[Pw_forum] Fwd: LDA+U problem on Mac

Gabriele Sclauzero sclauzer at sissa.it
Wed Nov 4 09:46:50 CET 2009



> Da: Gabriele Sclauzero <gurlomail at gmail.com>
> Data: 04 novembre 2009 9:43:12 GMT+01:00
> A: PWSCF Forum <pw_forum at pwscf.org>
> Oggetto: Re: [Pw_forum] LDA+U problem on Mac
>
>
> Il giorno 04/nov/09, alle ore 04:04, Mulazzi Mattia ha scritto:
>
>>
>>>> About the LDA+U, I did more testing on it both on my PC (using
>>>> cygwin under WinXP) and on the Mac. The cygwin compilation
>>>> recognises without problems the V pseudopotential included in the  QE
>>>> library. The same calculations done on the Mac require first a
>>>> small  modification of the tabd.f90 file.
>>>
>>> Not only! As I said also set_hubbard_l.f90 needs to be modified
>>> accordingly. For the moment, you can take those two files from 4.1.1
>>> and put in 4.0.5 version (and recompile). They should work.
>>
>> I did not need to modify the set_hubbard_l.f90 file to have the LDA+U
>> running, but just the tab_d.f90.
>
> This sounds strange to me! If I remember well, there should be a   
> check in both subroutines, and the program should stop complaining   
> if either of the two subroutines has not been modified properly. Be   
> careful that you might not be obtaining correct results. You should   
> check that Hubbard_l=2 (it shuold be written at the beginning of   
> output) and starting local occupation on V is 3 (nsum=3 before the   
> first iteration).
>
>>> This is strange. Anyway keep in mind that the problem shows up  only if
>>> you put the +U on the single-character named species.
>>
>> I checked and there is also a segmentation fault when I use the MnS2
>> system, and for it also when I do a calculation with one Mn atom per
>> cell there is a segmentation fault. On the other hand a calculation
>> with two Mn atoms, Mn1 and Mn2, does not give a segmentation fault.  So
>> the problem is not in the atomic pseudopotential naming.
>
>
>>>
>>> Yes it is not a problem of the PP. If you still have problems, after
>>> replacing those two files of above, or using V1 in place of V,  please
>>> report it.
>>
>> I did a test after changing from V.pbe-n-van.UPF to V1.pbe-n-van.UPF,
>> but I still have the segmentation fault.
>
> Can you please send me some input files which give the crash   
> (together with details of compilation and QE version used)
>
>
>>
>>> HTH
>>>
>>> GS
>>> PS: After all, I'm still a bit surprised that this problem I  mentioned
>>> shows up as a segmentaion fault. Maybe it is something else...
>> At first I thought about a compiler bug (as suggested in the QE
>> troubleshooting pages), but it is quite unlikely it is the case. Why
>> should the compiler be buggy only with a one atom per cell system.
>
> I've always been told that g95 is not much reliable to do serious   
> calculations, since it may contain many bugs. Can you recompile with  
>  gfortran, it is free software  as well. You could also try with   
> ifort, but you need a license for Mac.
>
> Gabriele -- SISSA & Democritos
>
>>
>> Mattia
>>
>> FPR Fellow of RIKEN at Spring8
>> Excitation Order Research Team
>> 1-1-1 Sayo-cho Sayo-gun, Hyogo
>> Japan
>> _______________________________________________
>> Pw_forum mailing list
>> Pw_forum at pwscf.org
>> http://www.democritos.it/mailman/listinfo/pw_forum
>
>
>
> § Gabriele Sclauzero §
>     email: gurlomail at gmail.com
>     office: sclauzer at sissa.it
>               +39 040 3787511
>     skype: gurlonotturno
>



§ Gabriele Sclauzero §
      email: gurlomail at gmail.com
      office: sclauzer at sissa.it
                +39 040 3787511
      skype: gurlonotturno



----- Fine del messaggio inoltrato -----


----------------------------------------------------------------
   SISSA Webmail https://webmail.sissa.it/
   Powered by Horde http://www.horde.org/

-------------- next part --------------


Inizio messaggio inoltrato:

> Da: Gabriele Sclauzero <gurlomail at gmail.com>
> Data: 04 novembre 2009 9:43:12 GMT+01:00
> A: PWSCF Forum <pw_forum at pwscf.org>
> Oggetto: Re: [Pw_forum] LDA+U problem on Mac
>
>
> Il giorno 04/nov/09, alle ore 04:04, Mulazzi Mattia ha scritto:
>
>>
>>>> About the LDA+U, I did more testing on it both on my PC (using
>>>> cygwin under WinXP) and on the Mac. The cygwin compilation
>>>> recognises without problems the V pseudopotential included in the  
>>>> QE
>>>> library. The same calculations done on the Mac require first a
>>>> small  modification of the tabd.f90 file.
>>>
>>> Not only! As I said also set_hubbard_l.f90 needs to be modified
>>> accordingly. For the moment, you can take those two files from 4.1.1
>>> and put in 4.0.5 version (and recompile). They should work.
>>
>> I did not need to modify the set_hubbard_l.f90 file to have the LDA+U
>> running, but just the tab_d.f90.
>
> This sounds strange to me! If I remember well, there should be a  
> check in both subroutines, and the program should stop complaining  
> if either of the two subroutines has not been modified properly. Be  
> careful that you might not be obtaining correct results. You should  
> check that Hubbard_l=2 (it shuold be written at the beginning of  
> output) and starting local occupation on V is 3 (nsum=3 before the  
> first iteration).
>
>>> This is strange. Anyway keep in mind that the problem shows up  
>>> only if
>>> you put the +U on the single-character named species.
>>
>> I checked and there is also a segmentation fault when I use the MnS2
>> system, and for it also when I do a calculation with one Mn atom per
>> cell there is a segmentation fault. On the other hand a calculation
>> with two Mn atoms, Mn1 and Mn2, does not give a segmentation fault.  
>> So
>> the problem is not in the atomic pseudopotential naming.
>
>
>>>
>>> Yes it is not a problem of the PP. If you still have problems, after
>>> replacing those two files of above, or using V1 in place of V,  
>>> please
>>> report it.
>>
>> I did a test after changing from V.pbe-n-van.UPF to V1.pbe-n-van.UPF,
>> but I still have the segmentation fault.
>
> Can you please send me some input files which give the crash  
> (together with details of compilation and QE version used)
>
>
>>
>>> HTH
>>>
>>> GS
>>> PS: After all, I'm still a bit surprised that this problem I  
>>> mentioned
>>> shows up as a segmentaion fault. Maybe it is something else...
>> At first I thought about a compiler bug (as suggested in the QE
>> troubleshooting pages), but it is quite unlikely it is the case. Why
>> should the compiler be buggy only with a one atom per cell system.
>
> I've always been told that g95 is not much reliable to do serious  
> calculations, since it may contain many bugs. Can you recompile with  
> gfortran, it is free software  as well. You could also try with  
> ifort, but you need a license for Mac.
>
> Gabriele -- SISSA & Democritos
>
>>
>> Mattia
>>
>> FPR Fellow of RIKEN at Spring8
>> Excitation Order Research Team
>> 1-1-1 Sayo-cho Sayo-gun, Hyogo
>> Japan
>> _______________________________________________
>> Pw_forum mailing list
>> Pw_forum at pwscf.org
>> http://www.democritos.it/mailman/listinfo/pw_forum
>
>
>
> § Gabriele Sclauzero §
>     email: gurlomail at gmail.com
>     office: sclauzer at sissa.it
>               +39 040 3787511
>     skype: gurlonotturno
>



§ Gabriele Sclauzero §
     email: gurlomail at gmail.com
     office: sclauzer at sissa.it
               +39 040 3787511
     skype: gurlonotturno

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20091104/d564f3cd/attachment.html>


More information about the users mailing list