[Wien] bugs in sumpara in WIEN2k_05

[Peter Blaha]

I can confirm these problems and a fix will be on the web with the next 
release. (It should be a problem since the 100-atom fix was introduced).

However, I was not able to reproduce the L= 5389.. problem.

> Since I installed the WIEN2k_05, I have been suffering the Segmentation fault when I tried to do the LDA+U correction with SO coupling calculation with k-points parallelizations. Now I think I find a bug in subroutine "dmatscf.f" under directory SRC_sumpara. In "dmatscf.f", lines 50 to 52 are read as,
>  
>      IF(MARGN(2:4).EQ.'TRA') THEN
>         if(MARGN(19:22).EQ.'UPUP') ispin=1
>         if(MARGN(19:22).EQ.'UPDN') ispin=2
>         if(MARGN(19:22).EQ.'DNDN') ispin=3
>  
> and lines 61 to 70 are 
>  
>        ELSE IF(MARGN(2:4).EQ.'POM') THEN
>         ispin=2
>         if(MARGN(7:8)=='UP') ispin=1
>         READ(MARGN,'(4x,i3,43x,3f9.5)')IATOM, X,y,z
>         pom(IATOM,1,ispin,lorb)=pom(IATOM,1,ispin,lorb)+X
>         pom(IATOM,2,ispin,lorb)=pom(IATOM,2,ispin,lorb)+y
>         pom(IATOM,3,ispin,lorb)=pom(IATOM,3,ispin,lorb)+z
>         IF(ILOOP.EQ.IPROC)THEN
>            WRITE(22,723)IATOM,MARGN(7:8),(pom(IATOM,ityp,ispin,lorb),ityp=1,3)
>         ENDIF
> 
> However, if you take a look at the file "case.scfdmup_1", you can find that MARGN(19:22) should be MARGN(20:23) and MARGN(7:8) should be MARGN(8:9).
>  
> Yes, after I made these modifications the program works. However, I got something else in the dayfile, they are:
>  L= 538976288
>  L= 2
>  L= 2
>  L= 2
>  L= 2
>  L= 2
>  L= 2
> ...
> These lines appeared both after  "Summary of lapw2para" of "lapw2 -c -up -so -p" and "Summary of lapwdmpara" of "lapwdm -up -p -so -c".  For "L= 2" I can accept, that's the orbital which LDA+U correction is applied for, but what is "L= 538976288"?  This scares me! 
>  
> Looking forward to your reply!
>  
> Best,
>  
> Stargmoon
> 
> 		
> ---------------------------------
> Do you Yahoo!?
>  Take Yahoo! Mail with you! Get it on your mobile phone.


                                      P.Blaha
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-15671             FAX: +43-1-58801-15698
Email: blaha at theochem.tuwien.ac.at    WWW: http://info.tuwien.ac.at/theochem/
--------------------------------------------------------------------------