[Wien] MIXER runtime error + solution on Mac OS X

[Laurence Marks]

Two comments:

a) The difference in energy between 12.1 and 13.1 is "nothing". The lower
value for 13.1 is due to slightly different linearization energies (lapw1)
and quadratures (lapw0) in the two version.  The difference in 13.1
versions on Mac and Linux is due to a stack of subtle numerical accuracy
issues, and won't change any physical results. I hope to presuade Peter to
impliment some changes to try and stabilize these as they effect the
convergence of MSR1a (they lead to noise), perhaps for release 14.2 (or
15.1).

b) The person who could repair quickly qmix8 is me -- but unfortunately I
don't have a clue what the issue is. Find a case where it crashes, then try
running "ddd mixer mixer.def" without the heap? (Maybe gdb, depends upon
the system.) If you can pin it down then a patch is plausible, but since I
have no Mac I cannot hope to reproduce the issue myself so I cannot help.
As I mentioned before, the issue may well be somewhere else in the code.


On Thu, Sep 4, 2014 at 2:05 PM, Kevin Jorissen <kevinjorissenpdx at gmail.com>
wrote:

>  Hi Martin,
>
>  thanks, that's reassuring that there shouldn't be a problem with the
> __APPLE__ statements.  It's indeed possible that the NaN's are a harmless
> unitialized field that will be initialized at the proper time -- I don't
> know.  Jianxin Zhu reported that issue and I merely passed it on.  I don't
> have much time this week or next, unfortunately.  My hope is that someone
> in the WIEN community can shed light on the behavior we observed.
>
>  Cheers,
>
>  Kevin
>
>
>
> On Wed, Sep 3, 2014 at 12:17 AM, Martin Kroeker <
> martin at ruby.chemie.uni-freiburg.de> wrote:
>
>> I would expect the __APPLE__ define to come in automatically from your
>> compiler and/or system headers.
>> The NaNs are "interesting" - could this be another case of uninitialized
>> array elements, just like the lapw0 bug you found last month ? Heap vs
>> stack would then just change the likelyhood of obtaining (non-)zeroed
>> memory on allocation...
>>
>> --
>> Dr. Martin Kroeker            martin at ruby.chemie.uni-freiburg.de
>> c/o Prof.Dr. Caroline Roehr
>> Institut fuer Anorganische und Analytische Chemie der Universitaet
>> Freiburg
>>
>> _______________________________________________
>> Wien mailing list
>> Wien at zeus.theochem.tuwien.ac.at
>> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>> SEARCH the MAILING-LIST at:
>> http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html
>>
>
>


-- 
Professor Laurence Marks
Department of Materials Science and Engineering
Northwestern University
www.numis.northwestern.edu
Corrosion in 4D: MURI4D.numis.northwestern.edu
Co-Editor, Acta Cryst A
"Research is to see what everybody else has seen, and to think what nobody
else has thought"
Albert Szent-Gyorgi
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20140904/f80feb2e/attachment.htm>