Re: ut_inv_calendar and f[:] => Segfault

Hi Mary,

I found the problem while trying to reduce the number of test data files. The
problem was a file which did not include the necessary variable. Actually, I
should have thought about this but the error behaviour was just to crazy.

So I guess you won't need test data anymore. Sorry for the confusion.

Sebastian

On Monday 03 May 2010 22:11:59 Mary Haley wrote:
> Hi Sebastian,
>
> Since I don't have your data files, I will try to reproduce this problem
> using dummy data. There's probably some weird memory hit here, and it
> could be in ut_inv_calendar or in the reading of the files.
>
> --Mary
>
> On May 3, 2010, at 11:32 AM, Sebastian Schubert wrote:
> > Dear all,
> >
> > I get a segmentation fault with the following script:
> >
> >
> > begin
> >
> > fils = systemfunc ("ls
> >
> > /home/schubert/scratch/clm/output/2003_0.0625/out01/lffd200308*.nc")
> >
> > f = addfiles (fils, "r")
> >
> > ; f =
> > addfile("/home/schubert/scratch/clm/output/2003_0.0625/out01/lffd20030825
> > 09.nc", "r")
> >
> > dalex = ut_inv_calendar(2003, 08, 1, 1,0,0,"seconds since 2003-08-01
> >
> > 00:00:00",0)
> > ; dalex = 1
> >
> > print(dalex)
> >
> > T = f[:]->T_S
> >
> > ; T = f->T_S
> >
> > printVarSummary(T)
> >
> > end
> >
> >
> > The files that are read are 246 netcdf files. The output is:
> >
> >
> > IPLEX - schubert@login01:~/plots> ncl test.ncl
> > ncl: /usr/lib64/libssl.so.0.9.8: no version information available
> > (required by ncl)
> > ncl: /usr/lib64/libcrypto.so.0.9.8: no version information available
> > (required by ncl)
> > Copyright (C) 1995-2010 - All Rights Reserved
> > University Corporation for Atmospheric Research
> > NCAR Command Language Version 5.2.0
> > The use of this software is governed by a License Agreement.
> > See http://www.ncl.ucar.edu/ for more details.
> >
> >
> > Variable: dalex
> > Type: double
> > Total Size: 8 bytes
> >
> > 1 values
> >
> > Number of Dimensions: 1
> > Dimensions and sizes: [1]
> > Coordinates:
> > Number Of Attributes: 2
> >
> > calendar : standard
> > units : seconds since 2003-08-01 00:00:00
> >
> > (0) 3600
> > Segmentation fault
> >
> >
> > No segfault when not getting T variable. No segfault when dalex not
> > defined or just the simple dalex=1; output for T then:
> >
> > Variable: T
> > Type: float
> > Total Size: 103935000 bytes
> >
> > 25983750 values
> >
> > Number of Dimensions: 3
> > Dimensions and sizes: [time | 246] x [rlat | 325] x [rlon | 325]
> >
> > Coordinates:
> > time: [ 0.. 0]
> > rlat: [-17..3.25]
> > rlon: [-12.5..7.75]
> >
> > Number Of Attributes: 4
> >
> > long_name : soil surface temperature
> > units : K
> > grid_mapping : rotated_pole
> > coordinates : lon lat
> >
> > Furthermore, and this is strange, no segfault when the order is changed:
> > first reading of T and then defining of dalex. Also, no segfault when
> > just reading one file.
> >
> > Bug?
> >
> > Cheers,
> > Sebastian

-- 
Potsdam Institute for Climate Impact Research (PIK)
Research Domain II - Climate Impacts & Vulnerabilities
http://www.pik-potsdam.de/members/schubert/