[Wien] gcc-gfortran compilation problem

[Peter Blaha]

First thing is to do a "sequential" installation. So forget about mpi 
first, but fix all problems due to the gfortran compiler.

The problems with
Error: Logicals at (1) must be compared with .eqv. instead of .eq.
have been mentioned before. Find in which lines and subroutines it 
occurs and fix it.

The other problems are probably form linking lapack/blas and you have to 
know where you installed them and how these libraries are called.

Once you installed and can run the sequential version, you can go on 
with the mpi-version. But note: it makes sense only if you have a larger 
cluster with infiniband network, or a "super-server" with shared memory 
and at least 8 cores).


On 01/21/2013 10:03 AM, Paweł Ficek wrote:
> Thank you very much for your answer. I tried to solve my problem in
> following way:
> -I installed latest version of openmpi
> -I installed latest versions of lapack, scalapack and their static libraries
> -As stated in openmpi FAQ I switched to mpi compilers wrapers (mpicc and
> mpif90)
> -I checked if mpif.h is present in the system
>
> The error messages at the end of compilation process are still the same:
>
> #######
>
> Compile time errors (if any) were:
> SRC_aim/compile.msg:make[1]: *** [aim] Error 1
> SRC_aim/compile.msg:make: *** [real] Error 2
> SRC_aim/compile.msg:make[1]: *** [aimc] Error 1
> SRC_aim/compile.msg:make: *** [complex] Error 2
> SRC_dipan/compile.msg:make: *** [dipan] Error 1
> SRC_hf/compile.msg:Error: Syntax error in array constructor at (1)
> SRC_hf/compile.msg:Error: Syntax error in array constructor at (1)
> SRC_hf/compile.msg:make[1]: *** [calc_exhfvv.o] Error 1
> SRC_hf/compile.msg:make: *** [real] Error 2
> SRC_hf/compile.msg:Error: Syntax error in array constructor at (1)
> SRC_hf/compile.msg:Error: Syntax error in array constructor at (1)
> SRC_hf/compile.msg:make[1]: *** [calc_exhfvv.o] Error 1
> SRC_hf/compile.msg:make: *** [complex] Error 2
> SRC_hf/compile.msg:modules_tmp_.F:182: Error: Can't open included file
> 'mpif.h'
> SRC_hf/compile.msg:make[1]: *** [modules.o] Error 1
> SRC_hf/compile.msg:make: *** [rp] Error 2
> SRC_hf/compile.msg:modules_tmp_.F:182: Error: Can't open included file
> 'mpif.h'
> SRC_hf/compile.msg:make[1]: *** [modules.o] Error 1
> SRC_hf/compile.msg:make: *** [cp] Error 2
> SRC_kgen/compile.msg:Error: Logicals at (1) must be compared with .eqv.
> instead of .eq.
> SRC_kgen/compile.msg:make: *** [arbmsh.o] Error 1
> SRC_lapw0/compile.msg:Error: Logicals at (1) must be compared with
> .neqv. instead of .ne.
> SRC_lapw0/compile.msg:Error: Logicals at (1) must be compared with .eqv.
> instead of .eq.
> SRC_lapw0/compile.msg:Error: Logicals at (1) must be compared with .eqv.
> instead of .eq.
> SRC_lapw0/compile.msg:make[1]: *** [lapw0.o] Error 1
> SRC_lapw0/compile.msg:make: *** [seq] Error 2
> SRC_lapw0/compile.msg:modules.F:22: Error: Can't open included file 'mpif.h'
> SRC_lapw0/compile.msg:make[1]: *** [modules.o] Error 1
> SRC_lapw0/compile.msg:make: *** [para] Error 2
> SRC_lapw1/compile.msg:make[1]: *** [lapw1] Error 1
> SRC_lapw1/compile.msg:make: *** [real] Error 2
> SRC_lapw1/compile.msg:make[1]: *** [lapw1c] Error 1
> SRC_lapw1/compile.msg:make: *** [complex] Error 2
> SRC_lapw1/compile.msg:modules_tmp_.F:33: Error: Can't open included file
> 'mpif.h'
> SRC_lapw1/compile.msg:make[1]: *** [modules.o] Error 1
> SRC_lapw1/compile.msg:make: *** [rp] Error 2
> SRC_lapw1/compile.msg:modules_tmp_.F:33: Error: Can't open included file
> 'mpif.h'
> SRC_lapw1/compile.msg:make[1]: *** [modules.o] Error 1
> SRC_lapw1/compile.msg:make: *** [cp] Error 2
> SRC_lapw2/compile.msg:make[1]: *** [lapw2] Error 1
> SRC_lapw2/compile.msg:make: *** [real] Error 2
> SRC_lapw2/compile.msg:make[1]: *** [lapw2c] Error 1
> SRC_lapw2/compile.msg:make: *** [complex] Error 2
> SRC_lapw2/compile.msg:modules_tmp_.F:47: Error: Can't open included file
> 'mpif.h'
> SRC_lapw2/compile.msg:make[1]: *** [modules.o] Error 1
> SRC_lapw2/compile.msg:make: *** [rp] Error 2
> SRC_lapw2/compile.msg:modules_tmp_.F:47: Error: Can't open included file
> 'mpif.h'
> SRC_lapw2/compile.msg:make[1]: *** [modules.o] Error 1
> SRC_lapw2/compile.msg:make: *** [cp] Error 2
> SRC_lapwdm/compile.msg:make[1]: *** [lapwdm] Error 1
> SRC_lapwdm/compile.msg:make: *** [real] Error 2
> SRC_lapwdm/compile.msg:make[1]: *** [lapwdmc] Error 1
> SRC_lapwdm/compile.msg:make: *** [complex] Error 2
> SRC_lapwso/compile.msg:make: *** [lapwso] Error 1
> SRC_mini/compile.msg:make: *** [mini] Error 1
> SRC_mixer/compile.msg:make: *** [mixer] Error 1
> SRC_pairhess/compile.msg:main.f:692:0: error: insn does not satisfy its
> constraints:
> SRC_pairhess/compile.msg:main.f:692:0: internal compiler error: in
> reload_cse_simplify_operands, at postreload.c:403
> SRC_pairhess/compile.msg:make: *** [main.o] Error 1
> SRC_qtl/compile.msg:make: *** [qtl] Error 1
>
> #####
>
> What I don't know how to do:
> -How to point WIEN to the correct mpi includes
> -How to point WIEN to the location of GotoBLAS2 library
> -How to specify lapack libraries
>
> I am very sorry that I am taking your time, but where it comes to the
> configuration of computational clusters I am just a beginner without
> much experience.
>
> Idea of using ifort is tempting but as I did not produce any results I
> must stick to the free alternatives.
>
>
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>

-- 

                                       P.Blaha
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-165300             FAX: +43-1-58801-165982
Email: blaha at theochem.tuwien.ac.at    WWW: 
http://info.tuwien.ac.at/theochem/
--------------------------------------------------------------------------