[molpro-user] Problem with spin-orbit calculation

[s13mc107 at mail.saitama-u.ac.jp]

Dear Molpro Users,

I'm having a serious problem.

When I performed spin-orbit calculation, got error message.  

--------------------------------------------------------------------
forrtl: severe (64): input conversion error, unit -5, file Internal 
Formatted Read
Image              PC                Routine            Line        
Source             
molpro.exe         000000000316AAEE  Unknown               Unknown  
Unknown
molpro.exe         0000000003169586  Unknown               Unknown  
Unknown
molpro.exe         00000000030F9960  Unknown               Unknown  
Unknown
molpro.exe         000000000308D82F  Unknown               Unknown  
Unknown
molpro.exe         000000000308CD37  Unknown               Unknown  
Unknown
molpro.exe         00000000030B3EEE  Unknown               Unknown  
Unknown
molpro.exe         00000000030B25E7  Unknown               Unknown  
Unknown
molpro.exe         0000000000C58A20  Unknown               Unknown  
Unknown
molpro.exe         0000000001E0948A  Unknown               Unknown  
Unknown
molpro.exe         0000000000944757  Unknown               Unknown  
Unknown
molpro.exe         0000000000943E7E  Unknown               Unknown  
Unknown
molpro.exe         0000000000A9BE09  Unknown               Unknown  
Unknown
molpro.exe         00000000009AF49D  Unknown               Unknown  
Unknown
molpro.exe         000000000040E769  Unknown               Unknown  
Unknown
molpro.exe         000000000040C8A4  Unknown               Unknown  
Unknown
molpro.exe         0000000000489F45  Unknown               Unknown  
Unknown
molpro.exe         0000000000409941  Unknown               Unknown  
Unknown
molpro.exe         00000000004098DC  Unknown               Unknown  
Unknown
libc.so.6          000000364621E576  Unknown               Unknown  
Unknown
molpro.exe         00000000004097D9  Unknown               Unknown  
Unknown

--------------------------------------------------------------------


Here I would like to show both of input and output files below
(log file is nothing).

---------------------------input file--------------------------------

***, spin-orbit coupling 
memory,1000,m
cartesian
basis=vdz
symmetry,x
geometry={Br1;
          H1 Br1 r1;
          Br2 H1 r2 Br1 a1;}

   r1=3.10577640 bohr
   r2=3.10571417 bohr
   a1=141.987501 ang

   {hf;maxit,500;wf,71,1,1}
   {multi;
    maxit,40;occ,27,10;closed,20,8;
    wf,71,1,1;state,2;
    wf,71,2,1;state,1}
   !dAp
   {ci;occ,27,10;closed,20,8;wf,71,1,1;state,1;save,4010.2}
   !dAdp
   {ci;occ,27,10;closed,20,8;wf,71,2,1;state,1;save,4020.2}
   !dAp2
   {ci;occ,27,10;closed,20,8;wf,71,1,1;state,2;save,4022.2}

   {ci;hlsmat,ls,4010.2,4020.2,4022.2;print,hls=0}

--------------------------end input file-----------------------------


--------------------------error output file--------------------------
1PROGRAM * CI (Multireference internally contracted CI)     Authors: H.-
J. Werner, P.J. Knowles, 1987

   ******************************
   *** Spin-orbit calculation ***
   ******************************


   Spin-orbit matrix elements
   ==========================
 
------------------------end error output file------------------------


So I edited Molpro/src/seward/sew_molpro.F.

Specifically I added codes 'subroutine seward_perm'
(labeled the arrows).  

---------------subroutine seward_perm(arround 620th line)------------

c.....loop over all SAOs
c     write(6,*) 'icart=',icart
      do 50 isk=1,nsk
      minj=nts(isk)+1
      do 20 i=nts(isk)+1,nts(isk)+nt(isk)
c     write(6,*) 'incoming lablist, i=',i,'  lablist=',lablist(i)
      fac=1.0d0
      label=lablist(i)
      do ll=0,mxlmom
        if(label(2:2).eq.anglab(ll)) then
          write(label(1:1),'(i1)') ll+1
          goto 200
        endif
      enddo
      errorstatus=0                                <=
      read(label,'(1x, 3i1)',err=51) ix,iy,iz      <=
      if ( errorstatus.eq.1 ) then                 <=
51      continue                                   <= 
      endif                                        <=
      ll=ix+iy+iz
      label(1:1)=anglab(ll)

---------------------------------------------------------------------

As a result, I didn't get error message.
Is this wrong?  


Molpro version is 2010.1 and patch level is 26.

Here is my machine spec.
CPU : Intel Core i7 3770S 3.1GHz 8MB 4Core 8Threads TDP65W
OS : Linux Fedora 17 (x86_64)
Fortrancompliler : Intel composer_xe_2011_sp1.7.256 version12.1.0

And compilekeywords are : -mpp -mppbase /opt/mpich2/include 
                          -auto-ga-tcgmsg-mpich2 -ifort -icc
                          -openmp -lapack -blas


Best regards.