[Wien] Bandstructure calc. with polarization and SO

[Peter Blaha]

Hi,
Your procedure should be ok.

I guess it is a problem with compiler/mkl version ???

You may also try to recompile SRC_lapwso   with   -Dold_scalapack

Also: You are using the mpi-parallel version + k-point parallel. (the 
error is in the diagonalization using mpi). Can you setup the 
calculation only with k-parallelization ?

500 k-points sounds as if your unit cell is quite small and 
mpi-parallelism is not necessary (or may even be slower than the non-mpi 
version).

Am 12.11.2022 um 19:32 schrieb pboulet:
> Hello,
> 
> I am trying to calculate the BS of a structure with spin polarization 
> and SOC included, but the job always ends up with a failure during the 
> lapwso step.
> 
> I am not particularly comfortable with spin-polarized SOC…
> 
> The sequence I am running is the following (is it wrong?):
> x lapw1 -p -up -band
> x lapw1 -p -dn -band
> x lapwso -p -up
> x lapw2 -p -up -c -so -band -qtl
> x lapw2 -p -dn -c -so -band -qtl
> x irrep -p -up -so
> x irrep -p -dn -so
> x spaghetti -p -up -so
> x spaghetti -p -dn -so
> 
> Some of the SO vectors seem to be created as the lsv command shows:
> -rw-r----- 1 bouletpa  29G Nov 12 11:23 Pb31TlTe32.vectorup_1
> -rw-r----- 1 bouletpa  29G Nov 12 11:24 Pb31TlTe32.vectorup_2
> -rw-r----- 1 bouletpa  29G Nov 12 11:41 Pb31TlTe32.vectorup_3
> -rw-r----- 1 bouletpa  28G Nov 12 11:42 Pb31TlTe32.vectorup_4
> -rw-r----- 1 bouletpa  29G Nov 12 12:00 Pb31TlTe32.vectordn_1
> -rw-r----- 1 bouletpa  29G Nov 12 12:00 Pb31TlTe32.vectordn_2
> -rw-r----- 1 bouletpa  29G Nov 12 12:18 Pb31TlTe32.vectordn_3
> -rw-r----- 1 bouletpa  28G Nov 12 12:18 Pb31TlTe32.vectordn_4
> 
> -rw-r----- 1 bouletpa 7.0G Nov 12 13:25 Pb31TlTe32.vectorsodn_2
> -rw-r----- 1 bouletpa 7.0G Nov 12 13:26 Pb31TlTe32.vectorsoup_2
> -rw-r----- 1 bouletpa 1.7G Nov 12 13:44 Pb31TlTe32.vectorsodn_4
> -rw-r----- 1 bouletpa 1.7G Nov 12 13:44 Pb31TlTe32.vectorsoup_4
> -rw-r----- 1 bouletpa 478M Nov 12 13:51 Pb31TlTe32.vectorsodn_3
> -rw-r----- 1 bouletpa 478M Nov 12 13:51 Pb31TlTe32.vectorsoup_3
> -rw-r----- 1 bouletpa  30G Nov 12 17:02 Pb31TlTe32.vectorsodn_1
> -rw-r----- 1 bouletpa  30G Nov 12 17:02 Pb31TlTe32.vectorsoup_1
> 
> After the lapw1 jobs are finished, the lapwso one is launched:
> + x lapwso -p -up
> 
> which ends up with a segmentation fault:
> forrtl: severe (174): SIGSEGV, segmentation fault occurred
> Image              PC                Routine            Line        Source
> lapwso_mpi         000000000047365A  Unknown               Unknown  Unknown
> libpthread-2.17.s  00002AB523B96630  Unknown               Unknown  Unknown
> libmkl_avx2.so     00002AB54A280AE7  mkl_blas_avx2_xdc     Unknown  Unknown
> libmkl_avx2.so     00002AB54A2D7584  mkl_blas_avx2_xzc     Unknown  Unknown
> libmkl_core.so     00002AB528505509  mkl_blas_xzcopy       Unknown  Unknown
> libmkl_intel_thre  00002AB525FA5ECC  mkl_blas_zcopy        Unknown  Unknown
> libmkl_intel_lp64  00002AB52539BBCA  zcopy                 Unknown  Unknown
> libmkl_scalapack_  00002AB524ADE494  pzhbrdb_              Unknown  Unknown
> libmkl_scalapack_  00002AB5249EE37C  pzherdb_              Unknown  Unknown
> libmkl_scalapack_  00002AB524AB5E46  pzheevr_              Unknown  Unknown
> lapwso_mpi         0000000000424E30  hmsec_                    783  hmsec.F
> lapwso_mpi         0000000000432B87  MAIN__                    606  lapwso.F
> lapwso_mpi         00000000004079E2  Unknown               Unknown  Unknown
> libc-2.17.so       00002AB52D853555  __libc_start_main     Unknown  Unknown
> lapwso_mpi         00000000004078E9  Unknown               Unknown  Unknown
> 
> As I can see, 165 k-points and eigenvalues are reported in 
> Pb31TlTe32.outputso out of 502 (in Pb31TlTe32.klist_band).
> 
> Does anyone know what is going wrong?
> 
> Thank you for your help and time,
> Pascal
> 
> 
> Pascal Boulet
> —
> /Professor in computational materials chemistry - DEPARTMENT OF CHEMISTRY/
> University of Aix-Marseille - Avenue Escadrille Normandie Niemen - 
> F-13013 Marseille - FRANCE
> Tél: +33(0)4 13 55 18 10 - Fax : +33(0)4 13 55 18 50
> Email : pascal.boulet at univ-amu.fr <mailto:pascal.boulet at univ-amu.fr>
> 
> 
> 
> 
> 
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> SEARCH the MAILING-LIST at:  http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html

-- 
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-165300
Email: peter.blaha at tuwien.ac.at    WIEN2k: http://www.wien2k.at
WWW:   http://www.imc.tuwien.ac.at
-------------------------------------------------------------------------