[Q-e-developers] QE 6.0 CRASH in GWW example2

[Paolo Giannozzi]

Paolo Umari has fixed the problem. Please try the attached patch (that
includes the one I sent a few days ago)

Paolo

On Sun, Oct 9, 2016 at 10:10 PM, Paolo Giannozzi <p.giannozzi at gmail.com>
wrote:

> On Sun, Oct 9, 2016 at 3:41 PM, Евгений Васильевич Тихонов <
> e.tikhonov at physics.msu.ru> wrote:
>
> [...] please check if GW examples are still working as intended.
>>
>
> unfortunately they aren't working at all. The crash you get is easily
> fixed (see patch below) but there is another problem
>
> Paolo
>
> Index: GWW/head/lanczos_k.f90
> ===================================================================
> --- GWW/head/lanczos_k.f90      (revision 13098)
> +++ GWW/head/lanczos_k.f90      (working copy)
> @@ -19,7 +19,7 @@
>    USE gvect
>    USE constants, ONLY : e2, pi, tpi, fpi
>    USE cell_base, ONLY: at, alat, tpiba, omega, tpiba2
> -  USE wvfct,    ONLY : g2kin, npwx, npw, nbnd
> +  USE wvfct,    ONLY : g2kin, npwx, nbnd
>    USE wavefunctions_module, ONLY : evc, psic
>    USE mp, ONLY : mp_sum, mp_barrier, mp_bcast
>    USE mp_world, ONLY : mpime, world_comm
> @@ -26,7 +26,7 @@
>    USE gvecs,                ONLY : nls, nlsm, doublegrid
>    USE g_psi_mod,            ONLY : h_diag, s_diag
>    USE uspp,                 ONLY : vkb, nkb, okvan
> -  USE klist,                ONLY : xk,igk_k
> +  USE klist,                ONLY : xk,igk_k, ngk
>    USE noncollin_module,     ONLY : noncolin, npol
>    USE qpoint,       ONLY : igkq
>
> @@ -51,7 +51,7 @@
>    REAL(kind=DP), ALLOCATABLE :: c(:)
>
>
> -  INTEGER :: is,ig,ii,jj,it,ipol
> +  INTEGER :: npw, is,ig,ii,jj,it,ipol
>    INTEGER :: iunlan
>    COMPLEX(kind=DP) :: csca
>
> @@ -61,6 +61,8 @@
>    allocate(c(nstates))
>    allocate(spsi(npwx,nstates))
>
> +  npw = ngk(ik)
> +  allocate(igkq(1:npw))
>    igkq(1:npw) = igk_k(1:npw,ik)
>
>    t_out(:,:,:)=(0.d0,0.d0)
> @@ -295,6 +297,7 @@
>
>    enddo
>
> +  deallocate(igkq)
>    deallocate(psi_1,psi_2,psi_3)
>    deallocate(u_0,u_1)
>    deallocate(alpha,beta,gamma,n_1)
>
>
>
>
>
>
> --
> Paolo Giannozzi, Dip. Scienze Matematiche Informatiche e Fisiche,
> Univ. Udine, via delle Scienze 208, 33100 Udine, Italy
> Phone +39-0432-558216, fax +39-0432-558222
>
>


-- 
Paolo Giannozzi, Dip. Scienze Matematiche Informatiche e Fisiche,
Univ. Udine, via delle Scienze 208, 33100 Udine, Italy
Phone +39-0432-558216, fax +39-0432-558222
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/developers/attachments/20161013/2b483424/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: gww.diffs
Type: application/octet-stream
Size: 14257 bytes
Desc: not available
URL: <http://lists.quantum-espresso.org/pipermail/developers/attachments/20161013/2b483424/attachment.obj>