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

20020318: McIDAS XCD uwgrid.sh script (cont.)

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 Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated.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: 20020318: McIDAS XCD uwgrid.sh script (cont.)
  • Date: Mon, 18 Mar 2002 11:07:23 -0700
>From: Eirh-Yu Hsie <address@hidden>
>Organization: Aeronomy Laboratory/NOAA/DOC
>Keywords: 200203151916.g2FJGga03699 McIDAS gtokserv INITBLOK g77

Hsie,

re: 

>>>Could you tell me what the error message means?
>>>
>>>Starting uwgrid.sh NF at 2002074/191421
>>>PRODUCT CODE=NF          2002074     191421
>>>GRDCOPY RTGRIDS/AVN MYDATA/GRIDS.9999 PARAM=T Z LEV=1000 850 700 500 400
>>>300 250 200 150 100 2 TRO DAY=2002074 GPRO=MERC TIME=12 FHOUR=0 NUM=ALL
>>>GRDCOPY: Grid token server unable to initialize McIDAS environment
>>>GRDCOPY - done
 ...
>>>uwgrid.k: Done....
>>>Ending uwgrid.sh NF at 2002074/191425

I originally offered the thought:

>It sounds like the McIDAS environment variables in uwgrid.sh are not
>set properly, but I would have anticipated a different error message
>than the one you show above.

After logging onto your Linux box rainbow, I discovered the cause of
the problem.  Several Fortran routines were declaring an argument to
the subroutine INITBLOK as INTEGER (which is INTEGER*4).  INITBLOK is
expecting an INTEGER*2, so the value being read by the calling routine
is accessing 2 bytes that are not set by INITBLOK.  For "standard"
Fortrans (i.e., the Fortran compilers from vendors), this does not
cause any problems.  g77, on the other hand, does not handle this
situation.

The solution for the problem was to go into all of the Fortran routines
that are calling INITBLOK with an INTEGER argument, and change the
argument to an INTEGER*2.  I made these changes on rainbow, and will
fold them into my next McIDAS addendum.

I would have thought that this should have been caught before now, but
several of the routines that had the mistake have been working with
no errors!

The upshot of all of this is 7 more routines that needed modification
to work correctly when built with g77...

Tom

>From address@hidden Mon Mar 18 11:24:55 2002
>Subject: Re: 20020318: McIDAS XCD uwgrid.sh script (cont.)

Hello:

Thanks, Tom.

Hsie
-------------------------
Eirh-Yu Hsie
Aeronomy Laboratory/NOAA
325 Broadway, R/AL4
Boulder, CO  80305-3328
voice:  303-497-3275
fax:    303-497-5373