[CP2K:2471] possible bug in cp_ddapc_types.F

[Teodoro Laino]

Yes, this is a bug. Will fix it asap.
Thanks
Teo

---------------------------------------------
Teodoro Laino
Zurich Switzerland

Contact info:
Tel.:     http://www.jajah.com/Teo
E-mail: teo... at laino.eu
              teodor... at gmail.com
---------------------------------------------

On 12 Dec 2009, at 00:12, Noam Bernstein <no... at bollweevil.gdbg.org>  
wrote:

> I'm seeing an odd crash, and I've traced it to a weird looking section
> in cp_ddapc_types.F, in cp_ddapc_ewald_create(), around line 317:
>
>    ! Determining the tasks and further check
>    IF (do_qmmm_periodic.AND.decoupling.AND.(output_unit>0)) THEN
>       ! Check than an additional POISSON section has not been
> defined. In case write a warning
>       WRITE(output_unit,'(T2,"WARNING",A)')&
>            "A calculation with the QMMM periodic model has been  
> requested.",&
>            "The explicit POISSON section in DFT section will be  
> IGNORED.",&
>            "QM Electrostatic controlled only by the PERIODIC section
> in QMMM section"
>       decoupling = .FALSE.
>    END IF
>    IF (decoupling) THEN
>       ! Simple decoupling technique
>     .
>     .
>     .
>
> In particular, the "decoupling = .FALSE." line is only executed if
> output_unit > 0.
> I see why that test in the if is needed for the warning statement, but
> if decoupling
> should be set to false, shouldn't that happen regardless of the  
> output_unit?  In
> particular, when I run in parallel I end up with output_unit > 0 only
> on 1 node, and
> then different nodes disagree on whether decoupling is true or  
> false, and things
> go horribly wrong.
>
> Am I correct about this being a bug?  If so, I suppose that  
> decoupling should
> be set to false regardless of output_unit, and only the WRITE  
> statement should
> be surrounded by a check for output_unit > 0 ?
>
>           thanks,
>           Noam
>
> P.S. Input file is here, just for completeness.  Presumably this got
> wasn't found
> before because we're doing something not quite right, triggering the  
> warning.
>
> &FORCE_EVAL
>  METHOD QMMM
>  &DFT
>    BASIS_SET_FILE_NAME BASIS_SET
>    POTENTIAL_FILE_NAME POTENTIAL
>    WFN_RESTART_FILE_NAME H2O-RESTART.wfn_end
>    &MGRID
>      COMMENSURATE
>      CUTOFF 180.0
>      NGRIDS 4
>      REL_CUTOFF 30.0
>    &END MGRID
>    &POISSON
>      PERIODIC NONE
>      POISSON_SOLVER MULTIPOLE
>      &MULTIPOLE
>        EWALD_PRECISION 1.0E-5
>        RCUT 5.0
>        ANALYTICAL_GTERM T
>      &END MULTIPOLE
>    &END POISSON
>    &QS
>      EXTRAPOLATION PS
>      EXTRAPOLATION_ORDER 1
>      EPS_DEFAULT 1.0E-3
>    &END QS
>    &SCF
>      SCF_GUESS restart #atomic
>      EPS_SCF 1.0E-3
>      MAX_SCF 100
>      &OUTER_SCF
>         EPS_SCF 1.0E-3
>         MAX_SCF 100
>      &END
>      &OT
>         MINIMIZER CG
>         PRECONDITIONER FULL_ALL
>      &END OT
>    &END SCF
>    &XC
>      &XC_FUNCTIONAL PADE
>      &END XC_FUNCTIONAL
>      &XC_GRID
> #        USE_FINER_GRID
>        XC_DERIV NN10_SMOOTH
>        XC_SMOOTH_RHO NN10
>      &END XC_GRID
>    &END XC
>  &END DFT
>  &MM
>    &FORCEFIELD
>      PARMTYPE CHM
>      PARM_FILE_NAME charmm.pot
>    &END FORCEFIELD
>    &POISSON
>      &EWALD
>        EWALD_TYPE ewald
>        EWALD_ACCURACY 1.E-5
>        ALPHA 0.35
>        GMAX 33
>      &END EWALD
>    &END POISSON
>  &END MM
>  &QMMM
> #    MM_POTENTIAL_FILE_NAME ../MM_POTENTIAL
>    USE_GEEP_LIB 15
>    &CELL
>      ABC 7.000000 7.000000 7.000000
>      PERIODIC XYZ
>    &END CELL
>    ECOUPL GAUSS
>    &PERIODIC
>      GMAX 1.0
>      NGRIDS 50 50 50
>      &MULTIPOLE
>        RCUT 5.0
>        EWALD_PRECISION 1.E-5
>        NGRIDS 50 50 50
>        ANALYTICAL_GTERM T
>        &INTERPOLATOR
>        &END INTERPOLATOR
>      &END MULTIPOLE
>    &END PERIODIC
>    &MM_KIND OT
>      RADIUS 0.780
>    &END MM_KIND
>    &MM_KIND HT
>      RADIUS 0.440
>    &END MM_KIND
> #    otherwise use default: 0.800
>    &QM_KIND H
>      MM_INDEX 2
>      MM_INDEX 3
>    &END QM_KIND
>    &QM_KIND O
>      MM_INDEX 1
>    &END QM_KIND
>    &PRINT
>      &DERIVATIVES
>      &END DERIVATIVES
>    &END PRINT
>  &END QMMM
>  &SUBSYS
>    &CELL
>      A 10. 0.000000 0.000000
>      B 0.000000 10. 0.000000
>      C 0.000000 0.000000 10.
>      PERIODIC XYZ
>    &END CELL
> #    &COORD
> #    &END COORD
> #    &VELOCITY
> #    &END VELOCITY
>    &TOPOLOGY
>      &DUMP_PSF
>      &END DUMP_PSF
>      &DUMP_PDB
>      &END DUMP_PDB
>      &GENERATE
>        BONDLENGTH_MAX     5.0
>        REORDER F
>        CREATE_MOLECULES F
>        &ISOLATED_ATOMS
>          LIST 1
>          LIST 2
>          LIST 3
>        &END ISOLATED_ATOMS
>      &END GENERATE
>      CHARGE_EXTENDED
>      COORDINATE EXYZ
>      COORD_FILE_NAME H2O2.exyz
>      CONN_FILE_NAME psf.H2O.psf
>      CONN_FILE_FORMAT PSF
>    &END TOPOLOGY
>    &KIND H
>      BASIS_SET DZVP-GTH-PADE
>      POTENTIAL GTH-PADE-q1
>    &END KIND
>    &KIND O
>      BASIS_SET DZVP-GTH-PADE
>      POTENTIAL GTH-PADE-q6
>    &END KIND
>  &END SUBSYS
> &END FORCE_EVAL
> &GLOBAL
>  PROJECT H2O
>  RUN_TYPE MD
>  PRINT_LEVEL DEBUG #MEDIUM
> &END GLOBAL
> &MOTION
>  &PRINT
>    &FORCES
>      FORMAT XMOL
>    &END FORCES
>  &END PRINT
>  &MD
>    ENSEMBLE NVE
>    STEPS 0
>    TIMESTEP 0.50
>    TEMPERATURE 300.00
>    TEMP_KIND
>    &PRINT
>      &TEMP_KIND
>      &END TEMP_KIND
>    &END PRINT
>  &END MD
> &END MOTION
>
> --
>
> You received this message because you are subscribed to the Google  
> Groups "cp2k" group.
> To post to this group, send email to cp... at googlegroups.com.
> To unsubscribe from this group, send email to cp2k+uns... at googlegroups.com 
> .
> For more options, visit this group at http://groups.google.com/group/cp2k?hl=en 
> .
>
>