Neko 1.0.1
A portable framework for high-order spectral element flow simulations
Loading...
Searching...
No Matches
euler_res.c
Go to the documentation of this file.
1/*
2 Copyright (c) 2025, The Neko Authors
3 All rights reserved.
4
5 Redistribution and use in source and binary forms, with or without
6 modification, are permitted provided that the following conditions
7 are met:
8
9 * Redistributions of source code must retain the above copyright
10 notice, this list of conditions and the following disclaimer.
11
12 * Redistributions in binary form must reproduce the above
13 copyright notice, this list of conditions and the following
14 disclaimer in the documentation and/or other materials provided
15 with the distribution.
16
17 * Neither the name of the authors nor the names of its
18 contributors may be used to endorse or promote products derived
19 from this software without specific prior written permission.
20
21 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
22 "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
23 LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
24 FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
25 COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
26 INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
27 BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
28 LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
29 CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
30 LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
31 ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
32 POSSIBILITY OF SUCH DAMAGE.
33*/
34
35#ifdef __APPLE__
36#include <OpenCL/cl.h>
37#else
38#include <CL/cl.h>
39#endif
40
41#include <stdio.h>
42#include <device/device_config.h>
43#include <device/opencl/jit.h>
44#include <device/opencl/prgm_lib.h>
45#include <device/opencl/check.h>
46
47#include "euler_res_kernel.cl.h"
48
49void euler_res_part_visc_opencl(void *rhs_u, void *Binv, void *lap_sol,
50 void *h, real *c_avisc, int *n) {
51 cl_int err;
52
53 if (euler_res_program == NULL)
54 opencl_kernel_jit(euler_res_kernel, (cl_program *) &euler_res_program);
55
56 cl_kernel kernel = clCreateKernel(euler_res_program,
57 "euler_res_part_visc_kernel", &err);
58 CL_CHECK(err);
59
60 CL_CHECK(clSetKernelArg(kernel, 0, sizeof(cl_mem), (void *) &rhs_u));
61 CL_CHECK(clSetKernelArg(kernel, 1, sizeof(cl_mem), (void *) &Binv));
62 CL_CHECK(clSetKernelArg(kernel, 2, sizeof(cl_mem), (void *) &lap_sol));
63 CL_CHECK(clSetKernelArg(kernel, 3, sizeof(cl_mem), (void *) &h));
64 CL_CHECK(clSetKernelArg(kernel, 4, sizeof(real), c_avisc));
65 CL_CHECK(clSetKernelArg(kernel, 5, sizeof(int), n));
66
67 const int nb = ((*n) + 256 - 1) / 256;
68 const size_t global_item_size = 256 * nb;
69 const size_t local_item_size = 256;
70
71 CL_CHECK(clEnqueueNDRangeKernel((cl_command_queue) glb_cmd_queue, kernel, 1,
72 NULL, &global_item_size, &local_item_size,
73 0, NULL, NULL));
74 CL_CHECK(clReleaseKernel(kernel));
75}
76
77void euler_res_part_mx_flux_opencl(void *f_x, void *f_y, void *f_z,
78 void *m_x, void *m_y, void *m_z,
79 void *rho_field, void *p, int *n) {
80 cl_int err;
81
82 if (euler_res_program == NULL)
83 opencl_kernel_jit(euler_res_kernel, (cl_program *) &euler_res_program);
84
85 cl_kernel kernel = clCreateKernel(euler_res_program,
86 "euler_res_part_mx_flux_kernel", &err);
87 CL_CHECK(err);
88
89 CL_CHECK(clSetKernelArg(kernel, 0, sizeof(cl_mem), (void *) &f_x));
90 CL_CHECK(clSetKernelArg(kernel, 1, sizeof(cl_mem), (void *) &f_y));
91 CL_CHECK(clSetKernelArg(kernel, 2, sizeof(cl_mem), (void *) &f_z));
92 CL_CHECK(clSetKernelArg(kernel, 3, sizeof(cl_mem), (void *) &m_x));
93 CL_CHECK(clSetKernelArg(kernel, 4, sizeof(cl_mem), (void *) &m_y));
94 CL_CHECK(clSetKernelArg(kernel, 5, sizeof(cl_mem), (void *) &m_z));
95 CL_CHECK(clSetKernelArg(kernel, 6, sizeof(cl_mem), (void *) &rho_field));
96 CL_CHECK(clSetKernelArg(kernel, 7, sizeof(cl_mem), (void *) &p));
97 CL_CHECK(clSetKernelArg(kernel, 8, sizeof(int), n));
98
99 const int nb = ((*n) + 256 - 1) / 256;
100 const size_t global_item_size = 256 * nb;
101 const size_t local_item_size = 256;
102
103 CL_CHECK(clEnqueueNDRangeKernel((cl_command_queue) glb_cmd_queue, kernel, 1,
104 NULL, &global_item_size, &local_item_size,
105 0, NULL, NULL));
106 CL_CHECK(clReleaseKernel(kernel));
107}
108
109void euler_res_part_my_flux_opencl(void *f_x, void *f_y, void *f_z,
110 void *m_x, void *m_y, void *m_z,
111 void *rho_field, void *p, int *n) {
112 cl_int err;
113
114 if (euler_res_program == NULL)
115 opencl_kernel_jit(euler_res_kernel, (cl_program *) &euler_res_program);
116
117 cl_kernel kernel = clCreateKernel(euler_res_program,
118 "euler_res_part_my_flux_kernel", &err);
119 CL_CHECK(err);
120
121 CL_CHECK(clSetKernelArg(kernel, 0, sizeof(cl_mem), (void *) &f_x));
122 CL_CHECK(clSetKernelArg(kernel, 1, sizeof(cl_mem), (void *) &f_y));
123 CL_CHECK(clSetKernelArg(kernel, 2, sizeof(cl_mem), (void *) &f_z));
124 CL_CHECK(clSetKernelArg(kernel, 3, sizeof(cl_mem), (void *) &m_x));
125 CL_CHECK(clSetKernelArg(kernel, 4, sizeof(cl_mem), (void *) &m_y));
126 CL_CHECK(clSetKernelArg(kernel, 5, sizeof(cl_mem), (void *) &m_z));
127 CL_CHECK(clSetKernelArg(kernel, 6, sizeof(cl_mem), (void *) &rho_field));
128 CL_CHECK(clSetKernelArg(kernel, 7, sizeof(cl_mem), (void *) &p));
129 CL_CHECK(clSetKernelArg(kernel, 8, sizeof(int), n));
130
131 const int nb = ((*n) + 256 - 1) / 256;
132 const size_t global_item_size = 256 * nb;
133 const size_t local_item_size = 256;
134
135 CL_CHECK(clEnqueueNDRangeKernel((cl_command_queue) glb_cmd_queue, kernel, 1,
136 NULL, &global_item_size, &local_item_size,
137 0, NULL, NULL));
138 CL_CHECK(clReleaseKernel(kernel));
139}
140
141void euler_res_part_mz_flux_opencl(void *f_x, void *f_y, void *f_z,
142 void *m_x, void *m_y, void *m_z,
143 void *rho_field, void *p, int *n) {
144 cl_int err;
145
146 if (euler_res_program == NULL)
147 opencl_kernel_jit(euler_res_kernel, (cl_program *) &euler_res_program);
148
149 cl_kernel kernel = clCreateKernel(euler_res_program,
150 "euler_res_part_mz_flux_kernel", &err);
151 CL_CHECK(err);
152
153 CL_CHECK(clSetKernelArg(kernel, 0, sizeof(cl_mem), (void *) &f_x));
154 CL_CHECK(clSetKernelArg(kernel, 1, sizeof(cl_mem), (void *) &f_y));
155 CL_CHECK(clSetKernelArg(kernel, 2, sizeof(cl_mem), (void *) &f_z));
156 CL_CHECK(clSetKernelArg(kernel, 3, sizeof(cl_mem), (void *) &m_x));
157 CL_CHECK(clSetKernelArg(kernel, 4, sizeof(cl_mem), (void *) &m_y));
158 CL_CHECK(clSetKernelArg(kernel, 5, sizeof(cl_mem), (void *) &m_z));
159 CL_CHECK(clSetKernelArg(kernel, 6, sizeof(cl_mem), (void *) &rho_field));
160 CL_CHECK(clSetKernelArg(kernel, 7, sizeof(cl_mem), (void *) &p));
161 CL_CHECK(clSetKernelArg(kernel, 8, sizeof(int), n));
162
163 const int nb = ((*n) + 256 - 1) / 256;
164 const size_t global_item_size = 256 * nb;
165 const size_t local_item_size = 256;
166
167 CL_CHECK(clEnqueueNDRangeKernel((cl_command_queue) glb_cmd_queue, kernel, 1,
168 NULL, &global_item_size, &local_item_size,
169 0, NULL, NULL));
170 CL_CHECK(clReleaseKernel(kernel));
171}
172
173void euler_res_part_E_flux_opencl(void *f_x, void *f_y, void *f_z,
174 void *m_x, void *m_y, void *m_z,
175 void *rho_field, void *p, void * E, int *n) {
176 cl_int err;
177
178 if (euler_res_program == NULL)
179 opencl_kernel_jit(euler_res_kernel, (cl_program *) &euler_res_program);
180
181 cl_kernel kernel = clCreateKernel(euler_res_program,
182 "euler_res_part_E_flux_kernel", &err);
183 CL_CHECK(err);
184
185 CL_CHECK(clSetKernelArg(kernel, 0, sizeof(cl_mem), (void *) &f_x));
186 CL_CHECK(clSetKernelArg(kernel, 1, sizeof(cl_mem), (void *) &f_y));
187 CL_CHECK(clSetKernelArg(kernel, 2, sizeof(cl_mem), (void *) &f_z));
188 CL_CHECK(clSetKernelArg(kernel, 3, sizeof(cl_mem), (void *) &m_x));
189 CL_CHECK(clSetKernelArg(kernel, 4, sizeof(cl_mem), (void *) &m_y));
190 CL_CHECK(clSetKernelArg(kernel, 5, sizeof(cl_mem), (void *) &m_z));
191 CL_CHECK(clSetKernelArg(kernel, 6, sizeof(cl_mem), (void *) &rho_field));
192 CL_CHECK(clSetKernelArg(kernel, 7, sizeof(cl_mem), (void *) &E));
193 CL_CHECK(clSetKernelArg(kernel, 8, sizeof(cl_mem), (void *) &p));
194 CL_CHECK(clSetKernelArg(kernel, 9, sizeof(int), n));
195
196 const int nb = ((*n) + 256 - 1) / 256;
197 const size_t global_item_size = 256 * nb;
198 const size_t local_item_size = 256;
199
200 CL_CHECK(clEnqueueNDRangeKernel((cl_command_queue) glb_cmd_queue, kernel, 1,
201 NULL, &global_item_size, &local_item_size,
202 0, NULL, NULL));
203 CL_CHECK(clReleaseKernel(kernel));
204}
205
206void euler_res_part_coef_mult_opencl(void *rhs_rho, void *rhs_m_x,
207 void *rhs_m_y, void *rhs_m_z,
208 void *rhs_E, void *mult, int *n) {
209 cl_int err;
210
211 if (euler_res_program == NULL)
212 opencl_kernel_jit(euler_res_kernel, (cl_program *) &euler_res_program);
213
214 cl_kernel kernel = clCreateKernel(euler_res_program,
215 "euler_res_part_coef_mult_kernel", &err);
216 CL_CHECK(err);
217
218 CL_CHECK(clSetKernelArg(kernel, 0, sizeof(cl_mem), (void *) &rhs_rho));
219 CL_CHECK(clSetKernelArg(kernel, 1, sizeof(cl_mem), (void *) &rhs_m_x));
220 CL_CHECK(clSetKernelArg(kernel, 2, sizeof(cl_mem), (void *) &rhs_m_y));
221 CL_CHECK(clSetKernelArg(kernel, 3, sizeof(cl_mem), (void *) &rhs_m_z));
222 CL_CHECK(clSetKernelArg(kernel, 4, sizeof(cl_mem), (void *) &rhs_E));
223 CL_CHECK(clSetKernelArg(kernel, 5, sizeof(cl_mem), (void *) &mult));
224 CL_CHECK(clSetKernelArg(kernel, 6, sizeof(int), n));
225
226 const int nb = ((*n) + 256 - 1) / 256;
227 const size_t global_item_size = 256 * nb;
228 const size_t local_item_size = 256;
229
230 CL_CHECK(clEnqueueNDRangeKernel((cl_command_queue) glb_cmd_queue, kernel, 1,
231 NULL, &global_item_size, &local_item_size,
232 0, NULL, NULL));
233 CL_CHECK(clReleaseKernel(kernel));
234}
235
236void euler_res_part_rk_sum_opencl(void *rho, void *m_x, void *m_y, void *m_z,
237 void *E, void *k_rho_i, void *k_m_x_i,
238 void *k_m_y_i, void *k_m_z_i, void *k_E_i,
239 real *dt, real *c, int *n) {
240 cl_int err;
241
242 if (euler_res_program == NULL)
243 opencl_kernel_jit(euler_res_kernel, (cl_program *) &euler_res_program);
244
245 cl_kernel kernel = clCreateKernel(euler_res_program,
246 "euler_res_part_rk_sum_kernel", &err);
247 CL_CHECK(err);
248
249 CL_CHECK(clSetKernelArg(kernel, 0, sizeof(cl_mem), (void *) &rho));
250 CL_CHECK(clSetKernelArg(kernel, 1, sizeof(cl_mem), (void *) &m_x));
251 CL_CHECK(clSetKernelArg(kernel, 2, sizeof(cl_mem), (void *) &m_y));
252 CL_CHECK(clSetKernelArg(kernel, 3, sizeof(cl_mem), (void *) &m_z));
253 CL_CHECK(clSetKernelArg(kernel, 4, sizeof(cl_mem), (void *) &E));
254 CL_CHECK(clSetKernelArg(kernel, 5, sizeof(cl_mem), (void *) &k_rho_i));
255 CL_CHECK(clSetKernelArg(kernel, 6, sizeof(cl_mem), (void *) &k_m_x_i));
256 CL_CHECK(clSetKernelArg(kernel, 7, sizeof(cl_mem), (void *) &k_m_y_i));
257 CL_CHECK(clSetKernelArg(kernel, 8, sizeof(cl_mem), (void *) &k_m_z_i));
258 CL_CHECK(clSetKernelArg(kernel, 9, sizeof(cl_mem), (void *) &k_E_i));
259 CL_CHECK(clSetKernelArg(kernel, 10, sizeof(real), dt));
260 CL_CHECK(clSetKernelArg(kernel, 11, sizeof(real), c));
261 CL_CHECK(clSetKernelArg(kernel, 12, sizeof(int), n));
262
263 const int nb = ((*n) + 256 - 1) / 256;
264 const size_t global_item_size = 256 * nb;
265 const size_t local_item_size = 256;
266
267 CL_CHECK(clEnqueueNDRangeKernel((cl_command_queue) glb_cmd_queue, kernel, 1,
268 NULL, &global_item_size, &local_item_size,
269 0, NULL, NULL));
270 CL_CHECK(clReleaseKernel(kernel));
271}
dt
__global__ void const T *__restrict__ const T *__restrict__ const T *__restrict__ const T *__restrict__ dt
Definition cdtp_kernel.h:109
dirichlet_apply_scalar_kernel
__global__ void dirichlet_apply_scalar_kernel(const int *__restrict__ msk, T *__restrict__ x, const T g, const int m)
Definition dirichlet_kernel.h:42
device_config.h
real
double real
Definition device_config.h:12
euler_res_part_coef_mult_opencl
void euler_res_part_coef_mult_opencl(void *rhs_rho, void *rhs_m_x, void *rhs_m_y, void *rhs_m_z, void *rhs_E, void *mult, int *n)
Definition euler_res.c:206
euler_res_part_E_flux_opencl
void euler_res_part_E_flux_opencl(void *f_x, void *f_y, void *f_z, void *m_x, void *m_y, void *m_z, void *rho_field, void *p, void *E, int *n)
Definition euler_res.c:173
euler_res_part_my_flux_opencl
void euler_res_part_my_flux_opencl(void *f_x, void *f_y, void *f_z, void *m_x, void *m_y, void *m_z, void *rho_field, void *p, int *n)
Definition euler_res.c:109
euler_res_part_mx_flux_opencl
void euler_res_part_mx_flux_opencl(void *f_x, void *f_y, void *f_z, void *m_x, void *m_y, void *m_z, void *rho_field, void *p, int *n)
Definition euler_res.c:77
euler_res_part_mz_flux_opencl
void euler_res_part_mz_flux_opencl(void *f_x, void *f_y, void *f_z, void *m_x, void *m_y, void *m_z, void *rho_field, void *p, int *n)
Definition euler_res.c:141
euler_res_part_rk_sum_opencl
void euler_res_part_rk_sum_opencl(void *rho, void *m_x, void *m_y, void *m_z, void *E, void *k_rho_i, void *k_m_x_i, void *k_m_y_i, void *k_m_z_i, void *k_E_i, real *dt, real *c, int *n)
Definition euler_res.c:236
euler_res_part_visc_opencl
void euler_res_part_visc_opencl(void *rhs_u, void *Binv, void *lap_sol, void *h, real *c_avisc, int *n)
Definition euler_res.c:49
opencl_kernel_jit
void opencl_kernel_jit(const char *kernel, cl_program *program)
Definition jit.c:50
CL_CHECK
#define CL_CHECK(err)
Definition check.h:12
prgm_lib.h
euler_res_program
void * euler_res_program