Skip to content

Commit f5b64a0

Browse files
author
Tanush Prathi
committed
Ran formatter
1 parent 6d9b0fb commit f5b64a0

File tree

6 files changed

+76
-77
lines changed

6 files changed

+76
-77
lines changed

‎src/common/include/3dHardcodedIC.fpp‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -119,7 +119,7 @@
119119
q_prim_vf(advxe)%sf(i, j, k) = patch_icpp(1)%alpha(2)
120120
end if
121121

122-
case (302) ! 3D Jet with IGR
122+
case (302) ! 3D Jet with IGR
123123
ux_th = 10*sqrt(1.4*0.4)
124124
ux_am = 0.0*sqrt(1.4)
125125
p_th = 2.0_wp

‎src/pre_process/m_boundary_conditions.fpp‎

Lines changed: 2 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -47,7 +47,7 @@ contains
4747
if (patch_bc(patch_id)%loc == ${LOC}$ .and. bc_x%${BOUND}$ < 0) then
4848
do j = 0, n
4949
if (y_cc(j) > y_boundary%beg .and. y_cc(j) < y_boundary%end) then
50-
if(${LOC}$ == -1) then
50+
if (${LOC}$ == -1) then
5151
bc_type(1, 1)%sf(0, j, 0) = patch_bc(patch_id)%type
5252
else
5353
bc_type(1, 2)%sf(0, j, 0) = patch_bc(patch_id)%type
@@ -71,7 +71,7 @@ contains
7171
if (patch_bc(patch_id)%loc == ${LOC}$ .and. bc_y%${BOUND}$ < 0) then
7272
do j = 0, m
7373
if (x_cc(j) > x_boundary%beg .and. x_cc(j) < x_boundary%end) then
74-
if(${LOC}$ == -1) then
74+
if (${LOC}$ == -1) then
7575
bc_type(2, 1)%sf(j, 0, 0) = patch_bc(patch_id)%type
7676
else
7777
bc_type(2, 2)%sf(j, 0, 0) = patch_bc(patch_id)%type

‎src/simulation/m_acoustic_src.fpp‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -442,7 +442,7 @@ contains
442442
source_spatials_num_points(ai) = count
443443

444444
! Allocate arrays with the correct size
445-
445+
446446
@:ALLOCATE(source_spatials(ai)%coord(1:3, 1:count))
447447
@:ALLOCATE(source_spatials(ai)%val(1:count))
448448
@:ALLOCATE(source_spatials(ai)%angle(1:count))

‎src/simulation/m_fftw.fpp‎

Lines changed: 67 additions & 68 deletions
Original file line numberDiff line numberDiff line change
@@ -157,6 +157,68 @@ contains
157157
end do
158158
#:endcall GPU_PARALLEL_LOOP
159159

160+
#:call GPU_HOST_DATA(use_device_addr='[data_real_gpu, data_cmplx_gpu, data_fltr_cmplx_gpu]')
161+
#if defined(__PGI)
162+
ierr = cufftExecD2Z(fwd_plan_gpu, data_real_gpu, data_cmplx_gpu)
163+
#else
164+
ierr = hipfftExecD2Z(fwd_plan_gpu, data_real_gpu, data_cmplx_gpu)
165+
call hipCheck(hipDeviceSynchronize())
166+
#endif
167+
#:endcall GPU_HOST_DATA
168+
Nfq = 3
169+
$:GPU_UPDATE(device='[Nfq]')
170+
171+
#:call GPU_PARALLEL_LOOP(collapse=3)
172+
do k = 1, sys_size
173+
do j = 0, m
174+
do l = 1, Nfq
175+
data_fltr_cmplx_gpu(l + j*cmplx_size + (k - 1)*cmplx_size*x_size) = data_cmplx_gpu(l + j*cmplx_size + (k - 1)*cmplx_size*x_size)
176+
end do
177+
end do
178+
end do
179+
#:endcall GPU_PARALLEL_LOOP
180+
181+
#:call GPU_HOST_DATA(use_device_addr='[data_real_gpu, data_cmplx_gpu, data_fltr_cmplx_gpu]')
182+
#if defined(__PGI)
183+
ierr = cufftExecZ2D(bwd_plan_gpu, data_fltr_cmplx_gpu, data_real_gpu)
184+
#else
185+
ierr = hipfftExecZ2D(bwd_plan_gpu, data_fltr_cmplx_gpu, data_real_gpu)
186+
call hipCheck(hipDeviceSynchronize())
187+
#endif
188+
#:endcall GPU_HOST_DATA
189+
190+
#:call GPU_PARALLEL_LOOP(collapse=3)
191+
do k = 1, sys_size
192+
do j = 0, m
193+
do l = 0, p
194+
data_real_gpu(l + j*real_size + 1 + (k - 1)*real_size*x_size) = data_real_gpu(l + j*real_size + 1 + (k - 1)*real_size*x_size)/real(real_size, dp)
195+
q_cons_vf(k)%sf(j, 0, l) = data_real_gpu(l + j*real_size + 1 + (k - 1)*real_size*x_size)
196+
end do
197+
end do
198+
end do
199+
#:endcall GPU_PARALLEL_LOOP
200+
201+
do i = 1, fourier_rings
202+
203+
#:call GPU_PARALLEL_LOOP(collapse=3)
204+
do k = 1, sys_size
205+
do j = 0, m
206+
do l = 1, cmplx_size
207+
data_fltr_cmplx_gpu(l + j*cmplx_size + (k - 1)*cmplx_size*x_size) = (0_dp, 0_dp)
208+
end do
209+
end do
210+
end do
211+
#:endcall GPU_PARALLEL_LOOP
212+
213+
#:call GPU_PARALLEL_LOOP(collapse=3, firstprivate='[i]')
214+
do k = 1, sys_size
215+
do j = 0, m
216+
do l = 0, p
217+
data_real_gpu(l + j*real_size + 1 + (k - 1)*real_size*x_size) = q_cons_vf(k)%sf(j, i, l)
218+
end do
219+
end do
220+
end do
221+
#:endcall GPU_PARALLEL_LOOP
160222

161223
#:call GPU_HOST_DATA(use_device_addr='[data_real_gpu, data_cmplx_gpu, data_fltr_cmplx_gpu]')
162224
#if defined(__PGI)
@@ -166,7 +228,8 @@ contains
166228
call hipCheck(hipDeviceSynchronize())
167229
#endif
168230
#:endcall GPU_HOST_DATA
169-
Nfq = 3
231+
232+
Nfq = min(floor(2_dp*real(i, dp)*pi), cmplx_size)
170233
$:GPU_UPDATE(device='[Nfq]')
171234

172235
#:call GPU_PARALLEL_LOOP(collapse=3)
@@ -188,81 +251,17 @@ contains
188251
#endif
189252
#:endcall GPU_HOST_DATA
190253

191-
#:call GPU_PARALLEL_LOOP(collapse=3)
254+
#:call GPU_PARALLEL_LOOP(collapse=3, firstprivate='[i]')
192255
do k = 1, sys_size
193256
do j = 0, m
194257
do l = 0, p
195258
data_real_gpu(l + j*real_size + 1 + (k - 1)*real_size*x_size) = data_real_gpu(l + j*real_size + 1 + (k - 1)*real_size*x_size)/real(real_size, dp)
196-
q_cons_vf(k)%sf(j, 0, l) = data_real_gpu(l + j*real_size + 1 + (k - 1)*real_size*x_size)
259+
q_cons_vf(k)%sf(j, i, l) = data_real_gpu(l + j*real_size + 1 + (k - 1)*real_size*x_size)
197260
end do
198261
end do
199262
end do
200263
#:endcall GPU_PARALLEL_LOOP
201-
202-
do i = 1, fourier_rings
203-
204-
#:call GPU_PARALLEL_LOOP(collapse=3)
205-
do k = 1, sys_size
206-
do j = 0, m
207-
do l = 1, cmplx_size
208-
data_fltr_cmplx_gpu(l + j*cmplx_size + (k - 1)*cmplx_size*x_size) = (0_dp, 0_dp)
209-
end do
210-
end do
211-
end do
212-
#:endcall GPU_PARALLEL_LOOP
213-
214-
#:call GPU_PARALLEL_LOOP(collapse=3, firstprivate='[i]')
215-
do k = 1, sys_size
216-
do j = 0, m
217-
do l = 0, p
218-
data_real_gpu(l + j*real_size + 1 + (k - 1)*real_size*x_size) = q_cons_vf(k)%sf(j, i, l)
219-
end do
220-
end do
221-
end do
222-
#:endcall GPU_PARALLEL_LOOP
223-
224-
#:call GPU_HOST_DATA(use_device_addr='[data_real_gpu, data_cmplx_gpu, data_fltr_cmplx_gpu]')
225-
#if defined(__PGI)
226-
ierr = cufftExecD2Z(fwd_plan_gpu, data_real_gpu, data_cmplx_gpu)
227-
#else
228-
ierr = hipfftExecD2Z(fwd_plan_gpu, data_real_gpu, data_cmplx_gpu)
229-
call hipCheck(hipDeviceSynchronize())
230-
#endif
231-
#:endcall GPU_HOST_DATA
232-
233-
Nfq = min(floor(2_dp*real(i, dp)*pi), cmplx_size)
234-
$:GPU_UPDATE(device='[Nfq]')
235-
236-
#:call GPU_PARALLEL_LOOP(collapse=3)
237-
do k = 1, sys_size
238-
do j = 0, m
239-
do l = 1, Nfq
240-
data_fltr_cmplx_gpu(l + j*cmplx_size + (k - 1)*cmplx_size*x_size) = data_cmplx_gpu(l + j*cmplx_size + (k - 1)*cmplx_size*x_size)
241-
end do
242-
end do
243-
end do
244-
#:endcall GPU_PARALLEL_LOOP
245-
246-
#:call GPU_HOST_DATA(use_device_addr='[data_real_gpu, data_cmplx_gpu, data_fltr_cmplx_gpu]')
247-
#if defined(__PGI)
248-
ierr = cufftExecZ2D(bwd_plan_gpu, data_fltr_cmplx_gpu, data_real_gpu)
249-
#else
250-
ierr = hipfftExecZ2D(bwd_plan_gpu, data_fltr_cmplx_gpu, data_real_gpu)
251-
call hipCheck(hipDeviceSynchronize())
252-
#endif
253-
#:endcall GPU_HOST_DATA
254-
255-
#:call GPU_PARALLEL_LOOP(collapse=3, firstprivate='[i]')
256-
do k = 1, sys_size
257-
do j = 0, m
258-
do l = 0, p
259-
data_real_gpu(l + j*real_size + 1 + (k - 1)*real_size*x_size) = data_real_gpu(l + j*real_size + 1 + (k - 1)*real_size*x_size)/real(real_size, dp)
260-
q_cons_vf(k)%sf(j, i, l) = data_real_gpu(l + j*real_size + 1 + (k - 1)*real_size*x_size)
261-
end do
262-
end do
263-
end do
264-
#:endcall GPU_PARALLEL_LOOP
265-
end do
264+
end do
266265

267266
#else
268267
Nfq = 3

‎src/simulation/m_rhs.fpp‎

Lines changed: 4 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -249,7 +249,7 @@ contains
249249
if (.not. igr) then
250250
@:ALLOCATE(flux_n(1:num_dims))
251251
@:ALLOCATE(flux_src_n(1:num_dims))
252-
@:ALLOCATE(flux_gsrc_n(1:num_dims))
252+
@:ALLOCATE(flux_gsrc_n(1:num_dims))
253253

254254
do i = 1, num_dims
255255

@@ -1010,9 +1010,9 @@ contains
10101010
if (acoustic_source) then
10111011
call nvtxStartRange("RHS-ACOUSTIC-SRC")
10121012
call s_acoustic_src_calculations(q_cons_qp%vf(1:sys_size), &
1013-
q_prim_qp%vf(1:sys_size), &
1014-
t_step, &
1015-
rhs_vf)
1013+
q_prim_qp%vf(1:sys_size), &
1014+
t_step, &
1015+
rhs_vf)
10161016
call nvtxEndRange
10171017
end if
10181018

‎src/simulation/m_start_up.fpp‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1219,7 +1219,7 @@ contains
12191219

12201220
stor = 1
12211221

1222-
if(time_stepper /= 1) then
1222+
if (time_stepper /= 1) then
12231223
#:call GPU_PARALLEL_LOOP(collapse=4, copyin='[idwbuff]')
12241224
do i = 1, sys_size
12251225
do l = idwbuff(3)%beg, idwbuff(3)%end

0 commit comments

Comments
 (0)