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

[netCDF #RGF-193009]: "ncdump -c" bug identifying coordinate variables in groups still present in 4.1.2

Hi John,

> Long time no talk. netCDF-4 use for MAST fusion data advances steadily here - 
> currently 140MB of netCDF-4 data stored for each MAST shot, with 20+ shots on 
> a good day.
> The "ncdump -c" bug supposedly fixed in 4.1.2-beta1 is still around, though 
> the behaviour is different now. The attached cdl file provides a cut-down 
> example. Do "ncgen -khdf5 grouptest.cdl; ncdump -c grouptest.nc". It hangs.

I just tried the example you sent. ncdump from the current snapshot, which
will soon be released in netcdf-4.2rc1, works fine, returning the attached CDL
quickly.  I remember fixing this bug on November 14, from Jira ticket

If you have code newer than Nov 14 and the bug symptom is still manifesting,
I'd like to get a test case that demonstrates it.  You could also check that
this patch was made to ncdump/dumplib.c:

Index: dumplib.c
--- dumplib.c   (.../trunk/ncdump/dumplib.c)    (revision 1646)
+++ dumplib.c   (.../branches/russ/ncdump/dumplib.c)    (working copy)
@@ -1770,7 +1770,11 @@
     do {         /* be safe in case someone is currently adding
                   * dimensions */
+#ifdef USE_NETCDF4
+       NC_CHECK( nc_inq_dimids(ncid, &ndims, NULL, include_parents ) );
        NC_CHECK( nc_inq_ndims(ncid, &ndims) );
        if (dims)
        dims = (ncdim_t *) emalloc((ndims + 1) * sizeof(ncdim_t));


Russ Rew                                         UCAR Unidata Program
address@hidden                      http://www.unidata.ucar.edu

Ticket Details
Ticket ID: RGF-193009
Department: Support netCDF
Priority: Normal
Status: Closed

NOTE: All email exchanges with Unidata User Support are recorded in the Unidata inquiry tracking system and then made publicly available through the web. If you do not want to have your interactions made available in this way, you must let us know in each email you send to us.