Page 1 of 1

ypp excitons wave plot :: Segmentation fault :: ver 3.4.0

Posted: Sat Apr 27, 2013 2:48 pm
by martin.gmitra
Dear all,
I am experiencing crash of ypp when asking for excitonic wave function plot [ypp -e w]. I have gfortran compilation of recent svn 3.4.0. The tesbed is Solid_LiF from tutorial. The ypp runs well for the sort case where I can obtain .exc_I_sorted and .exc_E_sorted files. Please see below stdout as well as my ypp.in. The same problem I do have with version 3.3.0 (from QEForge). In attachment I am enclosing valdgrind scan.

Do you have any ideas?
Thanks,
Martin

---------------------------------------------------------------------------------------------------------
ypp -N -F ypp.in -J PlotJOB

<---> [01] Y(ambo) P(ost)/(re) P(rocessor)
<---> [02] Core DB
<---> :: Electrons : 8.000000
<---> :: Temperature [ev]: 0.000000
<---> :: Lattice factors [a.u.]: 3.851738 3.851738 3.851738
<---> :: K points : 10
<---> :: Bands : 10
<---> :: Symmetries : 48
<---> :: RL vectors : 1885
<---> [03] K-point grid
<---> :: Q-points (IBZ): 19
<---> :: X K-points (IBZ): 10
<---> [04] CORE Variables Setup
<---> [04.01] Unit cells
<---> [04.02] Symmetries
<---> [04.03] RL shells
<---> [04.04] K-grid lattice
<---> [04.05] Energies [ev] & Occupations
<---> [05] Excitonic Properties
<---> :: Sorting energies
<---> [FFT-EXCWF] Mesh size: 15 15 15
<---> [WF-EXCWF loader] Wfs (re)loading |####################| [100%] --(E) --(X)
<---> [05.01] Excitonic Wave Function
<---> [05.01.01] Real-Space grid setup
<---> :: Extended grid : 45 45 45
<---> :: Hole position in the DL cell [cc]: 2.000000 2.000000 2.000000
<---> :: position in the FFT grid [cc]: 2.054260 2.054260 2.054260
<---> :: translated position [cc]: 9.757735 9.757735 9.757735
<---> :: [A]: 5.163571 5.163571 5.163571
<---> Processing 1 states
<---> :: State 3 Merged with states 1 -> 3
<---> ExcWF@3 | | [000%] --(E) --(X)Segmentation fault


---------------------------------------------------------------------------------------------------------
ypp.in

wavefunction # [R] Wavefunction
excitons # [R] Excitons
Format= "g" # Output format [(c)ube/(g)nuplot/(x)crysden]
Direction= "1" # [rlu] [1/2/3] for 1d or [12/13/23] for 2d [123] for 3D
FFTGvecs= 749 RL # [FFT] Plane-waves
States= "3 - 3" # Index of the BS state(s)
Degen_Step= 0.0100 eV # Maximum energy separation of two degenerate states
MinWeight= 0.0100 # Print transitions with weight => MinWeight
% Cells
3 | 3 | 3 | # Number of cell repetitions (even or 1)
%
% Hole
2.00 | 2.00 | 2.00 | # [cc] Hole position in unit cell

Re: ypp excitons wave plot :: Segmentation fault :: ver 3.4.

Posted: Sat Apr 27, 2013 3:05 pm
by Daniele Varsano
Dear Martin,
no idea on what's going on. In order to reproduce the error, can you also post the input/report file of your BSE calculation?
Thanks,
Daniele

Re: ypp excitons wave plot :: Segmentation fault :: ver 3.4.

Posted: Sat Apr 27, 2013 3:33 pm
by martin.gmitra
Dear Daniele,
Please see attached to the post the input and report files. I have realized that the report file contains
*ERR* line. Could it be the problem? If yes, could you drop more information how one avoids such cases?

Thanks,
Martin

Re: ypp excitons wave plot :: Segmentation fault :: ver 3.4.

Posted: Sat Apr 27, 2013 3:48 pm
by Daniele Varsano
Dear Martin,
that error it is not a problem, as you can see Yambo found an inconsistency in the kindx database, so it recalculate and write it again (see few lines below).In order to spot the problem I presume we have to reproduce the error. From your debug it looks a problem with the symmetries of the Wfs.
Best,
Daniele

Re: ypp excitons wave plot :: Segmentation fault :: ver 3.4.

Posted: Tue Apr 30, 2013 3:26 pm
by Daniele Varsano
Dear Martin,
indeed there was a bug in the exciton wavefunction building routine.
Thanks for pointing in out. We will solve it soon.
Stay tuned, I will post you when a corrected version is ready in the repository.

Best,
Daniele

Re: ypp excitons wave plot :: Segmentation fault :: ver 3.4.

Posted: Fri May 03, 2013 10:57 am
by Davide Sangalli
Dear Martin,
we have fixed this issue in the GPL svn repository. If you update and recompile yambo and ypp everything should work fine.
With the fix we also changed a module, so it is safer if, after the update, you first run "make clean_all" and then again "./configure" (+options) , make yambo interfaces ypp.

You should be able to use the yambo databases, i.e. you just need to run ypp

Best,
Davide

Re: ypp excitons wave plot :: Segmentation fault :: ver 3.4.

Posted: Sun May 12, 2013 4:45 pm
by ljzhou86
Dear Davide,
I also met such problem, but unfortunately, I cannot open up this website to get the lastest ver 3.4.0 : ''http://www.yambo-code.org/download.php' .
It shows the following err:
"Not Found

The requested URL /downloads.php was not found on this server."

Coulde you send me this code by email:ljzhou86@fjirsm.ac.cn. Thank u in advance!
Davide Sangalli wrote:Dear Martin,
we have fixed this issue in the GPL svn repository. If you update and recompile yambo and ypp everything should work fine.
With the fix we also changed a module, so it is safer if, after the update, you first run "make clean_all" and then again "./configure" (+options) , make yambo interfaces ypp.

You should be able to use the yambo databases, i.e. you just need to run ypp

Best,
Davide

Re: ypp excitons wave plot :: Segmentation fault :: ver 3.4.

Posted: Sun May 12, 2013 6:30 pm
by claudio
The link to download yambo is:

http://www.yambo-code.org/download/index.php

try to follow the instruction and let us know

claudio