This archive contains answers to questions sent to Unidata support through mid-2025. Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated. For the most up-to-date information on the use of NSF Unidata software and data services, please consult the Software Documentation first.
Huisheng, >Date: Wed, 18 Feb 2004 16:01:59 -0500 >From: Huisheng Bian <address@hidden> >Organization: NASA/GSFC >To: Steve Emmerson <address@hidden> >Subject: Re: 20040218: error in 64-bit libnetcdf.a >Keywords: 200402171327.i1HDRirV003011 The above message contained the following: > According to the instruction of gcc(1), I change the setting like this: > setenv CC /app/gnu/bin/gcc > setenv CPPFLAGS -DNDEBUG > setenv CFLAGS '-O -mfp64 -mcpu=r4000 -mips3B' > setenv FC /bin/f90 > setenv FFLAGS '-O -64 -mips4' > setenv F90 /bin/f90 > setenv F90FLAGS '-O -64 -mips4' > setenv CXX /app/gnu/bin/g++ > setenv CXXFLAGS '-O -mfp64 -mcpu=r4000 -mips3' > However, conftest.o is still n32. > > Config.log is in attachment. > > Thanks, > Huisheng ... The "config.log" file shows this problem: configure:1933: checking if Fortran "integer*1" is C "signed char" configure:1942: /app/gnu/bin/gcc -c -DNDEBUG -O -mfp64 -mcpu=r4000 -mips3 conftest.c configure:1944: /bin/f90 -O -64 -mips4 -c conftestf.f configure:1946: /bin/f90 -o conftest -O -64 -mips4 conftestf.o conftest.o ld64: FATAL 12 : Expecting n64 objects: conftest.o is n32. Apparently, the options you gave the gcc(1) compiler still didn't produce an n64 object module. You'll have to do one of the following: 1. Discover what gcc(1) options will produce an n64 object; or 2. Use another C compiler that WILL generate n64 objects; or 3. Give up on generating a 64-bit netCDF library. Regards, Steve Emmerson