Compilation error

44 views
Skip to first unread message

Kedar G. Bhide

unread,
Oct 21, 2016, 12:24:23 AM10/21/16
to pencil-code-discuss

I'm using latest version of PENCIL code from GitHub. It works fine on our workstation. When I try to compile it on supercomputing facility it is giving following error. Our workstation works on CentOS 6.3 where as supercomputing facility has CentOS 6.2. Previous version of PENCIL code worked fine on both the systems. I'm using mpif90 compiler.

mpif.h:549.16:
Included at mpicomm_double.f90:55:

   integer*8 MPI_DISPLACEMENT_CURRENT
            1

Error: Nonstandard type declaration INTEGER*8 at (1)
mpif.h:550.42:
Included at mpicomm_double.f90:55:

   PARAMETER (MPI_DISPLACEMENT_CURRENT=-54278278)
                                      1

Error: Symbol 'mpi_displacement_current' at (1) has no IMPLICIT type
make[2]: *** [mpicomm_double.o] Error 1
make[2]: Leaving directory /home/external/iitb/kedarb/pencil-code/samples/2d-te sts/2d_methane_flame/8_proc_parallel/src' make[1]: *** [code] Error 2 make[1]: Leaving directory/home/external/iitb/kedarb/pencil-code/samples/2d-te sts/2d_methane_flame/8_proc_parallel/src'
make: *** [default_to_be] Error 2
'make -j FFLAGS_DOUBLE=-fdefault-real-8 -fdefault-double-8 CFLAGS_DOUBLE=-DDOUBL E_PRECISION LD_MPI= CFLAGS_FFTW3= FFLAGS_FFTW3= LD_FFTW3= CFLAGS_FFTW2= FFLAGS_F FTW2= LD_FFTW2= FC=mpif90 F77=mpif77 FFLAGS=-O FSTD_95=-std=f95 FSTD_2003=-std=f 2003 CC=mpicc CFLAGS=-DFUNDERSC=1 default_to_be' failed: <>

Alex Richert

unread,
Oct 21, 2016, 12:27:14 AM10/21/16
to pencil-co...@googlegroups.com
I've run into something similar, I seem to recall fixing it by changing -std=f95 to something more recent (I think Fortran 08).
Alex
--
You received this message because you are subscribed to the Google Groups "pencil-code-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pencil-code-dis...@googlegroups.com.
To post to this group, send email to pencil-co...@googlegroups.com.
Visit this group at https://groups.google.com/group/pencil-code-discuss.
For more options, visit https://groups.google.com/d/optout.

kedar bhide

unread,
Oct 21, 2016, 1:12:21 AM10/21/16
to pencil-co...@googlegroups.com
My configuration file default.conf is like this. Still it is giving same error.

%section Makefile
        FC = mpif90
        FSTD_90=-std=f2008
        CC = mpicc
%endsection Makefile


On Fri, Oct 21, 2016 at 9:57 AM, Alex Richert <alexr...@gmail.com> wrote:
I've run into something similar, I seem to recall fixing it by changing -std=f95 to something more recent (I think Fortran 08).
Alex

On 10/21/2016 12:24 AM, Kedar G. Bhide wrote:

I'm using latest version of PENCIL code from GitHub. It works fine on our workstation. When I try to compile it on supercomputing facility it is giving following error. Our workstation works on CentOS 6.3 where as supercomputing facility has CentOS 6.2. Previous version of PENCIL code worked fine on both the systems. I'm using mpif90 compiler.

mpif.h:549.16:
Included at mpicomm_double.f90:55:

   integer*8 MPI_DISPLACEMENT_CURRENT
            1

Error: Nonstandard type declaration INTEGER*8 at (1)
mpif.h:550.42:
Included at mpicomm_double.f90:55:

   PARAMETER (MPI_DISPLACEMENT_CURRENT=-54278278)
                                      1

Error: Symbol 'mpi_displacement_current' at (1) has no IMPLICIT type
make[2]: *** [mpicomm_double.o] Error 1
make[2]: Leaving directory /home/external/iitb/kedarb/pencil-code/samples/2d-te sts/2d_methane_flame/8_proc_parallel/src' make[1]: *** [code] Error 2 make[1]: Leaving directory/home/external/iitb/kedarb/pencil-code/samples/2d-te sts/2d_methane_flame/8_proc_parallel/src'
make: *** [default_to_be] Error 2
'make -j FFLAGS_DOUBLE=-fdefault-real-8 -fdefault-double-8 CFLAGS_DOUBLE=-DDOUBL E_PRECISION LD_MPI= CFLAGS_FFTW3= FFLAGS_FFTW3= LD_FFTW3= CFLAGS_FFTW2= FFLAGS_F FTW2= LD_FFTW2= FC=mpif90 F77=mpif77 FFLAGS=-O FSTD_95=-std=f95 FSTD_2003=-std=f 2003 CC=mpicc CFLAGS=-DFUNDERSC=1 default_to_be' failed: <>

--
You received this message because you are subscribed to the Google Groups "pencil-code-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pencil-code-discuss+unsub...@googlegroups.com.
To post to this group, send email to pencil-code-discuss@googlegroups.com.

--
You received this message because you are subscribed to a topic in the Google Groups "pencil-code-discuss" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/pencil-code-discuss/tbP_vNZimmE/unsubscribe.
To unsubscribe from this group and all its topics, send an email to pencil-code-discuss+unsub...@googlegroups.com.
To post to this group, send email to pencil-code-discuss@googlegroups.com.



--
Kedar Govind Bhide

Philippe-A. Bourdin

unread,
Oct 21, 2016, 5:17:02 AM10/21/16
to pencil-code-discuss
Hello,


The error shown here arises from an MPI library that is not compatible with the Fortran standard (F95 or F2003, depending on which modules you use), as the error messages says:

   integer*8 MPI_DISPLACEMENT_CURRENT
            1

Error: Nonstandard type declaration INTEGER*8 at (1) 


Shifting the Fortran standard requirements to something even higher (F2008) can not work in this case. Instead, you should switch OFF the Fortran standard checking, which is simply done by setting:
FSTD=
in your configuration. Please never change the settings of "FSTD_*" because these are not meant to be changed.

Otherwise, it is always a good start to follow exactly the steps given in our "quick start guide" that is linked on our homepage, also if you want to submit a Bug report.

Thanks and best greetings,
Philippe.

Reply all
Reply to author
Forward
0 new messages