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

20000428: bug fix for SGI progs reading Linux written gem files

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 Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated.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: 20000428: bug fix for SGI progs reading Linux written gem files
  • Date: Fri, 28 Apr 2000 09:36:17 -0600
Pete,

I run on an SGI and read our grid files that are written from a Solaris x86
machine (whiuch will define the same MTMACH and byte order as Linux)
and have not seen your troubles- So we should probably make sure we are
comparing apples here and see what SGI OS and compilers you have.

I am running IRIX64 with:
--> uname -R
6.5 6.5.7m


--> showprods
I  c_dev                09/29/1998  C Headers and Libraries, 7.2.1
I  compiler_eoe         03/07/2000  IRIX Standard Execution Environment (Base
                                    Compiler Headers and Libraries, 7.2.1.3m)
I  ftn77_dev            09/29/1998  Fortran 77 Headers and Libraries, 7.2.1

The Makeinc.irix file defines -o32 in the compilation flags, just to
be sure you still have that defined too.

Steve Chiswell
Unidata User Support

On Thu, 27 Apr 2000, Pete Pokrandt wrote:

> 
> Steve,
> 
> We've been having a problem while trying to read gempak grid files
> written on an intel linux box, using gdinfo or gdplot or whatever on an
> SGI running Irix.  Both the writing and reading machines were using
> pl16.
> 
> The symptoms were that the SGI programs could read gempak files written
> on an SGI but not those written on a Linux box, whereas the linux gdinfo
> etc were able to read gempak files written on either an SGI or Linux
> box.
> 
> It turns out that the mvswp4 file defines the 1 byte arrays as type BYTE
> (just below the doc header in gempak5.4/source/gemlib/mv/mvswp4.f)
> 
> When I used this, for whatever reason, the routine did not run properly
> on the SGI. If I change the BYTE declaration to CHARACTER*1 instead, the
> routine runs right, and the sgi can read linux written gempak files.
> 
> I have not tested the CHARACTER definition on other platforms, but
> it works for sure on an SGI.
> 
> There are other instances of BYTE in the other routines in the mv
> directory (mvsw42, mvswp2), I wonder if they need to be changed
> as well?
> 
> Pete
> 
> 
> --
> +>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>+<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<+
> ^ Pete Pokrandt                    V 1447  AOSS Bldg  1225 W Dayton St^
> ^ Systems Programmer               V Madison,         WI     53706    ^
> ^                                  V      address@hidden       ^
> ^ Dept of Atmos & Oceanic Sciences V (608) 262-3086 (Phone/voicemail) ^
> ^ University of Wisconsin-Madison  V       262-0166 (Fax)             ^
> +<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<+>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>+
> 
>