This repository has been archived by the owner on Jun 20, 2024. It is now read-only.
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathadvec_cell_cuda_kernels.cu
executable file
·345 lines (302 loc) · 11.3 KB
/
advec_cell_cuda_kernels.cu
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
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
/*Crown Copyright 2012 AWE.
*
* This file is part of CloverLeaf.
*
* CloverLeaf is free software: you can redistribute it and/or modify it under
* the terms of the GNU General Public License as published by the
* Free Software Foundation, either version 3 of the License, or (at your option)
* any later version.
*
* CloverLeaf is distributed in the hope that it will be useful, but
* WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
* FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more
* details.
*
* You should have received a copy of the GNU General Public License along with
* CloverLeaf. If not, see http://www.gnu.org/licenses/.
*/
/*
* @brief CUDA cell advection kernel
* @author Michael Boulton NVIDIA COORP
* @details Performs a second order advective remap using van-Leer limiting
* with directional splitting.
*/
#include "cuda_common.cu"
#define _SHARED_KERNEL_ARGS_ \
const int x_min, \
const int x_max, \
const int y_min, \
const int y_max, \
const int swp_nmbr, \
const double* __restrict const volume, \
const double* __restrict const vol_flux_x, \
const double* __restrict const vol_flux_y, \
const double* __restrict const pre_vol, \
double* __restrict const density1, \
double* __restrict const energy1, \
double* __restrict const ener_flux
__global__ void device_pre_vol_kernel_x
(const int x_min, const int x_max,
const int y_min, const int y_max,
const int swp_nmbr,
double* __restrict const pre_vol,
double* __restrict const post_vol,
const double* __restrict const volume,
const double* __restrict const vol_flux_x,
const double* __restrict const vol_flux_y)
{
__kernel_indexes;
if (row >= (y_min + 1) - 2 && row <= (y_max + 1) + 2
&& column >= (x_min + 1) - 2 && column <= (x_max + 1) + 2)
{
if (swp_nmbr == 1)
{
pre_vol[THARR2D(0, 0, 1)] = volume[THARR2D(0, 0, 0)]
+(vol_flux_x[THARR2D(1, 0, 1)] - vol_flux_x[THARR2D(0, 0, 1)]
+ vol_flux_y[THARR2D(0, 1, 0)] - vol_flux_y[THARR2D(0, 0, 0)]);
post_vol[THARR2D(0, 0, 1)] = pre_vol[THARR2D(0, 0, 1)]
- (vol_flux_x[THARR2D(1, 0, 1)] - vol_flux_x[THARR2D(0, 0, 1)]);
}
else
{
pre_vol[THARR2D(0, 0, 1)] = volume[THARR2D(0, 0, 0)]
+ vol_flux_x[THARR2D(1, 0, 1)] - vol_flux_x[THARR2D(0, 0, 1)];
post_vol[THARR2D(0, 0, 1)] = volume[THARR2D(0, 0, 0)];
}
}
}
__global__ void device_ener_flux_kernel_x
(_SHARED_KERNEL_ARGS_,
const double* __restrict const vertexdx,
double* __restrict const mass_flux_x)
{
__kernel_indexes;
double sigmat, sigmam, sigma3, sigma4, diffuw, diffdw, limiter;
int upwind, donor, downwind, dif;
const double one_by_six = 1.0/6.0;
//
// if cell is within x area:
// +++++++++++++++++++++
// +++++++++++++++++++++
// ++xxxxxxxxxxxxxxxxxxx
// +++++++++++++++++++++
// +++++++++++++++++++++
//
if (row >= (y_min + 1) && row <= (y_max + 1)
&& column >= (x_min + 1) && column <= (x_max + 1) + 2)
{
// if flowing right
if (vol_flux_x[THARR2D(0, 0, 1)] > 0.0)
{
upwind = -2;
donor = -1;
downwind = 0;
dif = donor;
}
else
{
// tries to get from below, unless it would be reading from a cell
// which would be off the right, in which case read from cur cell
upwind = (column == (x_max + 1) + 2) ? 0 : 1;
donor = 0;
downwind = -1;
dif = upwind;
}
sigmat = fabs(vol_flux_x[THARR2D(0, 0, 1)]) / pre_vol[THARR2D(donor, 0, 1)];
sigma3 = (1.0 + sigmat) * (vertexdx[column] / vertexdx[column + dif]);
sigma4 = 2.0 - sigmat;
diffuw = density1[THARR2D(donor, 0, 0)] - density1[THARR2D(upwind, 0, 0)];
diffdw = density1[THARR2D(downwind, 0, 0)] - density1[THARR2D(donor, 0, 0)];
if (diffuw * diffdw > 0.0)
{
limiter = (1.0 - sigmat) * SIGN(1.0, diffdw)
* MIN(fabs(diffuw), MIN(fabs(diffdw), one_by_six
* (sigma3 * fabs(diffuw) + sigma4 * fabs(diffdw))));
}
else
{
limiter = 0.0;
}
mass_flux_x[THARR2D(0, 0, 1)] = vol_flux_x[THARR2D(0, 0, 1)]
* (density1[THARR2D(donor, 0, 0)] + limiter);
sigmam = fabs(mass_flux_x[THARR2D(0, 0, 1)])
/ (density1[THARR2D(donor, 0, 0)] * pre_vol[THARR2D(donor, 0, 1)]);
diffuw = energy1[THARR2D(donor, 0, 0)] - energy1[THARR2D(upwind, 0, 0)];
diffdw = energy1[THARR2D(downwind, 0, 0)] - energy1[THARR2D(donor, 0, 0)];
if (diffuw * diffdw > 0.0)
{
limiter = (1.0 - sigmam) * SIGN(1.0, diffdw)
* MIN(fabs(diffuw), MIN(fabs(diffdw), one_by_six
* (sigma3 * fabs(diffuw) + sigma4 * fabs(diffdw))));
}
else
{
limiter = 0.0;
}
ener_flux[THARR2D(0, 0, 0)] = mass_flux_x[THARR2D(0, 0, 1)]
* (energy1[THARR2D(donor, 0, 0)] + limiter);
}
}
__global__ void device_advec_cell_kernel_x
(_SHARED_KERNEL_ARGS_,
const double* __restrict const mass_flux_x)
{
__kernel_indexes;
double pre_mass, post_mass, advec_vol, post_ener;
//
// if cell is within x area:
// +++++++++++++++++++++
// +++++++++++++++++++++
// ++xxxxxxxxxxxxxxxxx++
// +++++++++++++++++++++
// +++++++++++++++++++++
//
if (row >= (y_min + 1) && row <= (y_max + 1)
&& column >= (x_min + 1) && column <= (x_max + 1))
{
pre_mass = density1[THARR2D(0, 0, 0)] * pre_vol[THARR2D(0, 0, 1)];
post_mass = pre_mass + mass_flux_x[THARR2D(0, 0, 1)]
- mass_flux_x[THARR2D(1, 0, 1)];
post_ener = (energy1[THARR2D(0, 0, 0)] * pre_mass
+ ener_flux[THARR2D(0, 0, 0)] - ener_flux[THARR2D(1, 0, 0)])
/ post_mass;
advec_vol = pre_vol[THARR2D(0, 0, 1)] + vol_flux_x[THARR2D(0, 0, 1)]
- vol_flux_x[THARR2D(1, 0, 1)];
density1[THARR2D(0, 0, 0)] = post_mass / advec_vol;
energy1[THARR2D(0, 0, 0)] = post_ener;
}
}
//////////////////////////////////////////////////////////////////////////
//y kernels
__global__ void device_pre_vol_kernel_y
(const int x_min, const int x_max,
const int y_min, const int y_max,
const int swp_nmbr,
double* __restrict const pre_vol,
double* __restrict const post_vol,
const double* __restrict const volume,
const double* __restrict const vol_flux_x,
const double* __restrict const vol_flux_y)
{
__kernel_indexes;
if (row >= (y_min + 1) - 2 && row <= (y_max + 1) + 2
&& column >= (x_min + 1) - 2 && column <= (x_max + 1) + 2)
{
if (swp_nmbr == 1)
{
pre_vol[THARR2D(0, 0, 1)] = volume[THARR2D(0, 0, 0)]
+(vol_flux_y[THARR2D(0, 1, 0)] - vol_flux_y[THARR2D(0, 0, 0)]
+ vol_flux_x[THARR2D(1, 0, 1)] - vol_flux_x[THARR2D(0, 0, 1)]);
post_vol[THARR2D(0, 0, 1)] = pre_vol[THARR2D(0, 0, 1)]
- (vol_flux_y[THARR2D(0, 1, 0)] - vol_flux_y[THARR2D(0, 0, 1)]);
}
else
{
pre_vol[THARR2D(0, 0, 1)] = volume[THARR2D(0, 0, 0)]
+ vol_flux_y[THARR2D(0, 1, 0)] - vol_flux_y[THARR2D(0, 0, 0)];
post_vol[THARR2D(0, 0, 1)] = volume[THARR2D(0, 0, 0)];
}
}
}
__global__ void device_ener_flux_kernel_y
(_SHARED_KERNEL_ARGS_,
const double* __restrict const vertexdy,
double* __restrict const mass_flux_y)
{
__kernel_indexes;
double sigmat, sigmam, sigma3, sigma4, diffuw, diffdw, limiter;
int upwind, donor, downwind, dif;
const double one_by_six = 1.0/6.0;
//
// if cell is within x area:
// +++++++++++++++++++++
// +++++++++++++++++++++
// ++xxxxxxxxxxxxxxxxx++
// ++xxxxxxxxxxxxxxxxx++
//
if (row >= (y_min + 1) && row <= (y_max + 1) + 2
&& column >= (x_min + 1) && column <= (x_max + 1))
{
// if flowing right
if (vol_flux_y[THARR2D(0, 0, 0)] > 0.0)
{
upwind = -2;
donor = -1;
downwind = 0;
dif = donor;
}
else
{
//
// tries to get from below, unless it would be reading from a cell
// which would be off the bottom, in which case read from cur cell
//
upwind = (row == (y_max + 1) + 2) ? 0 : 1;
donor = 0;
downwind = -1;
dif = downwind;
}
sigmat = fabs(vol_flux_y[THARR2D(0, 0, 0)]) / pre_vol[THARR2D(0, donor, 1)];
sigma3 = (1.0 + sigmat) * (vertexdy[row] / vertexdy[row + dif]);
sigma4 = 2.0 - sigmat;
diffuw = density1[THARR2D(0, donor, 0)] - density1[THARR2D(0, upwind, 0)];
diffdw = density1[THARR2D(0, downwind, 0)] - density1[THARR2D(0, donor, 0)];
if (diffuw * diffdw > 0.0)
{
limiter = (1.0 - sigmat) * SIGN(1.0, diffdw)
* MIN(fabs(diffuw), MIN(fabs(diffdw), one_by_six
* (sigma3 * fabs(diffuw) + sigma4 * fabs(diffdw))));
}
else
{
limiter = 0.0;
}
mass_flux_y[THARR2D(0, 0, 0)] = vol_flux_y[THARR2D(0, 0, 0)]
* (density1[THARR2D(0, donor, 0)] + limiter);
sigmam = fabs(mass_flux_y[THARR2D(0, 0, 0)])
/ (density1[THARR2D(0, donor, 0)] * pre_vol[THARR2D(0, donor, 1)]);
diffuw = energy1[THARR2D(0, donor, 0)] - energy1[THARR2D(0, upwind, 0)];
diffdw = energy1[THARR2D(0, downwind, 0)] - energy1[THARR2D(0, donor, 0)];
if (diffuw * diffdw > 0.0)
{
limiter = (1.0 - sigmam) * SIGN(1.0, diffdw)
* MIN(fabs(diffuw), MIN(fabs(diffdw), one_by_six
* (sigma3 * fabs(diffuw) + sigma4 * fabs(diffdw))));
}
else
{
limiter = 0.0;
}
ener_flux[THARR2D(0, 0, 0)] = mass_flux_y[THARR2D(0, 0, 0)]
* (energy1[THARR2D(0, donor, 0)] + limiter);
}
}
__global__ void device_advec_cell_kernel_y
(_SHARED_KERNEL_ARGS_,
const double* __restrict const mass_flux_y)
{
__kernel_indexes;
double pre_mass, post_mass, advec_vol, post_ener;
//
// if cell is within x area:
// +++++++++++++++++++++
// +++++++++++++++++++++
// ++xxxxxxxxxxxxxxxxx++
// +++++++++++++++++++++
// +++++++++++++++++++++
//
if (row >= (y_min + 1) && row <= (y_max + 1)
&& column >= (x_min + 1) && column <= (x_max + 1))
{
pre_mass = density1[THARR2D(0, 0, 0)] * pre_vol[THARR2D(0, 0, 1)];
post_mass = pre_mass + mass_flux_y[THARR2D(0, 0, 0)]
- mass_flux_y[THARR2D(0, 1, 0)];
post_ener = (energy1[THARR2D(0, 0, 0)] * pre_mass
+ ener_flux[THARR2D(0, 0, 0)] - ener_flux[THARR2D(0, 1, 0)])
/ post_mass;
advec_vol = pre_vol[THARR2D(0, 0, 1)] + vol_flux_y[THARR2D(0, 0, 0)]
- vol_flux_y[THARR2D(0, 1, 0)];
density1[THARR2D(0, 0, 0)] = post_mass / advec_vol;
energy1[THARR2D(0, 0, 0)] = post_ener;
}
}