[QE-users] [Solved]Error in SCF calculation with QE-7.1

[Yang Liu]

Promblem solved.

QE7.1 automatically applied the optimal parallel operation paramaters. In my calculation with only carbon, subspace diagonalization used a serial algorithm; while in the carbon+gold calculation, QE used a parallel algorithm automatically. I forces QE use a serial algorithm for subspace diagonalization "mpirun -np 48 pw.x -ndiag 1 -inp ....", the problem was gone.

I am not sure if the problem arises because I did not turn Scalapack on when I compiled QE.

Regards

--------------------------------------------------------------
Dear all

I compiled QE-7.1 on my Centos7 system with Intel oneAPI 2022; and I used pseudopotentials from pslibrary1.0.0.

When I ran a SCF calculation with only carbon, the calculation ended normally. When I ran a SCF calculation with carbon and gold, the calculation ceased with error. The output file went as:

......skipped contents......

     Initial potential from superposition of free atoms

     starting charge     527.9947, renormalised to     528.0000

     negative rho (up, down):  3.569E-03 0.000E+00
     Message from routine qexsd_readschema :
     xml data file ./Au.save/data-file-schema.xml not found
     Cannot read wfcs: file not found
     Starting wfcs are  432 randomized atomic wfcs

And I got error message in the command window as follow:

Abort(201931269) on node 4 (rank 4 in comm 0): Fatal error in PMPI_Comm_free: Invalid communicator, error stack:
PMPI_Comm_free(137): MPI_Comm_free(comm=0x7ffce2cfc908) failed
PMPI_Comm_free(85).: Null communicator
Abort(604584453) on node 8 (rank 8 in comm 0): Fatal error in PMPI_Comm_free: Invalid communicator, error stack:
PMPI_Comm_free(137): MPI_Comm_free(comm=0x7fffb2de8b88) failed
PMPI_Comm_free(85).: Null communicator
Abort(805911045) on node 20 (rank 20 in comm 0): Fatal error in PMPI_Comm_free: Invalid communicator, error stack:
PMPI_Comm_free(137): MPI_Comm_free(comm=0x7ffc7a797f08) failed
PMPI_Comm_free(85).: Null communicator

Could anyone tell me what was going wrong here?

Thanks

Mr. Yang LIU

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20221031/3a0855e8/attachment.html>