[gpaw-users] segmentation fault in HSE for surfaces

[Ivano E. Castelli]

Dear users,

I would like to run HSE calculations of a surface (like TiO2) with 
various adsorbates.
First I ran a PBE calculation using the calculator:

"
#calculator
calc = GPAW(mode=PW(400),
             parallel={'domain': 1, 'band': 1},
             xc='PBE',
             txt='pbe.txt',
             kpts=(1,1,1),
             eigensolver='rmm-diis',
             width=0.10,
             mixer=Mixer(0.05, 5, weight=100.0),
             convergence={'bands': -10},
             spinpol = False,
             symmetry = 'off',
             poissonsolver = DipoleCorrection(PoissonSolver(),2),
             maxiter = 1000)
slab.set_calculator(calc)
slab.get_potential_energy()
calc.write('TiO2.gpw', mode='all')
"

I have a segmentation fault error when I ran the HSE part:

"
from ase.parallel import paropen
from ase.units import Hartree

from gpaw import GPAW, PW
from gpaw.mpi import size, rank, world, serial_comm
from gpaw.xc.tools import vxc
from gpaw.xc.hybridg import HybridXC
import pickle
import sys
import numpy as np

calc = GPAW('TiO2.gpw',
             parallel={'domain': 1},
             txt='HSE06.txt')

E = calc.get_potential_energy()
print E
hyb_calc = HybridXC('HSE06', bandstructure=True)

de_skn = vxc(calc, hyb_calc) - vxc(calc, 'PBE')
E_diff = E + calc.get_xc_difference(HybridXC('HSE06'))

for kpt in calc.wfs.kpt_u:
     kpt.eps_n += de_skn[kpt.s, kpt.k]
calc.occupations.calculate(calc.wfs)
ef = calc.get_fermi_level()

pickle.dump((E_diff, de_skn, ef), open(name+'_HSE06.pckl', 'w'), -1)
"

The error is "_pmiu_daemon(SIGCHLD): [NID 00132] [c0-0c2s1n0] [Thu Apr  
7 16:08:05 2016] PE RANK 0 exit signal Segmentation fault",
and the output from the HSE script is:

"
Memory estimate
---------------
Process memory now: 259.26 MiB
Calculator  1972.66 MiB
     Density  287.43 MiB
         Arrays  94.48 MiB
         Localized functions  156.61 MiB
         Mixer  36.34 MiB
     Hamiltonian  66.30 MiB
         Arrays  61.77 MiB
         XC  0.00 MiB
         Poisson  0.00 MiB
         vbar  4.52 MiB
     Wavefunctions  1618.93 MiB
         Arrays psit_nG  508.69 MiB
         Eigensolver  560.72 MiB
         Projections  17.44 MiB
         Projectors  8.90 MiB
         Overlap op  513.57 MiB
         PW-descriptor  9.60 MiB
...
Total number of cores used: 1

Symmetries present (total): 1

   ( 1  0  0)
   ( 0  1  0)
   ( 0  0  1)

1 k-point (Gamma)
1 k-point in the Irreducible Part of the Brillouin Zone

           k-points in crystal coordinates                weights
    0:     0.00000000    0.00000000    0.00000000          1/1

Mixer Type: Mixer
Linear Mixing Parameter: 0.05
Mixing with 5 Old Densities
Damping of Long Wave Oscillations: 100

Convergence Criteria:
     Total Energy Change: 0.0005 eV / electron
     Integral of Absolute Density Change: 0.0001 electrons
     Integral of Absolute Eigenstate Change: 4e-08 eV^2
Number of Atoms: 197
Number of Atomic Orbitals: 1355
Number of Bands in Calculation: 800
Bands to Converge: -30 Lowest Bands
Number of Valence Electrons: 1549
"
and
"
Spins: 1
Number of occupied bands (min, max): 799, 799
Number of valence electrons: 1549
Calculating eigenvalue shifts.
Value of alpha parameter: 599.301 Bohr^2
Value of gamma parameter: 56.733 Bohr^2
Number of pairs: 320399
1 x 1 x 1 k-points
Distributing 1 IBZ k-points over 1 process(es).
Distributing 1 q-vectors over 1 process(es).
Cutoff energies:
     Wave functions:          400.000 eV
     Density:                 803.869 eV
Time for first pair-density:      0.845 seconds
Estimated total time:        270799.287 seconds
"

How can I solve the error?

Thank you in advance,
best regards,
Ivano