This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
configure problem w/ HAVE_SYS_RESOURCE_H not being defined for ix86 linux
- To: gcc-bugs at gcc dot gnu dot org
- Subject: configure problem w/ HAVE_SYS_RESOURCE_H not being defined for ix86 linux
- From: Bill Priest <bpriest at comspacecorp dot com>
- Date: Wed, 5 Sep 2001 10:57:37 -0500
All & Alan,
Building stock gcc 3.0 and 3.0.1 for linux-2.4.9 w/ ac7 patch
configure fails to define HAVE_SYS_RESOURCE_H which causes build failures.
Manually modifying auto-host.h to have #define HAVE_SYS_RESOURCE_H 1 allows
the build to get further in the compilation process.
Secondly the definitions for getrlimit and setrlimit in system.h
differ from those found in resource.h
extern int getrlimit PARAMS ((int, struct rlimit *));
extern int setrlimit PARAMS ((int, const struct rlimit *));
vs.
extern int getrlimit __P ((enum __rlimit_resource __resource,
struct rlimit *__rlimits));
extern int setrlimit __P ((enum __rlimit_resource __resource,
__const struct rlimit *__rlimits));
Debugging configure it seems that my sys/resource.h includes
bits/resource.h and this causes the following warning from
gcc -E conftest.c
In file included from /usr/include/sys/resource.h:25,
from configure:2190:
/usr/include/bits/resource.h:109: warning: `RLIM_INFINITY' redefined
/usr/include/asm/resource.h:26: warning: this is the location of the previous definition
from this warning configure concludes that resource.h is missing.
I commented out #define RLIM_INFINITY ... from the asm/resource.h
and configure and make bootstrap proceeded normally.
Is this a gcc problem, a linux kernel problem, or a problem w/ my
header files (stock RedHat 6.1 IIRC), or maybe a fix-includes problem?
Regards,
Bill
PS. I haven't checked if this is fixed in 2.4.9-ac8 (it came out today).