forked from stanfordhpccenter/HTR-solver
-
Notifications
You must be signed in to change notification settings - Fork 0
/
prometeo_metric.hpp
282 lines (262 loc) · 13 KB
/
prometeo_metric.hpp
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
// Copyright (c) "2019, by Stanford University
// Developer: Mario Di Renzo
// Affiliation: Center for Turbulence Research, Stanford University
// URL: https://ctr.stanford.edu
// Citation: Di Renzo, M., Lin, F., and Urzay, J. (2020).
// HTR solver: An open-source exascale-oriented task-based
// multi-GPU high-order code for hypersonic aerothermodynamics.
// Computer Physics Communications 255, 107262"
// All rights reserved.
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are met:
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above copyright
// notice, this list of conditions and the following disclaimer in the
// documentation and/or other materials provided with the distribution.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
// ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
// WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
// DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE FOR ANY
// DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
// (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
// LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
// ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
// SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
#ifndef __PROMETEO_METRIC_HPP__
#define __PROMETEO_METRIC_HPP__
#include "legion.h"
using namespace Legion;
//-----------------------------------------------------------------------------
// LOAD PROMETEO UTILITIES AND MODULES
//-----------------------------------------------------------------------------
#include "prometeo_types.h"
#include "task_helper.h"
#include "prometeo_metric.h"
#include "prometeo_metric.inl"
//-----------------------------------------------------------------------------
// UTILITY TO UNWARP COORDIATNES AROUND PERIODIC BCS
//-----------------------------------------------------------------------------
template<direction dir>
__CUDA_H__
inline double unwarpCoordinate(double x, const double w, const int off,
Point<3> p, const Rect<3> bounds);
template<>
__CUDA_H__
inline double unwarpCoordinate<Xdir>(double x, const double w, const int off,
Point<3> p, const Rect<3> bounds) {
p.x += off;
// if we are below the lower bound start by shifting:
// - x by -1 width
// - the index by 1
if (p.x < bounds.lo.x) { x -= w; p.x += 1; }
return x + w*coord_t((p.x - bounds.lo.x)/(bounds.hi.x - bounds.lo.x + 1));
}
template<>
__CUDA_H__
inline double unwarpCoordinate<Ydir>(double y, const double w, const int off,
Point<3> p, const Rect<3> bounds) {
p.y += off;
// if we are below the lower bound start by shifting:
// - y by -1 width
// - the index by 1
if (p.y < bounds.lo.y) { y -= w; p.y += 1; }
return y + w*coord_t((p.y - bounds.lo.y)/(bounds.hi.y - bounds.lo.y + 1));
}
template<>
__CUDA_H__
inline double unwarpCoordinate<Zdir>(double z, const double w, const int off,
Point<3> p, const Rect<3> bounds) {
p.z += off;
// if we are below the lower bound start by shifting:
// - z by -1 width
// - the index by 1
if (p.z < bounds.lo.z) { z -= w; p.z += 1; }
return z + w*coord_t((p.z - bounds.lo.z)/(bounds.hi.z - bounds.lo.z + 1));
}
//-----------------------------------------------------------------------------
// TASK THAT COMPUTES THE METRIC OF THE INTERNAL POINTS OF THE GRID
//-----------------------------------------------------------------------------
class InitializeMetricTask {
public:
struct Args {
uint64_t arg_mask[1];
LogicalRegion MetricGhost;
LogicalRegion XGhost;
LogicalRegion YGhost;
LogicalRegion ZGhost;
LogicalRegion Fluid;
Rect<3> Fluid_bounds;
double Grid_xWidth;
double Grid_yWidth;
double Grid_zWidth;
FieldID MetricGhost_fields [FID_last - 101];
FieldID XGhost_fields [FID_last - 101];
FieldID YGhost_fields [FID_last - 101];
FieldID ZGhost_fields [FID_last - 101];
FieldID Fluid_fields [FID_last - 101];
};
public:
static const char * const TASK_NAME;
static const int TASK_ID;
static const bool CPU_BASE_LEAF = true;
static const bool GPU_BASE_LEAF = true;
static const int MAPPER_ID = 0;
public:
// Direction dependent quantities
template<direction dir>
__CUDA_H__
static inline double reconstructCoordEuler(const AccessorRO<Vec3, 3> centerCoordinates,
const Point<3> p,
const double width,
const int nType,
const coord_t dsize,
const Rect<3> bounds) {
// TODO: implement some sort of static_if
int iN;
if (dir == Xdir) iN = 0;
else if (dir == Ydir) iN = 1;
else if (dir == Zdir) iN = 2;
// Compute stencil points
const Point<3> pM2 = warpPeriodic<dir, Minus>(bounds, p, dsize, offM2(nType));
const Point<3> pM1 = warpPeriodic<dir, Minus>(bounds, p, dsize, offM1(nType));
const Point<3> pP1 = warpPeriodic<dir, Plus >(bounds, p, dsize, offP1(nType));
const Point<3> pP2 = warpPeriodic<dir, Plus >(bounds, p, dsize, offP2(nType));
const Point<3> pP3 = warpPeriodic<dir, Plus >(bounds, p, dsize, offP3(nType));
return LinearReconstruct(
unwarpCoordinate<dir>(centerCoordinates[pM2][iN], width, offM2(nType), p, bounds),
unwarpCoordinate<dir>(centerCoordinates[pM1][iN], width, offM1(nType), p, bounds),
centerCoordinates[p ][iN],
unwarpCoordinate<dir>(centerCoordinates[pP1][iN], width, offP1(nType), p, bounds),
unwarpCoordinate<dir>(centerCoordinates[pP2][iN], width, offP2(nType), p, bounds),
unwarpCoordinate<dir>(centerCoordinates[pP3][iN], width, offP3(nType), p, bounds),
nType);
}
template<direction dir>
__CUDA_H__
static inline void ComputeDiffusionMetrics(const AccessorRW<double, 3> m_d,
const AccessorRW<double, 3> m_s,
const AccessorRO< Vec3, 3> centerCoordinates,
const Point<3> p,
const double width,
const int nType,
const coord_t dsize,
const Rect<3> bounds) {
// TODO: implement some sort of static_if
int iN;
if (dir == Xdir) iN = 0;
else if (dir == Ydir) iN = 1;
else if (dir == Zdir) iN = 2;
const Point<3> pM1 = warpPeriodic<dir, Minus>(bounds, p, dsize, offM1(nType));
const Point<3> pP1 = warpPeriodic<dir, Plus >(bounds, p, dsize, offP1(nType));
// Compute staggered metric for viscous fluxes
m_s[p] = 1.0/(unwarpCoordinate<dir>(centerCoordinates[pP1][iN], width, 1, p, bounds) -
centerCoordinates[p ][iN]);
// Compute collocated metric for viscous fluxes
m_d[p] = 1.0/(getDeriv(nType,
unwarpCoordinate<dir>(centerCoordinates[pM1][iN], width, -1, p, bounds),
centerCoordinates[p ][iN],
unwarpCoordinate<dir>(centerCoordinates[pP1][iN], width, 1, p, bounds), 1.0));
}
public:
static void cpu_base_impl(const Args &args,
const std::vector<PhysicalRegion> ®ions,
const std::vector<Future> &futures,
Context ctx, Runtime *runtime);
#ifdef LEGION_USE_CUDA
static void gpu_base_impl(const Args &args,
const std::vector<PhysicalRegion> ®ions,
const std::vector<Future> &futures,
Context ctx, Runtime *runtime);
#endif
};
//-----------------------------------------------------------------------------
// TASK THAT CORRECTS THE METRIC OF GHOST POINTS
//-----------------------------------------------------------------------------
template<direction dir>
class CorrectGhostMetricTask {
public:
struct Args {
uint64_t arg_mask[1];
LogicalRegion Fluid;
FieldID Fluid_fields [FID_last - 101];
};
public:
static const char * const TASK_NAME;
static const int TASK_ID;
static const bool CPU_BASE_LEAF = true;
static const bool GPU_BASE_LEAF = true;
static const int MAPPER_ID = 0;
private:
// Direction dependent quantities
static const FieldID FID_nType;
static const FieldID FID_m;
public:
__CUDA_H__
static inline void CorrectLeftStaggered(const AccessorRW<double, 3> m,
const AccessorRO< Vec3, 3> centerCoordinates,
const Point<3> p) {
int iN; Point<3> pp1; Point<3> pp2;
if (dir == Xdir) { iN = 0; pp1 = p + Point<3>(1, 0, 0); pp2 = p + Point<3>(2, 0, 0); }
else if (dir == Ydir) { iN = 1; pp1 = p + Point<3>(0, 1, 0); pp2 = p + Point<3>(0, 2, 0); }
else if (dir == Zdir) { iN = 2; pp1 = p + Point<3>(0, 0, 1); pp2 = p + Point<3>(0, 0, 2); }
m[p] = 1.0/(- 8.0/3.0*centerCoordinates[p ][iN]
+ 3.0 *centerCoordinates[pp1][iN]
- 1.0/3.0*centerCoordinates[pp2][iN]);
// The staggered metric should be scaled by 0.5 but we avoid this factor
// for computational efficiency. Remeber this comment when you compute the fluxes
};
__CUDA_H__
static inline void CorrectLeftCollocated(const AccessorRW<double, 3> m,
const AccessorRO< Vec3, 3> centerCoordinates,
const Point<3> p) {
int iN; Point<3> pp1; Point<3> pp2;
if (dir == Xdir) { iN = 0; pp1 = p + Point<3>(1, 0, 0); pp2 = p + Point<3>(2, 0, 0); }
else if (dir == Ydir) { iN = 1; pp1 = p + Point<3>(0, 1, 0); pp2 = p + Point<3>(0, 2, 0); }
else if (dir == Zdir) { iN = 2; pp1 = p + Point<3>(0, 0, 1); pp2 = p + Point<3>(0, 0, 2); }
m[p] = 1.0/(- 1.5*centerCoordinates[p ][iN]
+ 2.0*centerCoordinates[pp1][iN]
- 0.5*centerCoordinates[pp2][iN]);
};
__CUDA_H__
static inline void CorrectRightStaggered(const AccessorRW<double, 3> m,
const AccessorRO< Vec3, 3> centerCoordinates,
const Point<3> p) {
int iN; Point<3> pm1; Point<3> pm2;
if (dir == Xdir) { iN = 0; pm1 = p - Point<3>(1, 0, 0); pm2 = p - Point<3>(2, 0, 0); }
else if (dir == Ydir) { iN = 1; pm1 = p - Point<3>(0, 1, 0); pm2 = p - Point<3>(0, 2, 0); }
else if (dir == Zdir) { iN = 2; pm1 = p - Point<3>(0, 0, 1); pm2 = p - Point<3>(0, 0, 2); }
m[p] = 1.0/( 8.0/3.0*centerCoordinates[p ][iN]
- 3.0 *centerCoordinates[pm1][iN]
+ 1.0/3.0*centerCoordinates[pm2][iN]);
// The staggered metric should be scaled by 0.5 but we avoid this factor
// for computational efficiency. Remeber this comment when you compute the fluxes
};
__CUDA_H__
static inline void CorrectRightCollocated(const AccessorRW<double, 3> m,
const AccessorRO< Vec3, 3> centerCoordinates,
const Point<3> p) {
int iN; Point<3> pm1; Point<3> pm2;
if (dir == Xdir) { iN = 0; pm1 = p - Point<3>(1, 0, 0); pm2 = p - Point<3>(2, 0, 0); }
else if (dir == Ydir) { iN = 1; pm1 = p - Point<3>(0, 1, 0); pm2 = p - Point<3>(0, 2, 0); }
else if (dir == Zdir) { iN = 2; pm1 = p - Point<3>(0, 0, 1); pm2 = p - Point<3>(0, 0, 2); }
m[p] = 1.0/( 1.5*centerCoordinates[p ][iN]
- 2.0*centerCoordinates[pm1][iN]
+ 0.5*centerCoordinates[pm2][iN]);
};
public:
static void cpu_base_impl(const Args &args,
const std::vector<PhysicalRegion> ®ions,
const std::vector<Future> &futures,
Context ctx, Runtime *runtime);
#ifdef LEGION_USE_CUDA
static void gpu_base_impl(const Args &args,
const std::vector<PhysicalRegion> ®ions,
const std::vector<Future> &futures,
Context ctx, Runtime *runtime);
#endif
};
#endif // __PROMETEO_METRIC_HPP__