Sourish, > Non-standard line lengths is not the only problem. I tried adding > '-ffree-line-length-none' to gfortran to force it to accept free form > input, but then I end up with a different set of errors during make check: > > test_write.o: In function `test_nf_set_fill_': > /home/i/Downloads/netcdf-4.1.3/nf_test/test_write.F:1172: undefined > reference to `nf_set_fill_' ... > /home/i/Downloads/netcdf-4.1.3/nf_test/test_write.F:1345: undefined > reference to `nf_delete_' > > and many similar undefined references from other files. I can't duplicate the problem you are seeing. Version 4.1.3 builds fine here with gfortran and the flag you're using to get around the long-line problem: make clean && CC=gcc FC=gfortran FFLAGS=-ffree-line-length-none CPPFLAGS=-I/share/ed/local/spike/include LDFLAGS=-L/share/ed/local/spike/lib ./configure && make -j check One difference might be that you are including szlib. We don't test with that library for reasons explained in section 2.2 of the installation guide: http://www.unidata.ucar.edu/netcdf/docs/netcdf-install.html#Quick-Instructions Do you really need szlib, knowing that netCDF-4 can't write szlib-encoded files? Another possibility might be that you just need to do a "make clean" and run the configure script again, before "make check". If you still can't get "make check" to run without errors, could you try it without szlib, to see if that is the source of the problems? If that's not convenient, could you try the current release candidate, version 4.2-rc1 with the separate Fortran release, version 4.2? --Russ > On 01/10/2012 12:09 AM, Unidata netCDF Support wrote: > > Sourish, > > > >> Ah good, thanks. I recompiled HDF5 with zlib, and now there's a new > >> error in netcdf make check: > >> > >> libtool: link: gfortran -I../fortran -I../f90 -I../f90 -I../include -g > >> -O2 -o .libs/f90tst_vars4 f90tst_vars4.o ../fortran/.libs/libnetcdff.so > >> /home/i/Downloads/netcdf-4.1.3/liblib/.libs/libnetcdf.so -lsz -lhdf5_hl > >> -lhdf5 -lm -lz > >> gfortran -I../fortran -I../f90 -I../f90 -I../include -g -O2 -c -o > >> f90tst_vars2.o f90tst_vars2.f90 > >> f90tst_vars2.f90:37.34: > >> > >> integer, parameter :: CACHE_SIZE = 8, CACHE_NELEMS = 571, > >> CACHE_PREEMPTION = > >> 1 > >> Error: Symbol 'cache_size' at (1) already has basic type of INTEGER > >> f90tst_vars2.f90:86.37: > >> > >> CACHE_NELEMS, CACHE_PREEMPTION)) > >> 1 > >> Error: Symbol 'cache_preemption' at (1) has no IMPLICIT type > >> f90tst_vars2.f90:77.13: > >> > >> call check(nf90_def_var(ncid, VAR5_NAME, NF90_INT, dimids, varid5, > >> chunksizes > >> 1 > >> Error: There is no specific function for the generic 'nf90_def_var' at (1) > >> make[2]: *** [f90tst_vars2.o] Error 1 > >> make[2]: Leaving directory `/home/i/Downloads/netcdf-4.1.3/nf_test' > >> make[1]: *** [check-am] Error 2 > >> make[1]: Leaving directory `/home/i/Downloads/netcdf-4.1.3/nf_test' > >> make: *** [check-recursive] Error 1 > > Hmm, you must have built with szip enabled, which builds a test program > > that's not usually built and that we didn't test on gfortran, > > f90tst_vars2.f90. It has some lines longer than gfortran handles by > > default. > > > > Could you download this newer version of that file, taken from the latest > > release of netcdf-fortran-4.2, and use it to replace > > nf_test/f90tst_vars2.f90 > > in your source code tree, then try "make check again" > > > > > > http://svn.unidata.ucar.edu/repos/netcdf-fortran/trunk/nf_test/f90tst_vars2.f90 > > > > If you run into more problems like this, I'd recommend using the separate > > Fortran > > release that's later than what's in netCDF-4.1.3, as announced here: > > > > > > http://www.unidata.ucar.edu/mailing_lists/archives/netcdfgroup/2011/msg00370.html > > > > Thanks. > > > > --Russ > > > >> On 01/08/2012 05:34 AM, Unidata netCDF Support wrote: > >>> This is telling you that the HDF5 library you have was not built with > >>> zlib, > >>> or else the ZLIB library is missing on your machine. > >>> > >>> =Dennis Heimbigner > >>> Unidata > >>> > >>> =Dennis Heimbigner > >>> Unidata > >>> > >>> > >>> Ticket Details > >>> =================== > >>> Ticket ID: ODR-209369 > >>> Department: Support netCDF > >>> Priority: Normal > >>> Status: Closed > >>> > >> > > Russ Rew UCAR Unidata Program > > address@hidden http://www.unidata.ucar.edu > > > > > > > > Ticket Details > > =================== > > Ticket ID: ODR-209369 > > Department: Support netCDF > > Priority: Normal > > Status: Closed > > > > Russ Rew UCAR Unidata Program address@hidden http://www.unidata.ucar.edu Ticket Details =================== Ticket ID: ODR-209369 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.