[Pw_forum] compile error on AIX 5.2 for Hitachi SR11000

[Axel Kohlmeyer]

On Wed, 25 Feb 2009, Dr. Shu-jun Hu wrote:

SH> 
SH> Dear All,
SH> 
SH> I want to compile the QE-4.0.3 code on the Hitachi SR11000 system, and however get
SH> the error information:

a word of caution:

are you really, really, _really_ sure you want to do that???
don't you have access to a decent linux cluster instead?

i can tell you from my experiences with CPMD that you are
in for a lot of trouble and getting the code compiled is
only the least of it.

the CPMD code is full of hacks and workarounds to get stuff
working correctly on these monsters. ...and - unlike CPMD - 
Q-E doesn't have a "sancto mauro" watching over it.

cheers,
   axel.

SH> 
SH> _________________________________________________________
SH> checking build system type... powerpc-ibm-aix5.2
SH> checking architecture... aix
SH> checking for Fortran 77 compiler default output file name... a.out
SH> checking whether the Fortran 77 compiler works... yes
SH> checking whether we are cross compiling... yes
SH> checking for suffix of executables...
SH> checking for suffix of object files... o
SH> checking whether we are using the GNU Fortran 77 compiler... no
SH> checking whether f90 accepts -g... yes
SH> checking for mpif90... mpif90
SH> checking whether we are using the GNU Fortran 77 compiler... no
SH> checking whether mpif90 accepts -g... yes
SH> setting F90... f90
SH> setting MPIF90... mpif90
SH> checking whether we are using the GNU C compiler... no
SH> checking whether mpCC_r accepts -g... yes
SH> checking for mpCC_r option to accept ANSI C... none needed
SH> setting CC... mpCC_r
SH> checking how to run the C preprocessor... mpCC_r -E
SH> checking for egrep... grep -E
SH> checking for ANSI C header files... yes
SH> checking for sys/types.h... yes
SH> checking for sys/stat.h... yes
SH> checking for stdlib.h... yes
SH> checking for string.h... yes
SH> checking for memory.h... yes
SH> checking for strings.h... yes
SH> checking for inttypes.h... yes
SH> checking for stdint.h... yes
SH> checking for unistd.h... yes
SH> checking for int *... yes
SH> checking size of int *... 4
SH> checking malloc.h usability... yes
SH> checking malloc.h presence... yes
SH> checking for malloc.h... yes
SH> checking for struct mallinfo.arena... yes
SH> checking for f90... f90
SH> checking whether we are using the GNU Fortran 77 compiler... no
SH> checking whether f90 accepts -g... yes
SH> setting F77... f90
SH> using F90... f90
SH> setting FFLAGS... -Os -64 -precexp=4 -nompi -noparallel -qfree=f90 -cpp
SH> setting F90FLAGS... $(FFLAGS)
SH> setting FFLAGS_NOOPT... -O0
SH> setting CFLAGS... -O3
SH> setting CPP... cpp
SH> setting CPPFLAGS... -P -traditional
SH> setting LD... mpif90
SH> setting LDFLAGS...
SH> setting AR... ar
SH> setting ARFLAGS... -X64 ruv
SH> setting ARFLAGS_DYNAMIC...-X64 ruv
SH> checking whether make sets $(MAKE)... yes
SH> checking whether Fortran files must be preprocessed... yes
SH> checking how to get verbose linking output from f90... -###
SH> checking for Fortran libraries of f90...  -L/opt/hitachi/matmpp/lib
SH> -L/opt/hitachi/MSL2/lib -l/tmp/.200902251913551638458F90/conftest/conftest.L
SH> -L/opt/ofort90/lib -lhf90vecmath -lhf90math -lf90 -lfiexpmpi
SH> checking for dummy main to link with Fortran libraries... unknown
SH> configure: error: linking to Fortran libraries from C fails
SH> See `config.log' for more details.
SH> _________________________________________________________
SH> 
SH> 
SH> As I trace the config.log, the following information tells me the souce of the
SH> error may be the link of libs:
SH> 
SH> _________________________________________________________
SH> configure:5050: mpCC_r -o conftest -g   conftest.c   -L/opt/hitachi/matmpp/lib
SH> -L/opt/hitachi/MSL2/lib -l/tmp/.200902251913551638458F90/conftest/conftest.L
SH> -L/opt/ofort90/lib -lhf90vecmath -lhf90math -lf90 -lfiexpmpi >&5
SH> ld: 0706-006 Cannot find or open library file: -l
SH> /tmp/.200902251913551638458F90/conftest/conftest.L
SH>         ld:open(): No such file or directory
SH> configure:5056: $? = 255
SH> configure: failed program was:
SH> ...
SH> _________________________________________________________
SH> 
SH> The path:
SH> 
SH>  -l/tmp/.200902251913551638458F90/conftest/conftest.L
SH> 
SH> is suspicious. How can I solve such problem?
SH> 
SH> Any reply is appreciated.
SH> 
SH> Best
SH> 
SH> Shujun
SH> 
SH> 
SH> 
SH> 
SH> 
SH> 
SH> 
SH> 
SH> 
SH> _______________________________________________
SH> Pw_forum mailing list
SH> Pw_forum at pwscf.org
SH> http://www.democritos.it/mailman/listinfo/pw_forum
SH> 

-- 
=======================================================================
Axel Kohlmeyer   akohlmey at cmm.chem.upenn.edu   http://www.cmm.upenn.edu
   Center for Molecular Modeling   --   University of Pennsylvania
Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
=======================================================================
If you make something idiot-proof, the universe creates a better idiot.