[Wien] kgen and lapw1 on intel machines

[gpzhang]

Dear Peter, WIEN users and other developers,

I have some troubles running kgen and lapw1 on intel computers though
the installation is fine. I hope this is not a bug in lapw1 (also
kgen) but increasingly I feel this is something you may be interested
in (same for w2k_3 and w2k_4.7) (I have not seen anything similar
reported before, but maybe I miss something here). It would be
wonderful if you would like to log into my computer and check it
yourself. If so, please simply send me an e-mail (private one though).


(0) computer compiler and options.
library:  atlas3.6.0.tgz + lapack.tgz
two computers: single PC with (a) intel m (512 MB, 1.3GZ) and  (b) intel p4 
(2GB)
compiler: ifc
options:
   Current settings:
     O   Compiler options:        -O -FR -mp -w
     L   Linker Flags:            -L/home/gpzhang/first-principles/LIB -Vaxlib
     P   Preprocessor flags       '-DParallel'
     R   R_LIB (LAPACK+BLAS):     -llapack -lf77blas -lcblas -latlas

(1) What I found.

  (a) in the subroutine SRC_lapw1/dscgst.f, a variable called
      PANELB( N, * ) causes the problem.

      for a small case (meaning small matrix) or anthlon, it works fine (maybe 
those processors
      get round of this problem erroneously since it produces wrong results)

  (b) in *.dayfile or *.output1, it gives me the following message

     lapw1 error
     (not surprising though, see more below)

(2) What I did.
    Insert the following code in dscgst.f

      do i=1,n-1
         write(1,*)panelb(i+1,i),i,N
      enddo

    before  LDPANELB = N
    This will cause it to crash easier and earlier.


(3) Debug

    I turn on the debug switch and find the error messages (see two
attachments for two different cases on two different systems, quite
big files, sorry!). What makes things worse is that I found several
other routines (kgen in particular) also have problems and stop (they do not
warn me when I turn off debug, klist also looks fine, things run fine until 
lapw1)

Array bounds exceeded in 
(a) tetcnt Line  104
(b) Subroutine arbmsh  Line  154, 
(c) MAIN PROGRAM (kgen)  Line  224
(d) Subroutine ssico  Line  79
(e) Subroutine xcpot1  Line  152
(f) MAIN PROGRAM (lapw0)   Line  1015



(4) I would really appreciate any suggestions and comments that you can give 
me.


   If there is anything unclear, please let me know.

Many thanks in advance!

(There are two attached files if you are interested in some details)

Sincerely,

Guoping

PS: Peter, your example of TiC uses a wrong lattice constant (both the
userguide and the actual run are incorrect) 
Would you please  give us a corrected one? Thanks!

-------------- next part --------------
A non-text attachment was scrubbed...
Name: q1.kgen.error
Type: application/octet-stream
Size: 5313 bytes
Desc: not available
Url : http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20040719/074b4385/q1.kgen.obj
-------------- next part --------------
A non-text attachment was scrubbed...
Name: q1.lapw1.error
Type: application/octet-stream
Size: 7351 bytes
Desc: not available
Url : http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20040719/074b4385/q1.lapw1.obj