iotk issue in yambo-3.2.2-rev.5

[Letizia Chiodo]

Dear developers,

I tried to install yambo-3.2.2-rev.5 on a local cluster, where previous version yambo-3.2.1-r.448 works properly.
448 was compiled (ifort and mpi) with iotk from espresso 4.2,
and p2y reads correctly calculations done with espresso 4.0,5, 4.1, 4.2.

On the other side, compiling yambo-3.2.2-rev.5 with iotk coming from the three cited versions of espresso (including 4.0,5)
seems to work, in the sense that executables are build,
but when running p2y on espresso calculations,
I obtain the result:

<---> Atomic data...########################################################################################################################
# FROM IOTK LIBRARY, VERSION 1.1.0development
# UNRECOVERABLE ERROR (ierr=1)
# ERROR IN: iotk_getline (iotk_scan.spp:812)
# CVS Revision: 1.7
#
iostat=-1
# ERROR IN: iotk_scan_tag (iotk_scan.spp:516)
# CVS Revision: 1.7
# ERROR IN: iotk_scan (iotk_scan.spp:710)
# CVS Revision: 1.7
# ERROR IN: iotk_scan_begin (iotk_scan.spp:94)
# CVS Revision: 1.7
########################################################################################################################
*** An error occurred in MPI_Comm_f2c
*** before MPI was initialized
*** MPI_ERRORS_ARE_FATAL (your MPI job will now abort)

I observed the same problem with the version compiled as a module on cineca sp6
(I'm sorry I cannot try to compile there with different iotk as the queue system is giving problems since some days).

ZZZoooo, my question is:
am I doing something really stupid?
or, is there, as far as you know, a version of iotk working with yambo-3.2.2-rev.5?
or maybe I am missing some new option in the ./configure?
or, also, a combination of things: e.g. iotk from a certain version of espresso and calculations runned with another version, can work?

thanks a lot! cheers
letizia

[Conor Hogan]

Can you post your data-file.xml file please, and I'll have a look at it.
Also perhaps please check if rev.3 works.

[Letizia Chiodo]

Hi Conor,

thanks, here they are.
I'll try as soon as possible rev3
btw, on sp6 now it works

ciao
letizia

[Letizia Chiodo]

Dear all,

the issue has been solved in
viewtopic.php?f=7&t=254

the problem was in the header of the PPs,
the error can be due to small differences, as the < in that post,
ot to a different header structure (as, for example, when pseudos are generated via the quantum espresso code).

by modifying the header, the p2y works fine

thanks a lot for finding the bug!
ciao
letizia

[kopinjol]

Hi,

I have been running into exactly the same problem. This is the error message i get when I run p2y -S -N


<---> [01] P(W) 2 Y(ambo) Ver. 4.0
<---> DBs path set to .
<---> Index file set to data-file.xml
<---> Header/K-points/Energies...done
<---> Cell data...done
<---> Atomic data...########################################################################################################################
# FROM IOTK LIBRARY, VERSION 1.2.0
# UNRECOVERABLE ERROR (ierr=1)
# ERROR IN: iotk_getline (iotk_scan.spp:834)
# CVS Revision: 1.23
iostat=-1
# ERROR IN: iotk_scan_tag (iotk_scan.spp:518)
# CVS Revision: 1.23
# ERROR IN: iotk_scan (iotk_scan.spp:712)
# CVS Revision: 1.23
# ERROR IN: iotk_scan_begin (iotk_scan.spp:94)
# CVS Revision: 1.23
########################################################################################################################
Attempting to use an MPI routine before initializing MPICH

Now I am using yambo-3.2.2-rev5 and to compile p2y i linked the iotk from espresso-4.2. And i am using a norm conserving pseudopotential AgPBE_nc.UPF created with ld1.x in the espresso distribution. I have played around with the PP headers but still keep running into the same problem. I will really appreciate any help!

Thanks in advance,
kopinjol baishya
Graduate Student
Deptt. Of Physics,
University of Illinois at Chicago
IL 60607

[Conor Hogan]

Dear Kopinjol,
The quick answer is to use p2y from revision 3 - the slow answer is that I'm working on a patch, but it will take a few days.
Conor

[Conor Hogan]

can you post the AgPBE_nc.UPF file explicitly, please? It appears to be corrupted in the tarfile.

[kopinjol]

Hi Conor,

Thanks for the reply. I will start with rev 3. I tried to upload the AgPBE_nc.UPF file but it doesn't let me upload an UPF file. So I have gzipped the file to be within the size limit. I hope it works.

Thanks again,
kopinjol

[kopinjol]

Hi,

I have another quick question. I somehow cannot access the subversion repositories. Is there another place where i can download the yambo-3.2.2-rev3 from? It is not there in the qeforge page, unless i am mistaken.

Thanks again,
kopinjol

[Conor Hogan]

Hey,
Actually today I managed to fix the bug, it will appear soon in the GPL repository.
Probably you are not able to access it due to not having uploaded your public key?

In the meantime, the p2y from the old tarball yambo-3.2.1-r.448.tar.bz2 will work fine.
For a quick and dirty fix, comment out the line 297 of interfaces/p2y/mod_p2y.F
call check_pseudo(n_atomic_species,psfile_)
and recompile - only if you have a norm conserving pseudopotential of course!
Conor