scnc
/
quantum-espresso
mirror of https://gitlab.com/QEF/q-e.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
quantum-espresso/D3/d3_exc.f90

95 lines
2.6 KiB
Fortran
Raw Normal View History

O-sesame git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@2 c92efa57-630b-4861-b058-cf58834340f0
2003-01-20 05:58:50 +08:00
!
! Copyright (C) 2001 PWSCF group
! This file is distributed under the terms of the
! GNU General Public License. See the file `License'
! in the root directory of the present distribution,
! or http://www.gnu.org/copyleft/gpl.txt .
!
1) An experimental parallelization of NEB images has been implemented. This required a deep modification of the parallelism in PWscf: there are two new communicators (intra_image_comm and inter_image_comm) and the existing "pool" communicators (intra_pool_comm and inter_pool_comm) are now vectors of length given by the number of parallel images. #ifdef __PARA is no longer needed because all "parallel" variables are always initialized for a serial run and all parallel routines are, in the case of a serial run, dummy routines. The wrappers to MPI routines used only by PWscf are in the PW/para.f90 file. The others (mp_***) are in the Modules/mp.f90. All explicit referencies to mpif.h should be replaced by an "USE parallel_include" (in a serial run parallel_include is simply a dummy module). 2) The extrapolation of both potential and wavefunctions has been rewritten in order to be smarter than before: on the basis of the required extrapolation order, on the basis of the history and on the basis of which files are really present on the disk, the algorithm chooses the extrapolation order. All the algorithms in which ions are moved can use the extrapolation. These are both unstable features: I need the help of everybody to test them. C.S. git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@742 c92efa57-630b-4861-b058-cf58834340f0
2004-03-24 17:36:50 +08:00
#include "machine.h"
O-sesame git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@2 c92efa57-630b-4861-b058-cf58834340f0
2003-01-20 05:58:50 +08:00
!
!-----------------------------------------------------------------------
module allocate, calls to mallocate, mfree removed pointers replaced by allocatable whenever possible git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@56 c92efa57-630b-4861-b058-cf58834340f0
2003-02-08 00:04:36 +08:00
subroutine d3_exc
O-sesame git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@2 c92efa57-630b-4861-b058-cf58834340f0
2003-01-20 05:58:50 +08:00
!-----------------------------------------------------------------------
!
! Calculates the contribution to the derivative of the dynamical
! matrix due to the third derivative of the exchange and correlation
! energy
Removed references (now everywhere) to module basis for those variables already in module ions_base. C.S. git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@958 c92efa57-630b-4861-b058-cf58834340f0
2004-06-14 15:30:47 +08:00
!
USE ions_base, ONLY : nat
1) An experimental parallelization of NEB images has been implemented. This required a deep modification of the parallelism in PWscf: there are two new communicators (intra_image_comm and inter_image_comm) and the existing "pool" communicators (intra_pool_comm and inter_pool_comm) are now vectors of length given by the number of parallel images. #ifdef __PARA is no longer needed because all "parallel" variables are always initialized for a serial run and all parallel routines are, in the case of a serial run, dummy routines. The wrappers to MPI routines used only by PWscf are in the PW/para.f90 file. The others (mp_***) are in the Modules/mp.f90. All explicit referencies to mpif.h should be replaced by an "USE parallel_include" (in a serial run parallel_include is simply a dummy module). 2) The extrapolation of both potential and wavefunctions has been rewritten in order to be smarter than before: on the basis of the required extrapolation order, on the basis of the history and on the basis of which files are really present on the disk, the algorithm chooses the extrapolation order. All the algorithms in which ions are moved can use the extrapolation. These are both unstable features: I need the help of everybody to test them. C.S. git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@742 c92efa57-630b-4861-b058-cf58834340f0
2004-03-24 17:36:50 +08:00
USE kinds, ONLY : DP
O-sesame git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@2 c92efa57-630b-4861-b058-cf58834340f0
2003-01-20 05:58:50 +08:00
use pwcom
use phcom
use d3com
use para
Calls to poolbcast replaced by calls to mp_bcast(*,MPI_COMM_ROW). PW/broadcast.f90 PW/poolbcast.f90 removed. C.S. git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@707 c92efa57-630b-4861-b058-cf58834340f0
2004-03-09 01:21:27 +08:00
USE io_global, ONLY : ionode_id
1) An experimental parallelization of NEB images has been implemented. This required a deep modification of the parallelism in PWscf: there are two new communicators (intra_image_comm and inter_image_comm) and the existing "pool" communicators (intra_pool_comm and inter_pool_comm) are now vectors of length given by the number of parallel images. #ifdef __PARA is no longer needed because all "parallel" variables are always initialized for a serial run and all parallel routines are, in the case of a serial run, dummy routines. The wrappers to MPI routines used only by PWscf are in the PW/para.f90 file. The others (mp_***) are in the Modules/mp.f90. All explicit referencies to mpif.h should be replaced by an "USE parallel_include" (in a serial run parallel_include is simply a dummy module). 2) The extrapolation of both potential and wavefunctions has been rewritten in order to be smarter than before: on the basis of the required extrapolation order, on the basis of the history and on the basis of which files are really present on the disk, the algorithm chooses the extrapolation order. All the algorithms in which ions are moved can use the extrapolation. These are both unstable features: I need the help of everybody to test them. C.S. git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@742 c92efa57-630b-4861-b058-cf58834340f0
2004-03-24 17:36:50 +08:00
USE mp_global, ONLY : inter_pool_comm, my_image_id
Calls to poolbcast replaced by calls to mp_bcast(*,MPI_COMM_ROW). PW/broadcast.f90 PW/poolbcast.f90 removed. C.S. git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@707 c92efa57-630b-4861-b058-cf58834340f0
2004-03-09 01:21:27 +08:00
USE mp, ONLY : mp_bcast
1) An experimental parallelization of NEB images has been implemented. This required a deep modification of the parallelism in PWscf: there are two new communicators (intra_image_comm and inter_image_comm) and the existing "pool" communicators (intra_pool_comm and inter_pool_comm) are now vectors of length given by the number of parallel images. #ifdef __PARA is no longer needed because all "parallel" variables are always initialized for a serial run and all parallel routines are, in the case of a serial run, dummy routines. The wrappers to MPI routines used only by PWscf are in the PW/para.f90 file. The others (mp_***) are in the Modules/mp.f90. All explicit referencies to mpif.h should be replaced by an "USE parallel_include" (in a serial run parallel_include is simply a dummy module). 2) The extrapolation of both potential and wavefunctions has been rewritten in order to be smarter than before: on the basis of the required extrapolation order, on the basis of the history and on the basis of which files are really present on the disk, the algorithm chooses the extrapolation order. All the algorithms in which ions are moved can use the extrapolation. These are both unstable features: I need the help of everybody to test them. C.S. git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@742 c92efa57-630b-4861-b058-cf58834340f0
2004-03-24 17:36:50 +08:00
O-sesame git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@2 c92efa57-630b-4861-b058-cf58834340f0
2003-01-20 05:58:50 +08:00
implicit none
1) An experimental parallelization of NEB images has been implemented. This required a deep modification of the parallelism in PWscf: there are two new communicators (intra_image_comm and inter_image_comm) and the existing "pool" communicators (intra_pool_comm and inter_pool_comm) are now vectors of length given by the number of parallel images. #ifdef __PARA is no longer needed because all "parallel" variables are always initialized for a serial run and all parallel routines are, in the case of a serial run, dummy routines. The wrappers to MPI routines used only by PWscf are in the PW/para.f90 file. The others (mp_***) are in the Modules/mp.f90. All explicit referencies to mpif.h should be replaced by an "USE parallel_include" (in a serial run parallel_include is simply a dummy module). 2) The extrapolation of both potential and wavefunctions has been rewritten in order to be smarter than before: on the basis of the required extrapolation order, on the basis of the history and on the basis of which files are really present on the disk, the algorithm chooses the extrapolation order. All the algorithms in which ions are moved can use the extrapolation. These are both unstable features: I need the help of everybody to test them. C.S. git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@742 c92efa57-630b-4861-b058-cf58834340f0
2004-03-24 17:36:50 +08:00
module allocate, calls to mallocate, mfree removed pointers replaced by allocatable whenever possible git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@56 c92efa57-630b-4861-b058-cf58834340f0
2003-02-08 00:04:36 +08:00
integer :: errcode, ir, ipert, jpert, kpert, npert1, npert2
4 feb 2004 D3 code updated to work with max_irr_dim != 3, some cleanup. D3example added. (Gernot Deinzer) git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@558 c92efa57-630b-4861-b058-cf58834340f0
2004-02-04 19:06:45 +08:00
real (kind = dp) :: d2mxc, rhotot, xq0 (3)
real (kind = dp), allocatable :: d2muxc (:)
complex (kind = dp) :: aux
complex (kind = dp), allocatable :: work1 (:), work2 (:), work3 (:), d3dyn1 (:,:,:)
O-sesame git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@2 c92efa57-630b-4861-b058-cf58834340f0
2003-01-20 05:58:50 +08:00
module allocate, calls to mallocate, mfree removed pointers replaced by allocatable whenever possible git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@56 c92efa57-630b-4861-b058-cf58834340f0
2003-02-08 00:04:36 +08:00
allocate (d2muxc( nrxx))
allocate (work1 ( nrxx))
allocate (work2 ( nrxx))
allocate (work3 ( nrxx))
allocate (d3dyn1( 3*nat, 3*nat, 3*nat))
Support for lahey compiler added: "error" renamed to "errore", "rnd" to "rndx" bug in io_base fixed PARA => __PARA cinterpolate moved into interpolate git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@86 c92efa57-630b-4861-b058-cf58834340f0
2003-02-21 22:57:00 +08:00
#ifdef __PARA
module allocate, calls to mallocate, mfree removed pointers replaced by allocatable whenever possible git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@56 c92efa57-630b-4861-b058-cf58834340f0
2003-02-08 00:04:36 +08:00
if (mypool.ne.1) goto 100
O-sesame git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@2 c92efa57-630b-4861-b058-cf58834340f0
2003-01-20 05:58:50 +08:00
#endif
!
! Calculates third derivative of Exc
!
Cleanup in PH/ and D3/, setv removed (please verify collateral damages!) git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@698 c92efa57-630b-4861-b058-cf58834340f0
2004-03-07 21:47:42 +08:00
d2muxc(:) = 0.d0
module allocate, calls to mallocate, mfree removed pointers replaced by allocatable whenever possible git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@56 c92efa57-630b-4861-b058-cf58834340f0
2003-02-08 00:04:36 +08:00
do ir = 1, nrxx
rhotot = rho (ir, 1) + rho_core (ir)
Cleanup in PH/ and D3/, setv removed (please verify collateral damages!) git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@698 c92efa57-630b-4861-b058-cf58834340f0
2004-03-07 21:47:42 +08:00
if (rhotot > 1.d-30) d2muxc (ir) = d2mxc (rhotot)
if (rhotot < - 1.d-30) d2muxc (ir) = - d2mxc ( - rhotot)
O-sesame git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@2 c92efa57-630b-4861-b058-cf58834340f0
2003-01-20 05:58:50 +08:00
enddo
!
! Calculates the contribution to d3dyn
!
Cleanup in PH/ and D3/, setv removed (please verify collateral damages!) git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@698 c92efa57-630b-4861-b058-cf58834340f0
2004-03-07 21:47:42 +08:00
d3dyn1 (:,:,:) = (0.d0, 0.d0)
module allocate, calls to mallocate, mfree removed pointers replaced by allocatable whenever possible git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@56 c92efa57-630b-4861-b058-cf58834340f0
2003-02-08 00:04:36 +08:00
do ipert = 1, 3 * nat
if (q0mode (ipert) ) then
call davcio_drho (work1, lrdrho, iud0rho, ipert, - 1)
do jpert = 1, 3 * nat
call davcio_drho (work2, lrdrho, iudrho, jpert, - 1)
do kpert = 1, 3 * nat
call davcio_drho (work3, lrdrho, iudrho, kpert, - 1)
aux = DCMPLX (0.d0, 0.d0)
do ir = 1, nrxx
O-sesame git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@2 c92efa57-630b-4861-b058-cf58834340f0
2003-01-20 05:58:50 +08:00
aux = aux + d2muxc (ir) * work1 (ir) * conjg (work2 (ir) ) &
* work3 (ir)
enddo
Support for lahey compiler added: "error" renamed to "errore", "rnd" to "rndx" bug in io_base fixed PARA => __PARA cinterpolate moved into interpolate git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@86 c92efa57-630b-4861-b058-cf58834340f0
2003-02-21 22:57:00 +08:00
#ifdef __PARA
module allocate, calls to mallocate, mfree removed pointers replaced by allocatable whenever possible git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@56 c92efa57-630b-4861-b058-cf58834340f0
2003-02-08 00:04:36 +08:00
call reduce (2, aux)
O-sesame git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@2 c92efa57-630b-4861-b058-cf58834340f0
2003-01-20 05:58:50 +08:00
#endif
module allocate, calls to mallocate, mfree removed pointers replaced by allocatable whenever possible git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@56 c92efa57-630b-4861-b058-cf58834340f0
2003-02-08 00:04:36 +08:00
d3dyn1 (ipert, jpert, kpert) = omega * aux / (nr1 * nr2 * nr3)
O-sesame git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@2 c92efa57-630b-4861-b058-cf58834340f0
2003-01-20 05:58:50 +08:00
enddo
enddo
endif
enddo
Support for lahey compiler added: "error" renamed to "errore", "rnd" to "rndx" bug in io_base fixed PARA => __PARA cinterpolate moved into interpolate git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@86 c92efa57-630b-4861-b058-cf58834340f0
2003-02-21 22:57:00 +08:00
#ifdef __PARA
Again bugs in call to mp_bcast when replaced poolbcast, one should check that npool /= 1 (previously done in poolbcast) git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@709 c92efa57-630b-4861-b058-cf58834340f0
2004-03-10 02:01:31 +08:00
100 continue
1) An experimental parallelization of NEB images has been implemented. This required a deep modification of the parallelism in PWscf: there are two new communicators (intra_image_comm and inter_image_comm) and the existing "pool" communicators (intra_pool_comm and inter_pool_comm) are now vectors of length given by the number of parallel images. #ifdef __PARA is no longer needed because all "parallel" variables are always initialized for a serial run and all parallel routines are, in the case of a serial run, dummy routines. The wrappers to MPI routines used only by PWscf are in the PW/para.f90 file. The others (mp_***) are in the Modules/mp.f90. All explicit referencies to mpif.h should be replaced by an "USE parallel_include" (in a serial run parallel_include is simply a dummy module). 2) The extrapolation of both potential and wavefunctions has been rewritten in order to be smarter than before: on the basis of the required extrapolation order, on the basis of the history and on the basis of which files are really present on the disk, the algorithm chooses the extrapolation order. All the algorithms in which ions are moved can use the extrapolation. These are both unstable features: I need the help of everybody to test them. C.S. git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@742 c92efa57-630b-4861-b058-cf58834340f0
2004-03-24 17:36:50 +08:00
IF ( npool /= 1 ) &
The array of communicators was non necessary (thanks to SdG): removed everywhere. Fixed a serious bug in para.f90 (introduced in the last commit). C.S. git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@744 c92efa57-630b-4861-b058-cf58834340f0
2004-03-25 00:30:44 +08:00
CALL mp_bcast( d3dyn1, ionode_id, inter_pool_comm )
O-sesame git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@2 c92efa57-630b-4861-b058-cf58834340f0
2003-01-20 05:58:50 +08:00
#endif
Cleanup in PH/ and D3/, setv removed (please verify collateral damages!) git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@698 c92efa57-630b-4861-b058-cf58834340f0
2004-03-07 21:47:42 +08:00
d3dyn = d3dyn + d3dyn1
d3dyn_aux9 = d3dyn1
O-sesame git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@2 c92efa57-630b-4861-b058-cf58834340f0
2003-01-20 05:58:50 +08:00
module allocate, calls to mallocate, mfree removed pointers replaced by allocatable whenever possible git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@56 c92efa57-630b-4861-b058-cf58834340f0
2003-02-08 00:04:36 +08:00
deallocate (d2muxc)
deallocate (work1)
deallocate (work2)
deallocate (work3)
deallocate (d3dyn1)
return
O-sesame git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@2 c92efa57-630b-4861-b058-cf58834340f0
2003-01-20 05:58:50 +08:00
end subroutine d3_exc