20021030: netCDF on RH 8.0 (cont.)

>From: Chris Webster <address@hidden>
>Organization: NCAR Research Aviation Facility
>Keywords: 200210291753.g9THrAX19436 netCDF RedHat 8.0

Chris,

>At first it worked great, then I got to my normal make failed due to lack of
>fortran compiler.  So I "setenv FC", then distclean and .configure again.
>Fails again testing the g++/c++ compiler.

This is essentially the same test that Russ and I ran yesterday in
response to your earlier email.  I had no problems running configure
when using the following defines:

CC=/usr/bin/gcc
CPPFLAGS=-DNDEBUG -Df2cFortran
CFLAGS=-O
FC=/usr/bin/g77
FFLAGS=-O -Wuninitialized -fno-automatic
CXX=/usr/bin/g++

Here is the output from the configure I ran:

netcdf-3.5.1-beta5/src% ./configure
creating cache ./config.cache
checking for top-level source-directory
/home/mcidas/netcdf-3.5.1/netcdf-3.5.1-beta5/src
checking for m4 preprocessor
checking for m4... m4
checking m4 flags... -B10000
checking C compiler "/usr/bin/gcc"... works
checking how to make dependencies... false
checking for /usr/bin/g++... /usr/bin/g++
checking C++ compiler "/usr/bin/g++"... works
checking how to run the C preprocessor... /usr/bin/gcc -E
checking user-defined Fortran-77 compiler "/usr/bin/g77"... works
checking for Fortran .F compiler... 
checking if Fortran-77 compiler handles *.F files... yes
checking "/usr/bin/g77" as Fortran-90 compiler... failed to build test program
checking for xlf90... no
checking for f90... no
 ...

The listing shows that configure first tests for a C compiler, then for
a C++ compiler, and then for a Fortran compiler.  Given this ordering,
any change to FC should have no effect on the testing of g++.

Can you send us:

o the values of CC, CPPFLAGS, CFLAGS, FC, FFLAGS, and CXX that were defined
  when you ran configure

o the output from the configure run

o the contents of config.log

Please generate these after doing a 'make distclean' first.

Tom

>From: Chris Webster <address@hidden>
>Date: Wed, 30 Oct 2002 13:23:37 -0700
>Subject: Re: 20021029:netCDF on RH 8.0

Russ Rew wrote:

re:
>>At first it worked great, then I got to my normal make failed due to lack of
>>fortran compiler.  So I "setenv FC", then distclean and .configure again.
>>Fails again testing the g++/c++ compiler.

Stop.  I failed to 'cd' to the new 3.5.1 directory correctly, I had 
gone into 3.5.0.  All is well.

-- 
--Chris

>From address@hidden Fri Nov  1 09:38:07 2002
>Subject: Re: 20021030: netCDF on RH 8.0 (cont.)
> 
> Chris,
> 
> 
>>At first it worked great, then I got to my normal make failed due to lack of
>>fortran compiler.  So I "setenv FC", then distclean and .configure again.
>>Fails again testing the g++/c++ compiler.
> 
> 
> This is essentially the same test that Russ and I ran yesterday in
> response to your earlier email.  I had no problems running configure
> when using the following defines:

Not sure who got my response after the one above, but I had not 
correctly gotten into the 3.5.1beta5 directory, all is well.

-- 
--Chris

        There are 10 kinds of people in this world
        Those who think in binary and those who don't