1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
|
! Copyright (C) 2017 T. Mueller, J. K. Dewhurst, S. Sharma and E. K. U. Gross.
! This file is distributed under the terms of the GNU General Public License.
! See the file COPYING for license details.
subroutine rhomaguk(ik0,lock,evecu)
use modmain
use modulr
use modomp
implicit none
! arguments
integer, intent(in) :: ik0
integer(omp_lock_kind), intent(inout) :: lock(nqpt)
complex(8), intent(in) :: evecu(nstulr,nstulr)
! local variables
integer ik,ikpa,jkpa
integer nst,ist,jst,i,j
integer ngk0,is,ias
integer npc,ir,nthd
real(8) ts0,ts1
real(4) wo
! automatic arrays
integer idx(nstsv)
complex(8) zfft(nqpt)
! allocatable arrays
complex(8), allocatable :: apwalm(:,:,:,:),evecfv(:,:),evecsv(:,:)
complex(8), allocatable :: evectv(:,:,:),evecsvt(:,:)
complex(4), allocatable :: wfmt(:,:,:,:),wfir(:,:,:)
call timesec(ts0)
! central k-point
ik=(ik0-1)*nkpa+1
! number of G+k-vectors for central k-point
ngk0=ngk(1,ik)
! get the eigenvectors from file
allocate(evecfv(nmatmax,nstfv),evecsv(nstsv,nstsv))
call getevecfv(filext,ik,vkl(:,ik),vgkl(:,:,:,ik),evecfv)
call getevecsv(filext,ik,vkl(:,ik),evecsv)
! find the matching coefficients
allocate(apwalm(ngkmax,apwordmax,lmmaxapw,natmtot))
call match(ngk0,vgkc(:,:,1,ik),gkc(:,1,ik),sfacgk(:,:,1,ik),apwalm)
allocate(evectv(nstsv,nstsv,nqpt))
call holdthd(nqpt,nthd)
!$OMP PARALLEL DEFAULT(SHARED) &
!$OMP PRIVATE(zfft,evecsvt,wfmt,wfir) &
!$OMP PRIVATE(ikpa,jkpa,ist,jst,i,j) &
!$OMP PRIVATE(ir,wo,ias,is,npc) &
!$OMP NUM_THREADS(nthd)
allocate(evecsvt(nstsv,nstsv))
allocate(wfmt(npcmtmax,natmtot,nspinor,nstsv),wfir(ngtc,nspinor,nstsv))
! loop over long-range states in subsets of size nstsv
do jkpa=1,nkpa
! number of and index to occupied states in subset
!$OMP BARRIER
!$OMP SINGLE
nst=0
do jst=1,nstsv
j=(jkpa-1)*nstsv+jst
if (abs(occulr(j,ik0)) < epsocc) cycle
nst=nst+1
idx(nst)=jst
end do
!$OMP END SINGLE
if (nst == 0) cycle
!$OMP DO SCHEDULE(DYNAMIC)
do jst=1,nst
j=(jkpa-1)*nstsv+idx(jst)
do ist=1,nstsv
zfft(1:nqpt)=0.d0
do ikpa=1,nkpa
i=(ikpa-1)*nstsv+ist
! store the long-range state in FFT Q-space
zfft(iqfft(ikpa))=evecu(i,j)
end do
! Fourier transform to R-space
call zfftifc(3,ngridq,1,zfft)
evectv(ist,jst,1:nqpt)=zfft(1:nqpt)
end do
end do
!$OMP END DO
! parallel loop over R-points
!$OMP DO SCHEDULE(DYNAMIC)
do ir=1,nqpt
! convert third-variational states to second-variational states
call zgemm('N','N',nstsv,nst,nstsv,zone,evecsv,nstsv,evectv(:,:,ir), &
nstsv,zzero,evecsvt,nstsv)
! generate the wavefunctions in single-precision
call genwfsv_sp(.false.,.false.,nst,[0],ngdgc,igfc,ngk0,igkig(:,1,ik), &
apwalm,evecfv,evecsvt,wfmt,ngtc,wfir)
! loop over second-variational states
do jst=1,nst
j=(jkpa-1)*nstsv+idx(jst)
wo=occulr(j,ik0)*wkpt(ik)
! add to the density and magnetisation
call omp_set_lock(lock(ir))
! muffin-tin part
do ias=1,natmtot
is=idxis(ias)
npc=npcmt(is)
if (spinpol) then
if (ncmag) then
call rmk1(npc,wo,wfmt(:,ias,1,jst),wfmt(:,ias,2,jst), &
rhormt(:,ias,ir),magrmt(:,ias,1,ir),magrmt(:,ias,2,ir), &
magrmt(:,ias,3,ir))
else
call rmk2(npc,wo,wfmt(:,ias,1,jst),wfmt(:,ias,2,jst), &
rhormt(:,ias,ir),magrmt(:,ias,1,ir))
end if
else
call rmk3(npc,wo,wfmt(:,ias,1,jst),rhormt(:,ias,ir))
end if
end do
! interstitial part
if (spinpol) then
if (ncmag) then
call rmk1(ngtc,wo,wfir(:,1,jst),wfir(:,2,jst),rhorir(:,ir), &
magrir(:,1,ir),magrir(:,2,ir),magrir(:,3,ir))
else
call rmk2(ngtc,wo,wfir(:,1,jst),wfir(:,2,jst),rhorir(:,ir), &
magrir(:,1,ir))
end if
else
call rmk3(ngtc,wo,wfir(:,1,jst),rhorir(:,ir))
end if
call omp_unset_lock(lock(ir))
end do
! end parallel loop over R-points
end do
!$OMP END DO
end do
deallocate(evecsvt,wfmt,wfir)
!$OMP END PARALLEL
call freethd(nthd)
deallocate(apwalm,evecfv,evecsv,evectv)
call timesec(ts1)
!$OMP ATOMIC
timerho=timerho+ts1-ts0
contains
pure subroutine rmk1(n,wo,wf1,wf2,rho,mag1,mag2,mag3)
implicit none
! arguments
integer, intent(in) :: n
real(4), intent(in) :: wo
complex(4), intent(in) :: wf1(n),wf2(n)
real(8), intent(inout) :: rho(n),mag1(n),mag2(n),mag3(n)
! local variables
integer i
real(4) wo2,a1,b1,a2,b2,t1,t2
wo2=2.e0*wo
!$OMP SIMD PRIVATE(a1,b1,a2,b2,t1,t2) SIMDLEN(8)
do i=1,n
a1=real(wf1(i)); b1=aimag(wf1(i))
a2=real(wf2(i)); b2=aimag(wf2(i))
t1=a1**2+b1**2; t2=a2**2+b2**2
mag1(i)=mag1(i)+wo2*(a1*a2+b1*b2)
mag2(i)=mag2(i)+wo2*(a1*b2-b1*a2)
mag3(i)=mag3(i)+wo*(t1-t2)
rho(i)=rho(i)+wo*(t1+t2)
end do
end subroutine
pure subroutine rmk2(n,wo,wf1,wf2,rho,mag)
implicit none
! arguments
integer, intent(in) :: n
real(4), intent(in) :: wo
complex(4), intent(in) :: wf1(n),wf2(n)
real(8), intent(inout) :: rho(n),mag(n)
! local variables
integer i
real(4) t1,t2
!$OMP SIMD PRIVATE(t1,t2) SIMDLEN(8)
do i=1,n
t1=real(wf1(i))**2+aimag(wf1(i))**2
t2=real(wf2(i))**2+aimag(wf2(i))**2
mag(i)=mag(i)+wo*(t1-t2)
rho(i)=rho(i)+wo*(t1+t2)
end do
end subroutine
pure subroutine rmk3(n,wo,wf,rho)
implicit none
! arguments
integer, intent(in) :: n
real(4), intent(in) :: wo
complex(4), intent(in) :: wf(n)
real(8), intent(inout) :: rho(n)
rho(1:n)=rho(1:n)+wo*(real(wf(1:n))**2+aimag(wf(1:n))**2)
end subroutine
end subroutine
|
