[Pw_forum] SEGV error with ph.x

Silviu Zilberman silviu at Princeton.EDU
Tue Dec 7 05:46:06 CET 2004


Hi,

I am running a simple test case of HCN molecule in vacuum with the ph.x 
code (the gamma point code seems not to support yet ultrasoft PP). The 
program ends with a SEGV that I traced to be in subroutine solve_e(), at 
line 349:

 write(iunrec) dr2, dvscfin

I use Espresso2.1.1 on an intel cluster (RedHat) with intel compiler 
8.1. Below are more details (debugger output, and input files).

Any suggestions are greatly appreciated.

Thanks, Silviu.

==================================================================

Debbuger output:

Thread received signal SEGV
stopped at [subroutine solve_e():349 0x081b559f]
    349         write(iunrec) dr2, dvscfin
(idb) p dr2
7.3841959271995015e-08
(idb) p iunrec
99
(idb) p dvscfin
(1,1,1) (-2.8466913223266599, 0)
(2,1,1) (-3.6325173854827879, 0)
(3,1,1) (-3.985181331634521, 0)
(4,1,1) (-3.5353976726531982, 0)
(5,1,1) (-3.1116700172424316, 0)
(6,1,1) (-3.5572616577148435, 0)
(7,1,1) (-2.7180039644241329, 0)
(8,1,1) (-3.4988471031188961, 0)
(9,1,1) (-3.1163440227508543, 0)
(10,1,1) (-2.5414211511611935, 0)
(11,1,1) (-2.6584623813629147, 0)
(12,1,1) (-1.1985895156860351, 0)
(13,1,1) (-1.3288981556892394, 0)
(14,1,1) (-1.6060824871063231, 0)
(15,1,1) (-0.92308571338653556, 0)
(16,1,1) (-0.67997025251388543, 0)
(17,1,1) (-0.54657442569732662, 0)
(18,1,1) (0.72354159355163572, 0)
(19,1,1) (-2.5153556346893309, 0)
(20,1,1) (-2.5364445686340331, 0)
(21,1,1) (-0.10039620846509933, 0)
(22,1,1) (0.85398633480072017, 0)
(23,1,1) (0.20967396199703214, 0)
...

==================================================================

Input file for pw.x calculation:

 &control
    calculation='scf'
    restart_mode='from_scratch',
    pseudo_dir = '/home/lotus/silviu/Work/Pot/',
    outdir='/scratch/silviu/HCN/'
    title='Test HCN gamma only'
    prefix='HCN'
    tprnfor=.true.,
    etot_conv_thr=1e-8,
    forc_conv_thr=1e-6,
 /
 &system
    ibrav=1, celldm(1) =15.0, nat=3, ntyp= 3,
    ecutwfc =30.0,     nosym=.true.,
 /
  &electrons
    diagonalization='davidson',
    diago_david_ndim=8,
    mixing_mode='plain',
    mixing_beta = 0.7,
    conv_thr =  1.0d-8,
 /
  &ions
    ion_dynamics='bfgs',
    upscale=100,
 /
ATOMIC_SPECIES
   C 12.000  C_pbe_yw_1.van
   H 1.0079  H_pbe_yw_1.van
   N 14.0067 007-N-gpbe--bm.van
ATOMIC_POSITIONS (angstrom)
C         0.0000     0.0000  0.0000
H         1.06501    0.0000  0.0000
N        -1.15324    0.0000  0.0000
K_POINTS
1
0.0 0.0 0.0 1.0

==================================================================

Input file for ph.x calculation:

normal modes for HCN
 &inputph
  tr2_ph=1.0d-10, niter_ph=80,
  prefix='HCN',
  amass(1)=12.000,
  amass(2)=1.0079,
  amass(3)=14.0067,
  outdir='/scratch/silviu/HCN/',
  epsil=.true.,
  trans=.true.,
  fildyn='HCN.dyn'
/
0.0 0.0 0.0


-- 
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Zilberman Silviu
 
213 Frick Laboratory, 
Department of Chemistry, 
Princeton University
Princeton, NJ 08544, USA
 
phone: 609-258-1834
fax:   609-258-6746
silviu at Princeton.EDU
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

-------------- next part --------------
A non-text attachment was scrubbed...
Name: silviu.vcf
Type: text/x-vcard
Size: 271 bytes
Desc: not available
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20041206/cb7def08/attachment.vcf>


More information about the users mailing list