ypp_ph giving segmentation fault at ELPH databases: transfer and I/O step

[dmurphy]

Hello,
I'm having an issue using ypp_ph for a metallic system, TiN. I'm using ypp_ph -g to create the gkkp DBs and I get a segmentation fault at the 'ELPH databases: transfer and I/O' step. I've tried to recompile yambo with no openmp, --enable-lin-alg and get the same error at the same point. I've also used a different machine (unsure of config but normally no issues) and again, segmentation fault at ELPH DB step.
I've also tried this for specifying the q-grid (as in the tutorial) and for a regular q-grid, same segmentation fault.

I also have to specify BSEScatt and BSE K-grids in the when setting up the DB with yambo_ph otherwise I get an another error, but that seems to work fine.

The Si tutorial for el-ph coupling works without any issues so I'm wondering what could be causing the segmentation fault? As the system is metallic are there any extra/increased parameters I need to be aware of? Or is there a particular configuration step I need to do?
I have attached the pw.x & yambo inputs to this message, you should be able to run them without issue. The pseudo dir for the QE calculations should be in the folder above where the pw.x/ph.x inputs are running.
Thank you!

[claudio]

Dear Daniel

may you upload your inputs, in such a way I can try to reproduce your error

best
Claudio

[claudio]

Dear Daniel

I notice you already posted inputs..... I will check them

claudio

[minyez]

Dear Daniel and claudio,

I am having a similar ypp_ph segfault issue with Yambo 5.0.4 when calculating a semiconductor CuAlO2.
With 8 q-points (4 in IBZ), ypp_ph breaks and dumps core after "ELPH databases: transfer and I/O". Core dump indicates the break at memory_mp_large_mems_().
With 216 q-points (32 in IBZ), ypp_ph breaks and dumps core after "Debye Energy" output. Core dump indicates the break at `ypp/excitons/excitons_spin.F:164`.
In both cases, r_gkkp ends right after "[06] == Electron-Phonon Databases ==" banner.
I am wondering whether there is some progress with this issue.

The ypp.in is a simple one

Code: Select all

gkkp
DBsPATH= "../elph_dir/"
PHfreqF= "none"
PHmodeF= "none"

the inputs (pseudo files and QE inputs with the two q-mesh) are provided in the attachment so that they may help reproduce the error.

Thank you!

Best Regards,
Min-Ye

[claudio]

Dear Min-Ye Zhang,

I reproduced your bug, I will try to fix it in few days

best
Claudio

[claudio]

Dear Min-Ye Zhang,

your bug is due to a problem in the definition of alat between Yambo and QE.
In order to solve it, please specify Alat in the ypp.in file when you generate the Q-grid:

bzgrids # [R] BZ Grid generator
Q_grid # [R] Q-grid analysis
OutputAlat= 11.078094 # [a.u.] Lattice constant used for "alat" ouput format
#NoWeights # Do not print points weight
cooIn= "rlu" # Points coordinates (in) cc/rlu/iku/alat
cooOut= "alat" # Points coordinates (out) cc/rlu/iku/alat
ListPts # List the internal q/k points also in the parser format
#ExpandPts # Expand the internal q/k points in the BZ
#ForceUserPts # Do not check the correcteness of the user points
%Qpts # Q points list
0.000000| 0.000000| 0.000000|
%

where the value 11.078094 is the one of your QE input celldm(1)= 11.078094.
This will generate the correct list of q-points that you can put in the phonon calculations.

with best regards
Claudio Attaccalite