[Pw_forum] nscf error

Aritz Leonardo swblelia at sw.ehu.es
Mon Nov 8 11:17:11 CET 2004


Hello:

I am runing an electron-phonon calculation in an ALPHA machine with UNIX 5.1
The thing is that when the second nscf calculation is done with a DENSE 
automatic grid (24 24 18)
The execution stops (without aborting) and produces the following error 
file:

stty: tcgetattr: Not a typewriter
forrtl: error (72): floating overflow
   0: __FINI_00_remove_gp_range [0x3ff81a6de38]
   1: __FINI_00_remove_gp_range [0x3ff81a76a20]
   2: __FINI_00_remove_gp_range [0x3ff800d5b90]
   3: WFCINIT_K forrtl: error (78): process killed (SIGTERM)
   0: __FINI_00_remove_gp_range [0x3ff81a6de38]
   1: __FINI_00_remove_gp_range [0x3ff81a772c0]
   2: __FINI_00_remove_gp_range [0x3ff800d5b90]
   3: __FINI_00_remove_gp_range [0x3ff81ac9bcc]
   4: __FINI_00_remove_gp_range [0x3ff81aaf3c0]
   5: __FINI_00_remove_gp_range [0x3ff81a6de38]
   6: __FINI_00_remove_gp_range [0x3ff81a76a20]
   7: __FINI_00_remove_gp_range [0x3ff800d5b90]
   8: WFCINIT_K

The output file (mg.nscf2.out) where the process stops is:

      k(10798) = (  -0.3333333   0.7698004  -0.0855754), wk =   0.0000000
        k(10799) = (  -0.3333333   0.5773503  -0.2738413), wk =   0.0003858
        k(10800) = (  -0.3333333   0.7698004  -0.1198056), wk =   0.0000000

     G cutoff =   79.9791  (   4187 G-vectors)     FFT grid: ( 18, 18, 30)

     nbndx  =    24  nbnd   =     6  natomwfc =     8  npwx   =     549
     nelec  =    4.00 nkb   =     2  ngl    =     256

     The initial potential is read from file         mg.pot
     Starting wfc are atomic

And the subrutine where I think that crasses is the last part of--> 
wfcinit_k.f90, because I obtain
good results for gamma point. I think it has to be something with DEALLOCATE





More information about the users mailing list