960328: netCDF 2.4.1 for Unicos 7.0.6

John,

>Date: Thu, 28 Mar 1996 13:42:54 -0500 (EST) 
>From: address@hidden (John Sheldon)
>Organization: NOAA/GFDL
>Subject: Re: 960328: netCDF 2.4.1 for Unicos 7.0.6 
>Keywords: 199603281505.AA27313

In the above message you wrote:

> Ok, done.  One side note about the log of the "make test" I sent you
> (and I hope it doesn't mean I've neglected to send you a clue you
> needed...) :  I was using the command "make test | tee make_test.log"
> so I could watch as things progressed, oblivious to the fact that stderr
> wasn't going to my log file.  SO...I redid the make (after your edits)
> and used ">&!" redirection to capture everything.  The result is below,
> and shows (in addtion to the new diagnostics) a couple of c89 Warnings
> that might mean something to you.

The compiler warning messages don't appear to be relevant.

> making `test' in directory 
> /ftmp/jps/tmpdir.045859a/unidata.ucar.edu/pub/netcdf/netcdf-2.4.1/src/libsrc
> 
>         c89 -c -O -DDEBUG xdrffio.c
> cc-101 c89: WARNING xdrffio.c, Line = 43, File = /usr/include/rpc/types.h, 
> Line = 19
>   The macro "FALSE" is redefined.
> cc-101 c89: WARNING xdrffio.c, Line = 43, File = /usr/include/rpc/types.h, 
> Line = 20
>   The macro "TRUE" is redefined.
> 
> TOTAL WARNINGS DETECTED IN xdrffio.c: 2
>         ar rcuv libnetcdf.a array.o attr.o cdf.o dim.o file.o iarray.o 
> error.o  globdef.o putget.o putgetg.o sharray.o string.o var.o  xdrffio.o
> r - xdrffio.o
>         : libnetcdf.a
> c89 -o cdftest -O cdftest.o libnetcdf.a
>         ./cdftest > cdftest_out.new
> NCxdrfile_create(): path="test.nc"
> NCxdrfile_create(): fmode=0x502
> NCxdrfile_create(): ControlString="bufa:336:2"
> ffopens(): Error 0
> nccreate: filename "test.nc"
> trying again
> NCxdrfile_create(): path="test.nc"
> NCxdrfile_create(): fmode=0x302
> NCxdrfile_create(): ControlString="bufa:336:2"
> ffopens(): Error 0

Great!  The ffopens() call fails but doesn't set `errno'.  We'll have to
add some more code.  Put the following lines just before the call to
perror() within the `if':

    fprintf(stderr, "ffopens(): stat.sw_error=%d\n",
                stat.sw_error);
    fprintf(stderr, "ffopens(): %s\n", strerror(stat.sw_error));

This should tell us why the ffopens() call is failing.

--------
Steve Emmerson   <address@hidden>