Re: Unidata Support: 970115: Problem report: netcdf-3.1a

> Date:    Wed, 15 Jan 1997 07:50:36 -0700
> From:    Unidata Support <address@hidden>
> To:      address@hidden
> Subject: 970115: Problem report: netcdf-3.1a
>
> >To: address@hidden
> >From: Harvey DAVIES <address@hidden>
> >Subject: Problem report: netcdf-3.1a
> >Organization: .
> >Keywords: 199701150927.CAA17119
>
> I have struck a problem with netcdf-3.1a on a Sun system.
> uname -a
> gives
> SunOS trout 4.1.3 2 sun4c
>
> The problem does not occur on SGI system.
>
> Following demonstrates the problem.
> File t.cdl is:
>
> netcdf t {
> dimensions:
>     a = UNLIMITED;
> variables:
>     double d(a);
> }
>
> File try.c is:
>
> #include "netcdf.h"
> #include <assert.h>
>
> main(int argc, char *argv[])
> {
>     int ncid;
>     int varid=0;
>     int status;
>     size_t start[] = {0};
>     size_t count[] = {1};
>     double value = 3.0;
>
>     status = nc_open( "t.nc", NC_WRITE, &ncid );
>     assert( !status );
>     status = nc_put_vara_double(ncid, varid, start, count, &value);
>     assert( !status );
>     status = nc_close(ncid);
>     assert( !status );
>     return 0;
> }
>
> Trying:
>
> gcc -g -I../../libsrc -I../../libsrc -target sun4 -c  try.c
> gcc -g try.o ../../../lib/libnetcdf.a  -o try
> ncgen -b t.cdl
> try
> putget.c:494: failed assertion `!(varp->ndims == 1 && ncp->recsize <=
varp->len)'
> *** Abort - core dumped
>
>
> Problem disappears if call to nc_put_vara_double is replaced by
>     status = nc_put_var1_double(ncid, varid, start, &value);
>
>
> Harvey
>
> Harvey Davies, CSIRO Mathematical and Information Sciences,
> 723 Swanston Street, Carlton, Victoria 3053, Australia
> Email: address@hidden
> Phone: +61 3 9282 2623 or +61 3 9239 4556
>   Fax: +61 3 9282 2600

I was working last week with  address@hidden via telephone on this.
The logic in the assertion in NCedgeck() is wrong and may be be deleted.
The fix to putget.m4 has been checked in.

-glenn