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

Re: 980109: netcdf3 interface (bug?)

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 Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated.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.

  • Subject: Re: 980109: netcdf3 interface (bug?)
  • Date: Mon, 26 Jan 98 09:21:03 -0700
Jim,

>Date: Fri, 23 Jan 1998 16:42:21 -0700 
>From: Jim Edwards <address@hidden>
>Organization: NOAA/FSL
>To: Steve Emmerson <address@hidden>
>Subject: Re: 980109: netcdf3 interface (bug?) 
>Keywords: 199801092007.NAA06889

In the above message, you wrote:

> I'm sorry, I should have stated this more clearly.  It's the f90
> compiler which does this.  I cannot use the f77 compiler for other
> reasons.

Then there's still some sort of problem because the netCDF Fortran API
supports both "integer*1" and "integer*2" types on our SunOS 5.5 system
with the f90 compiler:

    /tmp/netcdf-3.3.1/src: nm -g libsrc/libnetcdf.a | grep '_int[12]'
    [26]    |      3976|     512|FUNC |GLOB |0    |2      |nf_get_att_int1_
    [23]    |      5064|     512|FUNC |GLOB |0    |2      |nf_get_att_int2_
    [28]    |      3416|     540|FUNC |GLOB |0    |2      |nf_put_att_int1_
    [24]    |      4504|     540|FUNC |GLOB |0    |2      |nf_put_att_int2_
    [26]    |      2720|     108|FUNC |GLOB |0    |2      |nf_get_var_int1_
    [23]    |      2976|     108|FUNC |GLOB |0    |2      |nf_get_var_int2_
    [28]    |      2592|     108|FUNC |GLOB |0    |2      |nf_put_var_int1_
    [24]    |      2848|     108|FUNC |GLOB |0    |2      |nf_put_var_int2_
    [26]    |      2912|     172|FUNC |GLOB |0    |2      |nf_get_var1_int1_
    [23]    |      3296|     172|FUNC |GLOB |0    |2      |nf_get_var1_int2_
    [28]    |      2720|     172|FUNC |GLOB |0    |2      |nf_put_var1_int1_
    [24]    |      3104|     172|FUNC |GLOB |0    |2      |nf_put_var1_int2_
    [26]    |      3104|     236|FUNC |GLOB |0    |2      |nf_get_vara_int1_
    [23]    |      3616|     236|FUNC |GLOB |0    |2      |nf_get_vara_int2_
    [28]    |      2848|     236|FUNC |GLOB |0    |2      |nf_put_vara_int1_
    [24]    |      3360|     236|FUNC |GLOB |0    |2      |nf_put_vara_int2_
    [26]    |      3464|     360|FUNC |GLOB |0    |2      |nf_get_varm_int1_
    [23]    |      4216|     360|FUNC |GLOB |0    |2      |nf_get_varm_int2_
    [28]    |      3088|     360|FUNC |GLOB |0    |2      |nf_put_varm_int1_
    [24]    |      3840|     360|FUNC |GLOB |0    |2      |nf_put_varm_int2_
    [26]    |      3296|     300|FUNC |GLOB |0    |2      |nf_get_vars_int1_
    [23]    |      3936|     300|FUNC |GLOB |0    |2      |nf_get_vars_int2_
    [28]    |      2976|     300|FUNC |GLOB |0    |2      |nf_put_vars_int1_
    [24]    |      3616|     300|FUNC |GLOB |0    |2      |nf_put_vars_int2_

And the nf_test/nf_test routine tests for this support.

Look at the file "fortran/nfconfig.inc" in the netCDF source tree.  Does
it contain the following lines?

    #define NF_INT1_T integer*1
    #define NF_INT2_T integer*2

> f90: WorkShop Compilers 4.2 10/22/96 FORTRAN 90 1.2

Our Fortran 90 compiler looks like it's the same as yours:

    /tmp/netcdf-3.3.1/src: /opt/SUNWspro/bin/f90 -V
    f90: WorkShop Compilers 4.2 01 Dec 1997 FORTRAN 90 1.2 patch 104366-04
    Usage: f90 [ options ] files.  Use 'f90 -flags' for details

--------
Steve Emmerson   <http://www.unidata.ucar.edu>