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

[Support #TIB-359748]: "make check" fails in NetCDF4

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 Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated.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: [Support #TIB-359748]: "make check" fails in NetCDF4
  • Date: Mon, 25 Feb 2008 05:59:06 -0700
> Problem:  When compiling NetCDF4 snapshot2008021202 using Gnu C/C++ and
> Portland Group F95 in Redhat Enterprise Linux (v4), "make check" failed
> after an otherwise successful build.  I've attached the relevant output, but
> note that I solved the problem by changing
>
> #include <netcdf.h>
> to
> #include "netcdf.h"
>
> in every *.c file in the libsrc4 directory that contained a call to include
> the netcdf.h header file.
>
> The problem is that the GCC preprocessor considers the current directory (in
> this case libsrc4) to be a "system" directory with regard to header file
> search paths.  This means that the preprocessor removes any instances of
> "-I.", or "-I../libsrc4" from the search path, because it intends to search
> the current directory anyway, but only after all non-system, user-specified
> directories had been searched.
>
> The problem, as you might imagine, is that another copy of netcdf.h resides
> in the directory libsrc (note the lack of a "4"), which does not contain the
> necessary header information for compiling the files in libsrc4.  By
> changing the angle brackets to quotation marks, the GCC preprocessor no
> longer considers the copy of netcdf.h residing in libsrc4 to be a system
> header, so it gets found according to the rules you guys set using multiple
> "-I" options to GCC (namely, you place libsrc4 before libsrc in the header
> file search path).
>
> After implementing this change, all of "make check" passed with flying
> colors, and I seem to have a working installation of NetCDF4.  If you can
> think of a strong argument against this change, please reply.  Thanks!
>
> -Josh Rigler
>
>

Howdy Josh!

This is not the correct answer to the problem you are having.

If you want me to try and sort out the problem, I need the output of "make
check" for a clean build, without the changes in the code files.

Then we can figure out why it is not working for you...

THanks,

Ed

Ticket Details
===================
Ticket ID: TIB-359748
Department: Support netCDF
Priority: Normal
Status: Closed