VMD-L Mailing List

From: Vlad Cojocaru (vlad.cojocaru_at_mpi-muenster.mpg.de)
Date: Tue Apr 07 2020 - 13:11:39 CDT

Hi John,

I was wondering whether you received my example trajectory for tackling
this issue ... I sent it some time ago via cryptshare.
The issue is not urgent at all, so no worries .. I am just asking to see
if you received those files (I believe they will expire at some point)...

Stay safe
Vlad

On 2020-01-06 21:20, John Stone wrote:
> Hi,
> I haven't seen any further discussion of this issue since December 19th.
>
> If there's any interest in actually resolving the problem, I will need
> an example file that I can use to reproduce the behavior that Vlad
> has reported. Vlad, can you (or anyone else) provide one?
>
> Best regards,
> John Stone
> vmd_at_ks.uiuc.edu
>
> On Thu, Dec 19, 2019 at 08:22:10AM -0600, John Stone wrote:
>> Hi,
>> Newer VMD 1.9.4 alpha builds include HDF5, whereas older revs did not.
>>
>> It seems possible that we're observing an issue where the HDF5 version of
>> NetCDF (or some shared internal routines used therein) is being picked
>> up rather than the old standalone NetCDF I compiled from source,
>> which is what was used for all prior builds.
>>
>> Vlad, I'm about to be offline for 2.5 days due to travel without network
>> access, but once I'm back in civilization, if you can post an affected
>> NetCDF file(s) that I can use for testing, I should be able to chase down
>> if this is caused by linking with HDF5 libs or if something less
>> obvious might be happening.
>>
>> Best,
>> John
>>
>> On Thu, Dec 19, 2019 at 02:09:50PM +0100, Luthaf wrote:
>>> Hi Vlad,
>>>
>>> What you are seeing is an assertion inside the netcdf library: https://github.com/Unidata/netcdf-c/blob/3acf69c6a01c9964532d85d9ec8a22acd6ca2488/libsrc/putget.m4#L821
>>>
>>> So either the the netcdf **library** version changed, or is was
>>> compiled without assertions (-DNDEBUG) for 1.9.4a12, and with
>>> assertions for 1.9.4a38.
>>>
>>> Anyway, hitting an assertion in the netcdf library might indicate
>>> that the files you have are invalid netcdf files. Are you able to
>>> open them outside of vmd? For example, using `ncdump file.nc`?
>>>
>>> Best,
>>> Guillaume
>>>
>>> Vlad Cojocaru a écrit le 19.12.19 à 10:46 :
>>>> Well, the only thing I can say at this point is that vmd 1.9.4a12
>>>> reads the trajectories perfectly, whereas vmd 1.9.4a38 does not
>>>> and gives the error I reported. First I thought its one trajectory
>>>> that is corrupted but then I checked it with other programs and it
>>>> was fine. After that I realized that the a38 version of vmd does
>>>> not ready any other netcdf trajectory I have ...
>>>>
>>>> Except the VMD version, I have not changed anything to the system ...
>>>>
>>>> I will look further into it and see if I can figure out what is
>>>> happening.
>>>>
>>>> Best,
>>>> Vlad
>>>>
>>>> On 12/18/19 5:19 PM, John Stone wrote:
>>>>> Vlad,
>>>>>   The only change to the NetCDF plugin (since 2016 actually) has been
>>>>> to correct a memory leak for a cell angles unit string in the AMBER
>>>>> reader code, nothing that would impact the eror you reported:
>>>>>
>>>>> @@ -133,6 +133,9 @@
>>>>>     if (cdf->amber.cell_lengths_units)
>>>>>       free(cdf->amber.cell_lengths_units);
>>>>> +  if (cdf->amber.cell_angles_units)
>>>>> +    free(cdf->amber.cell_angles_units);
>>>>> +
>>>>>     /* MMTK stuff */
>>>>>     if (cdf->mmtk.comment)
>>>>>       free(cdf->mmtk.comment);
>>>>> @@ -1121,7 +1124,7 @@
>>>>>     plugin.prettyname = "NetCDF (AMBER, MMTK)";
>>>>>     plugin.author = "Konrad Hinsen, John Stone";
>>>>>     plugin.majorv = 1;
>>>>> -  plugin.minorv = 1;
>>>>> +  plugin.minorv = 2;
>>>>>     plugin.is_reentrant = VMDPLUGIN_THREADSAFE;
>>>>>     plugin.filename_extension = "nc,ncrst";
>>>>>     plugin.open_file_read = open_cdf_read;
>>>>>
>>>>>
>>>>> I expect that the problem you're seeing is something deeper in
>>>>> the NetCDF library itself.
>>>>>
>>>>> Best,
>>>>>    John Stone
>>>>>
>>>>> On Wed, Dec 18, 2019 at 02:25:11PM +0100, Vlad Cojocaru wrote:
>>>>>> Dear all,
>>>>>>
>>>>>> Did VMD change the netcdf plugin between version 1.9.4a12 and
>>>>>> 1.9.4a38 ?
>>>>>> I am using netcdf trajectories created with CPPTRAJ program from
>>>>>> AmberTools 19.
>>>>>>
>>>>>> When I load these trajectories in VMD 1.9.4a12, everything
>>>>>> works well ....
>>>>>>
>>>>>> However, with VMD 1.9.4a38, I cannot load exactly the same
>>>>>> trajectories loaded correctly with the a12 version. I am getting the
>>>>>> error below. I will revert for now to the a12 version, but maybe
>>>>>> this is worth investigating.
>>>>>>
>>>>>> Best wishes
>>>>>> Vlad
>>>>>>
>>>>>> netcdfplugin) conventions: 'AMBER'
>>>>>> netcdfplugin) trajectory follows AMBER conventions version '1.0'
>>>>>> netcdfplugin) AMBER: program 'cpptraj'
>>>>>> netcdfplugin) AMBER: program version 'V18.01'
>>>>>> netcdfplugin) AMBER: title 'Cpptraj Generated trajectory'
>>>>>> netcdfplugin) AMBER: application 'AMBER'
>>>>>> netcdfplugin) AMBER: spatial dimension: 3
>>>>>> netcdfplugin) AMBER: atom dimension: 13042
>>>>>> netcdfplugin) AMBER: frame dimension: 80000
>>>>>> netcdfplugin) AMBER: coordinates units: 'angstrom'
>>>>>> netcdfplugin) AMBER: no coordinates scalefactor attribute, 1.0 assumed
>>>>>> netcdfplugin) AMBER: coordinates scalefactor: 1.000000
>>>>>> netcdfplugin) AMBER trajectory contains periodic cell information
>>>>>> netcdfplugin) AMBER: cell lengths units: 'angstrom'
>>>>>> netcdfplugin) AMBER: no cell lengths scalefactor attribute, 1.0 assumed
>>>>>> netcdfplugin) AMBER: cell lengths scalefactor: 1.000000
>>>>>> netcdfplugin) AMBER: cell angles units: 'degree'
>>>>>> netcdfplugin) AMBER: no cell angles scalefactor attribute, 1.0 assumed
>>>>>> netcdfplugin) AMBER: cell angles scalefactor: 1.000000
>>>>>> Info) Using plugin netcdf for coordinates from file ../test.ncdf
>>>>>> vmd_LINUXAMD64: putget.c:3593: getNCvx_float_float: Assertion `value
>>>>>> != ((void *)0)' failed.
>>>>>> Abort (core dumped)
>>>>>>
>>>>>> --
>>>>>> Vlad Cojocaru, PD (Habil.), Ph.D.
>>>>>> -----------------------------------------------
>>>>>> Project Group Leader
>>>>>> Department of Cell and Developmental Biology
>>>>>> Max Planck Institute for Molecular Biomedicine
>>>>>> Röntgenstrasse 20, 48149 Münster, Germany
>>>>>> -----------------------------------------------
>>>>>> Tel: +49-251-70365-324; Fax: +49-251-70365-399
>>>>>> Email: vlad.cojocaru[at]mpi-muenster.mpg.de
>>>>>> http://www.mpi-muenster.mpg.de/43241/cojocaru
>> --
>> NIH Center for Macromolecular Modeling and Bioinformatics
>> Beckman Institute for Advanced Science and Technology
>> University of Illinois, 405 N. Mathews Ave, Urbana, IL 61801
>> http://www.ks.uiuc.edu/~johns/ Phone: 217-244-3349
>> http://www.ks.uiuc.edu/Research/vmd/ 

-- 
Vlad Cojocaru, PD (Habil.), Ph.D.
-----------------------------------------------
Project Group Leader
Department of Cell and Developmental Biology
Max Planck Institute for Molecular Biomedicine
Röntgenstrasse 20, 48149 Münster, Germany
-----------------------------------------------
Tel: +49-251-70365-324; Fax: +49-251-70365-399
Email: vlad.cojocaru[at]mpi-muenster.mpg.de
http://www.mpi-muenster.mpg.de/43241/cojocaru