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

20040204: detecting proper incorporation of gdbm library

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 Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated.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: 20040204: detecting proper incorporation of gdbm library
  • Date: Wed, 04 Feb 2004 08:38:05 -0700
Neil,

>Date: 03 Feb 2004 18:36:54 -0600
>From: "Neil R. Smith" <address@hidden>
>Organization: Texas A&M University
>To: Steve Emmerson <address@hidden>
>Subject: Re: 20040203: detecting proper incorporation of gdbm library
>Keywords: 200401280025.i0S0PCp2028769 LDM gdbm

The above message contained the following:

> I spoke too soon. The DBFILE output is not understandable
> by the weather program. I forgot that I'd config'd weather
> to use only flat ascii files for metar data instead of
> gdbm, so I wasn't actually reading the pqsurf GDFILE output
> with the weather program. I've now corrected that and have
> done some sample executions of the weather program to extract
> reports from those DBFILE output files.
> 
> Currently our linux built weather will exit with error
> "gdbm fatal: read error". Our Irix built weather will 
> exit with "Bus error(coredump)" (the data volume is nfs-mounted
> on the linux and Irix hosts).
> 
> I ran 'ldd pqsurf' to see if it'd confirm that the correct libgdbm
> was used, but that only shows dynamically linked libraries:
> pqsurf:
>         libm.so.2 => /usr/lib/libm.so.2 (0x2807a000)
>         libc.so.4 => /usr/lib/libc.so.4 (0x28095000)
> 
> Is ours a static incorportation of the libgdbm.a in our build?

Unless you did something special with the makefiles, the GDBM library
should be statically-linked into the pqsurf(1) program.  The value of
the GDBMLIB makefile macro (possibly together with the contents of any
referenced directory) should tell you.

> Am I barking up the wrong tree here?

I don't know what's wrong.  You relinked the pqsurf(1) program; perhaps
it's time to relink the weather program.  Make sure that the same GDBM
library is used in both.

Regards,
Steve Emmerson