Hi Alberto, I digged a bit more into the code. Now my understanding is that I should set SolutionMethod to ELSI in the fdf, instead of PEXSI. In the compilation, -DSIESTA__PEXSI should not be used.
Thanks, Chong ________________________________ From: Chong Wang <ch-w...@outlook.com> Sent: Wednesday, January 19, 2022 2:05:23 AM To: siesta-l@uam.es <siesta-l@uam.es> Subject: RE: How to compile the psml-support branch with PEXSI Hi Alberto, Thanks for the reply. However, the same issue occurs on rel-MaX-1 branch. For example: m_pexsi_dos.F90 and m_pexsi_drvier.F90 both call subroutine f_ppexsi_inertia_count_real_symmetric_matrix, which is not defined in f_interface.f90 in the PEXSI code distributed with ELSI. Chong From: Alberto Garcia<mailto:alber...@icmab.es> Sent: Tuesday, January 18, 2022 1:14 PM To: siesta-l<mailto:siesta-l@uam.es> Subject: Re: [SIESTA-L] How to compile the psml-support branch with PEXSI Hi, As explained in https://gitlab.com/siesta-project/siesta/-/wikis/Guide-to-Siesta-versions you need to use the rel-MaX-1 branch, which contains both PSML support and the ELSI interface. Best regards, Alberto ----- El 17 de Ene de 2022, a las 02:50, Chong Wang ch-w...@outlook.com escribió: | Hi, | | | | I want to use psml pseudopotential and PEXSI solver. As far as I understand, the | psml-support has a new interface to PEXSI through ELSI. Most of the compilation | are done, except for the density of states part, where the error is: | | | | Name in only-list does not exist or is not accessible. | [F_PPEXSI_INERTIA_COUNT_REAL_SYMMETRIC_MATRI] | | use f_ppexsi_interface, only: f_ppexsi_inertia_count_real_symmetric_matrix | | | | and | | | | Name in only-list does not exist or is not accessible. | [F_PPEXSI_LOAD_REAL_SYMMETRIC_HS_MATRIX] | | use f_ppexsi_interface, only: f_ppexsi_load_real_symmetric_hs_matrix | | | | I checked that the f_interface.f90 file (of PEXSI library) is indeed missing the | two subroutines. In addition, I noticed that the manual says: “Newer versions | of SIESTA (in the Gitlab development site) can use the current PEXSI library | through the ELSI library interface, which offers some more options, although | not currently the density-of-states calculation.” Therefore, I think maybe the | source code needs to be modified to make it work. | | | | Does anyone know how to make PEXSI work in the psml-support branch? I don’t need | the density of states. | | | | Chong Wang | | | | | -- | SIESTA is supported by the Spanish Research Agency (AEI) and by the European | H2020 MaX Centre of Excellence (http://www.max-centre.eu/)
-- SIESTA is supported by the Spanish Research Agency (AEI) and by the European H2020 MaX Centre of Excellence (http://www.max-centre.eu/)