Skip to content

Commit 2cad739

Browse files
prathi-windprathi-wind
committed
Ran formatter
1 parent d15e2ff commit 2cad739

File tree

1 file changed

+58
-58
lines changed

1 file changed

+58
-58
lines changed

‎src/simulation/m_fftw.fpp‎

Lines changed: 58 additions & 58 deletions
Original file line numberDiff line numberDiff line change
@@ -161,119 +161,119 @@ contains
161161
end do
162162
#:endcall GPU_PARALLEL_LOOP
163163

164-
#:if not USING_NVHPC
165-
p_real => data_real_gpu
166-
p_cmplx => data_cmplx_gpu
167-
p_fltr_cmplx => data_fltr_cmplx_gpu
168-
#:endif
164+
#:if not USING_NVHPC
165+
p_real => data_real_gpu
166+
p_cmplx => data_cmplx_gpu
167+
p_fltr_cmplx => data_fltr_cmplx_gpu
168+
#:endif
169169

170170
#:call GPU_DATA(attach='[p_real, p_cmplx, p_fltr_cmplx]')
171171
#:call GPU_HOST_DATA(use_device_ptr='[p_real, p_cmplx, p_fltr_cmplx]')
172172
#if defined(__PGI)
173-
ierr = cufftExecD2Z(fwd_plan_gpu, data_real_gpu, data_cmplx_gpu)
173+
ierr = cufftExecD2Z(fwd_plan_gpu, data_real_gpu, data_cmplx_gpu)
174174
#else
175175
ierr = hipfftExecD2Z(fwd_plan_gpu, c_loc(p_real), c_loc(p_cmplx))
176176
call hipCheck(hipDeviceSynchronize())
177177
#endif
178178
#:endcall GPU_HOST_DATA
179-
Nfq = 3
180-
$:GPU_UPDATE(device='[Nfq]')
179+
Nfq = 3
180+
$:GPU_UPDATE(device='[Nfq]')
181181

182-
#:call GPU_PARALLEL_LOOP(collapse=3)
183-
do k = 1, sys_size
184-
do j = 0, m
185-
do l = 1, Nfq
186-
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)
182+
#:call GPU_PARALLEL_LOOP(collapse=3)
183+
do k = 1, sys_size
184+
do j = 0, m
185+
do l = 1, Nfq
186+
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)
187+
end do
187188
end do
188189
end do
189-
end do
190-
#:endcall GPU_PARALLEL_LOOP
190+
#:endcall GPU_PARALLEL_LOOP
191191

192192
#:call GPU_HOST_DATA(use_device_ptr='[p_real, p_fltr_cmplx]')
193193
#if defined(__PGI)
194-
ierr = cufftExecZ2D(bwd_plan_gpu, data_fltr_cmplx_gpu, data_real_gpu)
194+
ierr = cufftExecZ2D(bwd_plan_gpu, data_fltr_cmplx_gpu, data_real_gpu)
195195
#else
196196
ierr = hipfftExecZ2D(bwd_plan_gpu, c_loc(p_fltr_cmplx), c_loc(p_real))
197197
call hipCheck(hipDeviceSynchronize())
198198
#endif
199199
#:endcall GPU_HOST_DATA
200200

201-
#:call GPU_PARALLEL_LOOP(collapse=3)
202-
do k = 1, sys_size
203-
do j = 0, m
204-
do l = 0, p
205-
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)
206-
q_cons_vf(k)%sf(j, 0, l) = data_real_gpu(l + j*real_size + 1 + (k - 1)*real_size*x_size)
207-
end do
208-
end do
209-
end do
210-
#:endcall GPU_PARALLEL_LOOP
211-
212-
do i = 1, fourier_rings
213-
214201
#:call GPU_PARALLEL_LOOP(collapse=3)
215202
do k = 1, sys_size
216203
do j = 0, m
217-
do l = 1, cmplx_size
218-
data_fltr_cmplx_gpu(l + j*cmplx_size + (k - 1)*cmplx_size*x_size) = (0_dp, 0_dp)
204+
do l = 0, p
205+
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)
206+
q_cons_vf(k)%sf(j, 0, l) = data_real_gpu(l + j*real_size + 1 + (k - 1)*real_size*x_size)
219207
end do
220208
end do
221209
end do
222210
#:endcall GPU_PARALLEL_LOOP
223211

224-
#:call GPU_PARALLEL_LOOP(collapse=3, firstprivate='[i]')
225-
do k = 1, sys_size
226-
do j = 0, m
227-
do l = 0, p
228-
data_real_gpu(l + j*real_size + 1 + (k - 1)*real_size*x_size) = q_cons_vf(k)%sf(j, i, l)
212+
do i = 1, fourier_rings
213+
214+
#:call GPU_PARALLEL_LOOP(collapse=3)
215+
do k = 1, sys_size
216+
do j = 0, m
217+
do l = 1, cmplx_size
218+
data_fltr_cmplx_gpu(l + j*cmplx_size + (k - 1)*cmplx_size*x_size) = (0_dp, 0_dp)
219+
end do
229220
end do
230221
end do
231-
end do
232-
#:endcall GPU_PARALLEL_LOOP
222+
#:endcall GPU_PARALLEL_LOOP
223+
224+
#:call GPU_PARALLEL_LOOP(collapse=3, firstprivate='[i]')
225+
do k = 1, sys_size
226+
do j = 0, m
227+
do l = 0, p
228+
data_real_gpu(l + j*real_size + 1 + (k - 1)*real_size*x_size) = q_cons_vf(k)%sf(j, i, l)
229+
end do
230+
end do
231+
end do
232+
#:endcall GPU_PARALLEL_LOOP
233233

234234
#:call GPU_HOST_DATA(use_device_ptr='[p_real, p_cmplx]')
235235
#if defined(__PGI)
236-
ierr = cufftExecD2Z(fwd_plan_gpu, data_real_gpu, data_cmplx_gpu)
236+
ierr = cufftExecD2Z(fwd_plan_gpu, data_real_gpu, data_cmplx_gpu)
237237
#else
238238
ierr = hipfftExecD2Z(fwd_plan_gpu, c_loc(p_real), c_loc(p_cmplx))
239239
call hipCheck(hipDeviceSynchronize())
240240
#endif
241241
#:endcall GPU_HOST_DATA
242242

243-
Nfq = min(floor(2_dp*real(i, dp)*pi), cmplx_size)
244-
$:GPU_UPDATE(device='[Nfq]')
243+
Nfq = min(floor(2_dp*real(i, dp)*pi), cmplx_size)
244+
$:GPU_UPDATE(device='[Nfq]')
245245

246-
#:call GPU_PARALLEL_LOOP(collapse=3)
247-
do k = 1, sys_size
248-
do j = 0, m
249-
do l = 1, Nfq
250-
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)
246+
#:call GPU_PARALLEL_LOOP(collapse=3)
247+
do k = 1, sys_size
248+
do j = 0, m
249+
do l = 1, Nfq
250+
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)
251+
end do
251252
end do
252253
end do
253-
end do
254-
#:endcall GPU_PARALLEL_LOOP
254+
#:endcall GPU_PARALLEL_LOOP
255255

256256
#:call GPU_HOST_DATA(use_device_ptr='[p_real, p_fltr_cmplx]')
257257
#if defined(__PGI)
258-
ierr = cufftExecZ2D(bwd_plan_gpu, data_fltr_cmplx_gpu, data_real_gpu)
258+
ierr = cufftExecZ2D(bwd_plan_gpu, data_fltr_cmplx_gpu, data_real_gpu)
259259
#else
260260
ierr = hipfftExecZ2D(bwd_plan_gpu, c_loc(p_fltr_cmplx), c_loc(p_real))
261261
call hipCheck(hipDeviceSynchronize())
262262
#endif
263263
#:endcall GPU_HOST_DATA
264264

265-
#:call GPU_PARALLEL_LOOP(collapse=3, firstprivate='[i]')
266-
do k = 1, sys_size
267-
do j = 0, m
268-
do l = 0, p
269-
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)
270-
q_cons_vf(k)%sf(j, i, l) = data_real_gpu(l + j*real_size + 1 + (k - 1)*real_size*x_size)
265+
#:call GPU_PARALLEL_LOOP(collapse=3, firstprivate='[i]')
266+
do k = 1, sys_size
267+
do j = 0, m
268+
do l = 0, p
269+
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)
270+
q_cons_vf(k)%sf(j, i, l) = data_real_gpu(l + j*real_size + 1 + (k - 1)*real_size*x_size)
271+
end do
271272
end do
272273
end do
273-
end do
274-
#:endcall GPU_PARALLEL_LOOP
274+
#:endcall GPU_PARALLEL_LOOP
275275

276-
end do
276+
end do
277277
#:endcall GPU_DATA
278278

279279
#else

0 commit comments

Comments
 (0)