EPW Forum

In recent version of EPW (v.5.3.1) I am getting segmentation fault error just before completion. Can any body help? My inputs, part of the outputs and error messages are given below:
INPUTS:

ep_coupling = .true.
elph = .true.
kmaps = .false.
epbwrite = .true.
epbread = .false.
epwwrite = .true.
epwread = .false.
etf_mem = 1
max_memlt =8.5
nbndsub = 5,
wannierize = .true.
num_iter = 500
! dis_win_min =-10.00
dis_win_max =10.00
dis_froz_max = 0.69
proj(1) = 'c=-0.798194,1.102645,1.665936:s'
proj(2) = 'c=-1.910058,1.102741,1.665359:s'
proj(3) = 'c=-1.354043,0.139765,1.665573:s'
proj(4) = 'c=-1.353691,0.781778,1.156993:s'
proj(5) = 'c=-0.000010,1.563462,-0.242859:s'
wdata(1) = 'search_shells= 24'
wdata(2) = 'guiding_centres=.true.'
wdata(3) = 'dis_num_iter= 500'
iverbosity = 2
eps_acustic = 2.0 ! Lowest boundary for the phonon frequency
ephwrite = .true. ! Writes .ephmat files used when Eliasberg = .true.
fsthick = 0.8 ! eV
degaussw = 0.10 ! eV
nsmear = 1
delta_smear = 0.08 ! eV
degaussq = 0.5 ! meV
nqstep = 500
eliashberg = .true.
laniso = .true.
limag = .true.
! imag_read = .true.
lpade = .true.

conv_thr_iaxis = 1.0d-5

wscut = 1.00 ! eV Upper limit over frequency integration/summation in the Elisashberg eq

nstemp = 1
temps = 5.00
! nstemp = 8
! temps(1) = 10.00
! temps(2) = 80.00
nsiter = 1000
muc = 0.10
dvscf_dir = './phonons/save'
nk1 = 40
nk2 = 40
nk3 = 1

nq1 = 20
nq2 = 20
nq3 = 1

mp_mesh_k = .true.
nkf1 =200
nkf2 =200
nkf3 = 1

nqf1 =200
nqf2 =200
nqf3 = 1
/
44 cartesian

Last part of OUTPUT
Finish reading .ephmat files
Electron-phonon coupling strength = 0.4985667
Estimated Allen-Dynes Tc = 8.388066 K for muc = 0.10000
Estimated w_log in Allen-Dynes Tc = 60.146591 meV
Estimated BCS superconducting gap = 1.272177 meV
temp( 1) = 5.00000 K
Solve anisotropic Eliashberg equations on imaginary-axis
Total number of frequency points nsiw( 1) = 369
Cutoff frequency wscut = 1.0003
Size of allocated memory per pool: ~= 2.9923 Gb
iter ethr znormi deltai [meV]
1 2.971508E+00 1.491053E+00 1.431025E+00
2 3.740524E-01 1.490389E+00 1.847887E+00
3 1.197325E-01 1.489012E+00 2.245419E+00
4 1.926536E-01 1.490425E+00 1.884749E+00
5 4.602331E-01 1.479267E+00 3.403280E+00
6 2.042513E-02 1.483029E+00 3.546419E+00
7 8.545420E-02 1.481398E+00 3.841414E+00
8 6.067198E-02 1.482562E+00 3.624939E+00
........
..........
110 5.317355E-05 1.475504E+00 4.943397E+00
111 3.144410E-06 1.475504E+00 4.943404E+00

Error File
forrtl: severe (174): SIGSEGV, segmentation fault occurred
Image PC Routine Line Source
epw.x 0000000000F90BBD Unknown Unknown Unknown
libpthread-2.17.s 00002AD28E9775D0 Unknown Unknown Unknown
libc-2.17.so 00002AD28EF0B59C cfree Unknown Unknown
epw.x 0000000000FD2AED Unknown Unknown Unknown
epw.x 00000000004E650E io_eliashberg_mp_ 2072 io_eliashberg.f90
epw.x 00000000004E5197 io_eliashberg_mp_ 257 io_eliashberg.f90
epw.x 00000000005299C3 supercond_aniso_m 540 supercond_aniso.f90
epw.x 0000000000521F11 supercond_aniso_m 104 supercond_aniso.f90
epw.x 000000000042B8F1 eliashberg_eqs_ 66 eliashberg.f90
epw.x 0000000000407BCC MAIN__ 163 epw.f90
epw.x 000000000040719E Unknown Unknown Unknown
libc-2.17.so 00002AD28EEA83D5 __libc_start_main Unknown Unknown
epw.x 00000000004070A9 Unknown Unknown Unknown

Dear paritosh_modak:

I think that I already identified it; under some situations, ibin in https://gitlab.com/QEF/q-e/-/blob/7b583 ... .f90#L2048 can be 0, leading to the segmentation fault.

I found this issue when using the underconverged computation parameters, but it seems that it is not your case.

This will be fixed in the next version of EPW, but if you want to try to fix it temporarily at the moment, one of the things you can try is to revert the merge request at https://gitlab.com/QEF/q-e/-/commit/7b5 ... 7c813c5b08 .

Lastly, regarding the use of EPW for 2D materials, please check the page at viewtopic.php?f=3&t=1311 .

Sincerely,

H. Lee

Thank you very much. I will try this.

Dear Prof. Lee,
I have tried by reverting those merged lines. It works fine for the first temps. But in the next run, when I am giving
imag_read=.true.,
nstemp=16
and temps=5.0 80.0
during solving for temps=10 K
It is giving the same error at the same place as earlier.