[Wien] Highly confusing bug -- any ideas? (LD_ASSUME_KERNEL)

[L. D. Marks]

Good idea, I've seen this one; alas no. Even after "save -f -d XYZ ; rm 
case.* ; restore -f -d XYZ" it only does 1-2 cycles before the symptoms 
reappear.

On Mon, 16 Jan 2006, georg at chem.au.dk wrote:

> Just a suggestion:
> It might be that you have some old up/dn files and you are now running a
> non-magnetic calculation ?
>
> Georg
>
> Quoting "L. D. Marks" <L-marks at northwestern.edu>:
>
>> As a follow up to my own email, has anyone used LD_ASSUME_KERNEL and got
>> anywhere?
>>
>> On Mon, 16 Jan 2006, L. D. Marks wrote:
>>
>>> I have a bug which is crazy, albeit reproducable. For a relatively big
>>> calculation after a few scf cycles the clm's before mixer appear to become
>>> constant, i.e. are no longer changing at all. For instance, :FER becomes
>>> constant. I think it is linked to the presence of some output files, since
>> if
>>> I do "save -d XYZ ; rm case.* ; restore -d XYZ" it goes away for a cycle or
>>
>>> two. This is not anything simple, my current thinking is some version
>>> incompatibilities between the intel and system libraries, perhaps the C++
>>> I/O. I'd be interested if anyone has seen anything similar, or seen any
>>> compatibility issues between Red; I'm having a hard time tracking down
>> which
>>> program (lapw1, mixer?) is the culprit.
>>>
>>> System details:
>>> Linux version 2.6.9-22.ELsmp (buildcentos at louisa.home.local) (gcc version
>>> 3.4.4 20050721 (Red Hat 3.4.4-2)) #1 SMP Sat Oct 8 19:11:43 CDT 2005
>>> Rocks 4.1
>>> Intel(R) Fortran Compiler for 32-bit applications, Version 8.1    Build
>>> 20050207Z Package ID: l_fc_pc_8.1.024
>>> Intel(R) Cluster Math Kernel Library v7.2.1
>>>
>>> ldd /opt2/Wien2k_ifc8/lapw1 gives
>>>        libguide.so => /opt2/intel_fc_81/lib/libguide.so (0x4000c000)
>>>        libpthread.so.0 => /lib/tls/libpthread.so.0 (0x00dbc000)
>>>        libm.so.6 => /lib/tls/libm.so.6 (0x00cb8000)
>>>        libdl.so.2 => /lib/libdl.so.2 (0x00cdd000)
>>>        libc.so.6 => /lib/tls/libc.so.6 (0x00b8c000)
>>>        /lib/ld-linux.so.2 (0x00b73000)
>>>
>>> (N.B., LD_LIBRARY_PATH is not being passed correctly via "at", so there is
>> a
>>> file /etc/ld.so.conf.d/ldm.so.conf with
>>> /opt2/intel_fc_81/lib
>>> /opt2/intel_cc_81/lib
>>> /opt2/intel/mkl721cluster/lib/32
>>> )M I)
>>> Note: if you have an old email address for me, please note that "nwu" has
>>> been changed to "northwestern".
>>> -----------------------------------------------
>>> Laurence Marks
>>> Department of Materials Science and Engineering
>>> MSE Rm 2036 Cook Hall
>>> 2220 N Campus Drive
>>> Northwestern University
>>> Evanston, IL 60201, USA
>>> Tel: (847) 491-3996 Fax: (847) 491-7820
>>> email: L-marks at northwestern dot edu
>>> http://www.numis.northwestern.edu
>>> -----------------------------------------------
>>>
>>> _______________________________________________
>>> Wien mailing list
>>> Wien at zeus.theochem.tuwien.ac.at
>>> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>>>
>>
>> Note: if you have an old email address for me, please note that "nwu" has
>> been changed to "northwestern".
>> -----------------------------------------------
>> Laurence Marks
>> Department of Materials Science and Engineering
>> MSE Rm 2036 Cook Hall
>> 2220 N Campus Drive
>> Northwestern University
>> Evanston, IL 60201, USA
>> Tel: (847) 491-3996 Fax: (847) 491-7820
>> email: L-marks at northwestern dot edu
>> http://www.numis.northwestern.edu
>> -----------------------------------------------
>>
>> _______________________________________________
>> Wien mailing list
>> Wien at zeus.theochem.tuwien.ac.at
>> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>>
>
>
> -- 
> Georg Madsen
> Department of Chemistry
> Aarhus University
> DK-8000 Århus C
> Denmark
> tlf (+45) 89423885
> http://www.chem.au.dk/~webuorg/gm.html
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>

Note: if you have an old email address for me, please note that "nwu" has
been changed to "northwestern".
-----------------------------------------------
Laurence Marks
Department of Materials Science and Engineering
MSE Rm 2036 Cook Hall
2220 N Campus Drive
Northwestern University
Evanston, IL 60201, USA
Tel: (847) 491-3996 Fax: (847) 491-7820
email: L-marks at northwestern dot edu
http://www.numis.northwestern.edu
-----------------------------------------------