[Pw_forum] set_irr sets incorrect npert when trans=.false. and elph=.true.

[Andrea Dal Corso]

Please check the following:

http://www.democritos.it/pipermail/pw_forum/2010-November/018732.html


HTH,

Andrea


On Tue, 2010-12-21 at 00:45 +0100, Alaska Subedi wrote:
> Dear All,
> 
> I would like to calculate lambda using different sets of dense k
> points for double delta integration. In order to do so I calculate
> dvscf using the following ph.in file:
> 
> --
> &inputph
>   tr2_ph = 1.0d-16
>   prefix='Al-ep2',
>   fildvscf='dvscf'
>   amass(1) = 26.98
>   outdir = './tmp/'
>   fildyn = 'Al-ep0.dyn2'
> /
>    -0.250000000000000E+00   0.250000000000000E+00   -0.250000000000000E+00
> 
> I now want to use dvscf and dyn files generated from this run to
> calculate lambda. For this, I use the following input file:
> 
> --
> &inputph
>   tr2_ph = 1.0d-16
>   prefix='Al-ep2',
>   fildvscf='dvscf'
>   amass(1) = 26.98
>   outdir = './tmp/'
>   fildyn = 'Al-ep0.dyn2'
>   elph=.true.
>   trans=.false.
> /
>    -0.250000000000000E+00   0.250000000000000E+00   -0.250000000000000E+00
> 
> 
> However, the code  crashes with the error:
> forrtl: severe (41): insufficient virtual memory
> 
> After some debugging, I noticed that npertx is very large and the
> error happens when matrices of dimension npertx is being allocated in
> allocate_pert(). npertx is calculated as:
> 
>        DO irr = 1, nirr
>            npertx = max (npertx, npert (irr) )
>         ENDDO
> 
> The error occurs because npert has incorrect values assigned in the
> subroutine set_irr.
> 
> This error does not occur when I calculate phonon and lambda in the
> same run (using trans=.true. and elph=.true.).
> 
> Any pointer to fix this error would be greatly appreciated.
> 
> Thanks,
> Alaska
> 
> The scf file is:&control
>     calculation='scf'
>     restart_mode='from_scratch',
>     tprnfor = .true.
>     prefix='Al-ep2',
>     pseudo_dir = './pseudo/',
>     outdir='./tmp/'
> /
> 
> &system
>    ibrav=  2,
>    celldm(1) =7.5,
>    nat= 1, ntyp= 1,
>    ecutwfc =45.0,
>    occupations='smearing',
>    smearing='methfessel-paxton',
>    degauss=0.05
> /
> 
> &electrons
>     conv_thr = 1.0d-10
> !    diagonalization = 'cg'
> !    mixing_mode = 'local-TF'
>      mixing_beta = 0.3d0
> /
> 
> ATOMIC_SPECIES
>  Al  26.98 Al.vbc.UPF
> ATOMIC_POSITIONS
>  Al 0.00 0.00 0.00
> K_POINTS {automatic}
>   8  8  8  0  0  0
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
-- 
Andrea Dal Corso                    Tel. 0039-040-3787428
SISSA, Via Bonomea 265              Fax. 0039-040-3787249
I-34136 Trieste (Italy)             e-mail: dalcorso at sissa.it