionalized on HAVE_GRANTPT. Maybe it should be
> something like this?
>
> #if defined(HAVE_GRANTPT) && !defined(__osf__) && !defined(linux) && !defined(__linux__)
No, actually the fix would be to autoconf for stropts.h and if so, use
the I_PUSH variant.
Can you submit a fix?
Here goes. It turns out that there were also problems with statfs
(independent); this patch fixes both problems.
diff -rc xxxxxxx/mc-4.5.39/configure.in mc-4.5.39/configure.in
*** xxxxxxx/mc-4.5.39/configure.in Thu Sep 23 13:01:07 1999
--- mc-4.5.39/configure.in Wed Sep 29 21:04:06 1999
***************
*** 222,227 ****
--- 222,232 ----
AC_CHECK_HEADERS(sys/select.h)
dnl
+ dnl We really want to check for stropts.h
+ dnl
+ AC_CHECK_HEADERS(stropts.h)
+
+ dnl
dnl On SCO and some SVR4, crypt is on libcrypt.a
dnl grantpt in libpt.a
dnl
diff -rc xxxxxxx/mc-4.5.39/mc-aclocal.m4 mc-4.5.39/mc-aclocal.m4
*** xxxxxxx/mc-4.5.39/mc-aclocal.m4 Thu Sep 23 13:01:07 1999
--- mc-4.5.39/mc-aclocal.m4 Wed Sep 29 22:00:05 1999
***************
*** 775,786 ****
AC_CHECKING(how to get filesystem space usage)
space=no
! # Here we'll compromise a little (and perform only the link test)
! # since it seems there are no variants of the statvfs function.
if test $space = no; then
# SVR4
! AC_CHECK_FUNCS(statvfs)
! if test $ac_cv_func_statvfs = yes; then
space=yes
AC_DEFINE(STAT_STATVFS)
fi
--- 775,795 ----
AC_CHECKING(how to get filesystem space usage)
space=no
! # Perform only the link test since it seems there are no variants of the
! # statvfs function. This check is more than just AC_CHECK_FUNCS(statvfs)
! # because that got a false positive on SCO OSR5. Adding the declaration
! # of a `struct statvfs' causes this test to fail (as it should) on such
! # systems. That system is reported to work fine with STAT_STATFS4 which
! # is what it gets when this test fails.
if test $space = no; then
# SVR4
! AC_CACHE_CHECK([statvfs function (SVR4)], fu_cv_sys_stat_statvfs,
! [AC_TRY_LINK([#include <sys/types.h>
! #include <sys/statvfs.h>],
! [struct statvfs fsd; statvfs (0, &fsd);],
! fu_cv_sys_stat_statvfs=yes,
! fu_cv_sys_stat_statvfs=no)])
! if test $fu_cv_sys_stat_statvfs = yes; then
space=yes
AC_DEFINE(STAT_STATVFS)
fi
diff -rc xxxxxxx/mc-4.5.39/src/subshell.c mc-4.5.39/src/subshell.c
*** xxxxxxx/mc-4.5.39/src/subshell.c Thu Sep 23 13:01:20 1999
--- mc-4.5.39/src/subshell.c Wed Sep 29 21:09:49 1999
***************
*** 62,68 ****
# define WIFEXITED(stat_val) (((stat_val) & 255) == 0)
#endif
! #ifdef HAVE_GRANTPT
# include <stropts.h> /* For I_PUSH */
#else
# include <grp.h> /* For the group struct & getgrnam() */
--- 62,68 ----
# define WIFEXITED(stat_val) (((stat_val) & 255) == 0)
#endif
! #ifdef HAVE_STROPTS_H
# include <stropts.h> /* For I_PUSH */
#else
# include <grp.h> /* For the group struct & getgrnam() */
--
Robert Krawitz <rlk@alum.mit.edu> http://www.tiac.net/users/rlk/
Tall Clubs International -- http://www.tall.org/ or 1-888-IM-TALL-2
Member of the League for Programming Freedom -- mail lpf@uunet.uu.net
"Linux doesn't dictate how I work, I dictate how Linux works."
--Eric Crampton
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]