[mpich-discuss] configure: different behaviour for bitypes.h on two machines
Siegmar Gross
Siegmar.Gross at informatik.fh-fulda.de
Thu Jan 29 06:18:33 CST 2009
Hi,
I have an old Cygwin-1.5 installation with X11R6 (machine "eiger")
and a new one with X11R7 (machine "wi23"). I was able to install
mpich2-1.0.8 with the following configure-command on the old machine:
configure --prefix=/usr/local/mpich2-1.0.8 \
CFLAGS="" CXXFLAGS="" FFLAGS="" LDFLAGS="" \
CPPFLAGS="" C_INCL_PATH="" C_INCLUDE_PATH="" CPLUS_INCLUDE_PATH="" \
OBJC_INCLUDE_PATH="" \
--with-thread-package=posix --with-mpe \
--enable-threads=multiple:global_mutex --enable-mpe \
--enable-f77 --enable-cxx --enable-romio \
|& tee log.configure.$SYSTEM_ENV.$MACHINE_ENV
Now I wanted to install mpich2-1.0.8 on the new machine with the
same configure-command and got an error in "make":
...
gcc -DHAVE_CONFIG_H -I. -I. -I. -I../../../include
-I/home/fd1026/MPI/MPICH2/mpich2-1.0.8/src/include -O2
...
mpidu_process_locks.c
In file included from ../../../include/mpiimpl.h:98,
from
/home/fd1026/MPI/MPICH2/mpich2-1.0.8/src/mpid/ch3/include/mpidimpl.h:22,
from mpidu_process_locks.h:20,
from mpidu_process_locks.c:6:
../../../include/mpitypedefs.h:17:25: sys/bitypes.h: No such file or directory
make[4]: *** [mpidu_process_locks.o] Error 1
...
The file "bitypes.h" exists neither on the old nor on the new machine.
On the new machine I could find the following output in file
"config.log".
...
configure:23487: checking for sys/bitypes.h
configure:23508: conftest.c
configure:23515: $? = 0
configure:23562: result: yes
...
wi23 mpich2-1.0.8 22 grep bitypes.h config.log
configure:23487: checking for sys/bitypes.h
ac_cv_header_sys_bitypes_h=yes
wi23 mpich2-1.0.8 23 grep bitypes.h log.configure.Cygwin.x86
checking for sys/bitypes.h... yes
checking sys/bitypes.h usability... no
checking sys/bitypes.h presence... no
checking for sys/bitypes.h... no
wi23 mpich2-1.0.8 24
wi23 mpich2-1.0.8 24 uname -a
CYGWIN_NT-5.1 wi23 1.5.25(0.156/4/2) 2008-06-12 19:34 i686 Cygwin
wi23 mpich2-1.0.8 25 gcc --version
gcc (GCC) 3.4.4 (cygming special, gdc 0.12, using dmd 0.125)
...
On the old machine I found the following outputs.
eiger mpich2-1.0.8-gcc 15 uname -a
CYGWIN_NT-5.1 eiger 1.5.25(0.156/4/2) 2008-06-12 19:34 i686 Cygwin
eiger mpich2-1.0.8-gcc 16 gcc --version
gcc (GCC) 3.4.4 (cygming special, gdc 0.12, using dmd 0.125)
...
eiger mpich2-1.0.8-gcc 13 grep bitypes.h config.log
configure:23487: checking for sys/bitypes.h
| #include <sys/bitypes.h>
configure:23534:25: sys/bitypes.h: No such file or directory
#include <sys/bitypes.h>
ac_cv_header_sys_bitypes_h=no
eiger mpich2-1.0.8-gcc 14 grep bitypes.h log.configure.Cygwin.x86
checking for sys/bitypes.h... no
checking sys/bitypes.h usability... no
checking sys/bitypes.h presence... no
checking for sys/bitypes.h... no
I found a lot more output in config.log on the old machine.
...
configure:23487: checking for sys/bitypes.h
configure:23508: conftest.c
./conftest.c: line 1: /bin: is a directory
./conftest.c: line 56: /bin: is a directory
configure:23515: $? = 126
configure: failed program was:
| /* confdefs.h. */
| #define PACKAGE_NAME ""
| #define PACKAGE_TARNAME ""
| #define PACKAGE_VERSION ""
| #define PACKAGE_STRING ""
| #define PACKAGE_BUGREPORT ""
| #define HAVE_ERROR_CHECKING MPID_ERROR_LEVEL_ALL
| #define MPICH_ERROR_MSG_LEVEL MPICH_ERROR_MSG_ALL
| #define USE_LOGGING MPID_LOGGING_NONE
| #define MPICH_THREAD_LEVEL MPI_THREAD_MULTIPLE
| #define USE_THREAD_IMPL MPICH_THREAD_IMPL_GLOBAL_MUTEX
| #define HAVE_ROMIO 1
| #define HAVE_LONG_LONG 1
| #define F77_NAME_LOWER_2USCORE 1
| #define STDCALL
| #define HAVE_FORTRAN_BINDING 1
| #define HAVE_CXX_EXCEPTIONS /**/
| #define HAVE_NAMESPACES /**/
| #define HAVE_NAMESPACE_STD /**/
| #define HAVE_CXX_BINDING 1
| #define HAVE_NAMEPUB_SERVICE 1
| #define restrict __restrict
| #define HAVE_GCC_ATTRIBUTE 1
| #define HAVE_SYS_TYPES_H 1
| #define HAVE_SYS_STAT_H 1
| #define HAVE_STDLIB_H 1
| #define HAVE_STRING_H 1
| #define HAVE_MEMORY_H 1
| #define HAVE_STRINGS_H 1
| #define HAVE_INTTYPES_H 1
| #define HAVE_STDINT_H 1
| #define HAVE_UNISTD_H 1
| #define WORDS_LITTLEENDIAN 1
| #define HAVE_LONG_DOUBLE 1
| #define HAVE_LONG_LONG_INT 1
| #define HAVE_MAX_INTEGER_ALIGNMENT 8
| #define HAVE_MAX_STRUCT_ALIGNMENT 8
| #define HAVE_MAX_FP_ALIGNMENT 4
| #define HAVE_MAX_DOUBLE_FP_ALIGNMENT 8
| #define HAVE_MAX_LONG_DOUBLE_FP_ALIGNMENT 4
| #define SIZEOF_CHAR 1
| #define SIZEOF_UNSIGNED_CHAR 1
| #define SIZEOF_SHORT 2
| #define SIZEOF_UNSIGNED_SHORT 2
| #define SIZEOF_INT 4
| #define SIZEOF_UNSIGNED_INT 4
| #define SIZEOF_LONG 4
| #define SIZEOF_UNSIGNED_LONG 4
| #define SIZEOF_LONG_LONG 8
| #define SIZEOF_UNSIGNED_LONG_LONG 8
| #define SIZEOF_FLOAT 4
| #define SIZEOF_DOUBLE 8
| #define SIZEOF_LONG_DOUBLE 12
| #define SIZEOF_WCHAR_T 2
| #define SIZEOF_VOID_P 4
| /* end confdefs.h. */
| #include <sys/bitypes.h>
configure:23539: gcc -c -O2 conftest.c 1>&5
configure:23534:25: sys/bitypes.h: No such file or directory
configure:23542: $? = 1
configure: failed program was:
#line 23533 "configure"
#include "confdefs.h"
#include <sys/bitypes.h>
int conftest() {return 0;}
configure:23571: result: no
...
As far as I can see both machines have the same Cygwin version and use
the same compiler. Does anybody know why "configure" behaves in such
a strange way? The only difference between both machines is that I have
installed a lot more packages on the old machine. Does anybody know if
"configure" needs a special package which I may have not installed on
the new machine? Thank you very much for any help in advance.
Siegmar
More information about the mpich-discuss
mailing list