Broken PPA Claculation for Germanium

[rvatanme]

Dear all,

I'm trying to calculate indirect band gap of Ge by using PPA approximation, but when I run the code I get a wierd error message like:

Signal: Segmentation fault (11)
Signal code: Address not mapped (1)
Failing at address: 0x7ff18fec8a10
/lib/x86_64-linux-gnu/libpthread.so.0(+0x10340) [0x7ff196468340]

FYI, I didn't get any error while I run yambo for GW approximation of Ge. I also didn't get any error while I run yambo for GW and PPA calculations for Silicon.
Could any one please help me with fixing this error?

Best,
Reza Vatan Meidanshahi
Electrical Engineering Department
Arizona State University
USA

[Daniele Varsano]

Dear Reza,
it is impossible to help you with the few information you provided.
Please post input, report (r_*), standard output (l_*) and all other info can help to spot the problem.

Best,

Daniele

[rvatanme]

Dear Daniel,

Thanks for your prompt reply. I tried a lot to find a way to attach my report file, but I couldn't.
Any way this is the last part of my report file:

[WR./GoWo_PPA_HF7Ry//db.HF_and_locXC]-----------------------
Brillouin Zone Q/K grids (IBZ/BZ): 44 864 28 864
RL vectors (WF): 609
Coulomb cutoff potential :none
Electronic Temperature [K]: 0.000000
Bosonic Temperature [K]: 0.000000
Green`s function wavefunctions :Becke 88(X)+Lee, Yang & Parr(C)
Total number of QP states : 280
Exchange RL vectors : 137
Exchange summation bands : 4
RIM RL components :0
RIM random points :0
- S/N 003890 --------------------------- v.03.04.01 r.3187 -

CpuTiming [Min/Max/Average]: 01m-04s/01m-04s/01m-04s

[05] Dynamic Dielectric Matrix (PPA)
====================================

[WR./GoWo_PPA_HF7Ry//db.dipoles]----------------------------
Brillouin Zone Q/K grids (IBZ/BZ): 44 864 28 864

Following is also my input file:

gw0 # [R GW] GoWo Quasiparticle energy levels
ppa # [R Xp] Plasmon Pole Approximation
em1d # [R Xd] Dynamical Inverse Dielectric Matrix
HF_and_locXC # [R XX] Hartree-Fock Self-energy and Vxc
EXXRLvcs= 7 Ry # [XX] Exchange RL components
Chimod= "Hartree" # [X] IP/Hartree/ALDA/LRC/BSfxc
% QpntsRXs
1 | 8 | # [Xp] Transferred momenta
%
% BndsRnXs
1 | 30 | # [Xp] Polarization function bands
%
NGsBlkXp= 7 RL # [Xp] Response block size
% LongDrXs
1.000000 | 0.000000 | 0.000000 | # [Xp] [cc] Electric Field
%
PPAPntXs= 27.21138 eV # [Xp] PPA imaginary energy
% GbndRnge
1 | 30 | # [GW] G[W] bands range
%
GDamping= 0.100000 eV # [GW] G[W] damping
dScStep= 0.100000 eV # [GW] Energy step to evalute Z factors
DysSolver= "n" # [GW] Dyson Equation solver (`n`,`s`,`g`)
%QPkrange # [GW] QP generalized Kpoint/Band indices
1| 40| 1| 10|
%
%QPerange # [GW] QP generalized Kpoint/Energy indices
1| 40| 0.0|-1.0|
%

I hope this help.

Best,
Reza

[Daniele Varsano]

Dear Reza,
Please post your entire report file, and also the standard output (l_*) as I need to see exactly where the run crashes.
In order to attach the files, there is a button named "Upload attachment" it is just below the Submit button.
In case the error is not evident we will need to reproduce the error, in this case also your ground state input and pseudopotential files are needed, but you can post them later if needed.

Best,
Daniele

[rvatanme]

Dear Daniel,

Thanks for the reply. The attached ziped file contains all requested files.

Best,
Reza

[Daniele Varsano]

Dear Reza,
please fill you signature with your complete affiliation, this is a rule of the forum, you can do it in your profile once for all.
Unfortunately you did not send me the standard output the l_* file, so I cannot see where exactly the code stop.
Anyway I can see something strange in your input file:

Code: Select all

% QpntsRXs
 1 | 8 |                     # [Xp] Transferred momenta
%
% BndsRnXs
  1 | 30 |                   # [Xp] Polarization function bands
%
NGsBlkXs= 7            RL    # [Xp] Response block size
% LongDrXs
 1.000000 | 0.000000 | 0.000000 |        # [Xp] [cc] Electric Field
%
PPAPntXs= 27.21138

All this variables terminates with "s" which is meant for "static", ie for the static dielctric function "em1s", while in the Plasmon Pole approximation you should have the corresponding "dynamical" variable for the plasmon pole approximation terminating with "p":

Code: Select all

%BndsRnXp
%NGsBlkXp
%LongDrXp
%PPAPntXp

did you generate the input using the command line tool, or you edited by scratch by yourself?
The command line:

Code: Select all

yambo -p p -g n 

should give you the correct variables.
Most probably this is what creating confusion and make the code to crash.
Best,
Daniele

[rvatanme]

Dear Daneil,

Thanks for your reply.
First of all I don't know how to work properly with the following commend for editing my input file:

Code: Select all

yambo -p p -g n 

The problem is that I can not properly save and exit the file after opening the file with this commend. I use vi commend (wq! or wq) to save and exit the file but it doesn't work, because when I open it again, some of my changes is not included.

There is a very strange thing about the replacement of p by s. When I open my input file with "yambo -p p -g n" they are all "p" and yambo also reads "p" while running. But when I open the file with "vi" or gedit, they are all s.

The program didn't generate any "l_" file for me to send you. Maybe it is because when I run yambo, all calculation details showes up at the same terminal which I'm working with not in a different file as log file. I use this commend for running yambo:

Code: Select all

yambo -F input -J output 

Please let me know if I'm doing something wrong.

Best,
Reza

[Daniele Varsano]

Dear Reza,
I do not know if I well understood your problem with the editor.
Have you the right permission to write in the directory you are running.

Code: Select all

When I open my input file with "yambo -p p -g n" they are all "p" and yambo also reads "p" while running. But when I open the file with "vi" or gedit, they are all s. 

How do you know yambo reads "p"?

Code: Select all

The program didn't generate any "l_" file for me to send you. Maybe it is because when I run yambo, all calculation details showes up at the same terminal which I'm working with not in a different file as log file.

Very strange, you should have a file called l-output*. Again check your permission. Anyway you can redirect your output to a file.
In any case if you think that your vi is not working you can change editor but adding the flag in the configure:
--with-editor=
and recompile.

In any case, if you do solve the problem, I suggest you to talk with your system administrator about it.

Best,

Daniele