[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[netCDF #FUW-295578]: netcdf Make check error



Alex,

> Thank you for your help with the previous 2 problems. I sorry, but I've
> managed to have another issue now. As before, I have successfully installed
> the netcdf C library (4.3.1) already, and now I'm trying to install the
> fortran library (4.2). I have exported the env variable $LD_LIBRARY_PATH
> that points to the lib directory of the netcdf C installation as you
> suggested, which allowed the `configure` step to complete without errors.
> 
> However, after I call `configure` on the fortran library as below:
> 
> CC=$HOME/apps/gcc/gcc/bin/gcc CXX=$HOME/apps/gcc/gcc/bin/g++
> FC=$HOME/apps/gcc/gcc/bin/gfortran
> CPPFLAGS=-I$HOME/apps/netcdf/4.3.1/include
> LDFLAGS=-L$HOME/apps/netcdf/4.3.1/lib ./configure
> --prefix=$HOME/apps/netcdf/4.3.1
> 
> I then get the following error when I call `make check`:
> 
> libtool: compile:  /home/fispalma25/apps/gcc/gcc/bin/gcc -DHAVE_CONFIG_H
> -I. -I.. -DDLL_EXPORT -I/home/fispalma25/apps/netcdf/4.3.1/include -g -O2
> -MT fort-attio.lo -MD -MP -MF .deps/fort-attio.Tpo -c fort-attio.c  -fPIC
> -DPIC -o .libs/fort-attio.o
> In file included from fort-attio.c:8:0:
> *ncfortran.h:24:42: fatal error: NF_INT_IS_C_... not defined: No such file
> or directory*
> * #   include "NF_INT_IS_C_... not defined"*
> *                                          ^*
> compilation terminated.
> 
> Once again, I'd be grateful for any help you could give me. I've never had
> these problems installing netcdf before. I suppose I have overlooked some
> detail again. The config.log and fortran/nfconfig.in, fortran/nfconfig.inc
> files are attached if they help.

I've seen this problem before in config.log, here's the description:

  http://www.unidata.ucar.edu/support/help/MailArchives/netcdf/msg11949.html

The solution explained there is:

   You either need to rebuild your gfortran configured with 
   --disable-libquadmath (which means it won't support REAL(16) 
   or COMPLEX(16) types, not required by the Fortran standards 
   as far as I know, or you'll have to install libquadmath to 
   make such support available.

You might find it easier to just install a gcc and gfortran
compiler pair that are known to have been built compatibly.
There is no need to have the very latest gcc/gfortran.  I'd
recommend just getting the compilers from a package management
system that makes binaries available, such as "yum".

--Russ

> address@hidden> wrote:
> 
> > Hi Alex,
> >
> > > Thanks for the help. I recompiled hdf5 with zlib enabled and now the all
> > > the tests for the netcdf c library have been passed, and I installed
> > > successfully.
> > >
> > > However, now I am not able to successfully install the fortran
> > > (netcdf-fortran-4.2) library. I get the following error:
> > >
> > > checking size of off_t... configure: error: in
> > > `/home/fispalma25/apps/netcdf/netcdf-fortran-4.2':
> > > configure: error: cannot compute sizeof (off_t)
> > > See `config.log' for more details
> > >
> > > I have attached `config.log` in case that helps. I would appreciate any
> > > advice you can provide.
> >
> > Since you installed the netCDF C library as a shared library in a location
> > that is
> > not searched by default, you need to set the LD_LIBRARY_PATH environment
> > variable (or the equivalent DYLD_LIBRARY_PATH for OSX) , as described here:
> >
> >   http://www.unidata.ucar.edu/netcdf/docs/netcdf-fortran-install.html
> >
> > This is in addition to the LDFLAGS you are giving the compiler, because the
> > LD_LIBRARY_PATH is used by the loader to find shared libraries.
> >
> > --Russ
> >
> > > Cheers,
> > > Alex
> > >
> > >
> > >
> > >
> > >
> > > address@hidden> wrote:
> > >
> > > > On 1/23/14, 9:27 AM, Alex Robinson wrote:
> > > > > New Ticket: netcdf Make check error
> > > > >
> > > > > This test failed during `Make check`:
> > > > >
> > > > > *** Checking HDF5 variable functions.
> > > > > *** Checking HDF5 boolen variables...ok.
> > > > > *** Checking HDF5 variable with unlimited dimension...ok.
> > > > > *** Checking HDF5 deflate filter setting and getting...Sorry!
> > Unexpected
> > > > > result, tst_h_vars.c, line: 230
> > > > > FAIL: tst_h_vars
> > > > >
> > > > > I am installing netCDF version 4.3.1 with HDF5 version 1.8.12.
> > > > >
> > > > > Am I doing something wrong?
> > > > >
> > > > > Thanks,
> > > > > Alex
> > > > >
> > > > >
> > > > Frequently this is caused by an HDF5 install that was not compiled
> > > > with zlib compression enabled.  It may also be caused if HDF5 was
> > > > compiled with SZIP support but if netCDF was not.  If you provide
> > > > the config.log file, I will be happy to take a look, but the first step
> > > > will be to ensure that you are using an HDF5 compiled with zlib
> > > > support.
> > > >
> > > > Thanks!
> > > >
> > > > -Ward
> > > >
> > > >
> > > >
> > > > Ticket Details
> > > > ===================
> > > > Ticket ID: FUW-295578
> > > > Department: Support netCDF
> > > > Priority: Normal
> > > > Status: Open
> > > >
> > > >
> > >
> > >
> > > --
> > > Alexander Robinson
> > > Dpto. AstrofÃsica y CC de la AtmÃsfera
> > > Facultad de CC FÃsicas
> > > Universidad Complutense de Madrid
> > > 28040 Madrid, Spain
> > > ---------------------------------
> > > Tel (Fax): +34 91 394 4457 (4635)
> > >
> > >
> > Russ Rew                                         UCAR Unidata Program
> > address@hidden                      http://www.unidata.ucar.edu
> >
> >
> >
> > Ticket Details
> > ===================
> > Ticket ID: FUW-295578
> > Department: Support netCDF
> > Priority: Normal
> > Status: Closed
> >
> >
> 
> 
> --
> Alexander Robinson
> Dpto. AstrofÃsica y CC de la AtmÃsfera
> Facultad de CC FÃsicas
> Universidad Complutense de Madrid
> 28040 Madrid, Spain
> ---------------------------------
> Tel (Fax): +34 91 394 4457 (4635)
> 
> 
Russ Rew                                         UCAR Unidata Program
address@hidden                      http://www.unidata.ucar.edu



Ticket Details
===================
Ticket ID: FUW-295578
Department: Support netCDF
Priority: Normal
Status: Closed


NOTE: All email exchanges with Unidata User Support are recorded in the Unidata inquiry tracking system and then made publicly available through the web. If you do not want to have your interactions made available in this way, you must let us know in each email you send to us.