Neko  0.9.0
A portable framework for high-order spectral element flow simulations
schwarz.c
Go to the documentation of this file.
1 /*
2  Copyright (c) 2022, 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 "schwarz_kernel.cl.h"
48 
49 void opencl_schwarz_extrude(void *arr1, int * l1, real * f1,
50  void *arr2, int * l2, real * f2,
51  int * nx, int * nel) {
52  cl_int err;
53 
54  if (schwarz_program == NULL)
55  opencl_kernel_jit(schwarz_kernel, (cl_program *) &schwarz_program);
56 
57  const size_t global_item_size = 256 * (*nel);
58  const size_t local_item_size = 256;
59 
60 #define STR(X) #X
61 #define CASE(NX) \
62  case NX: \
63  { \
64  \
65  cl_kernel kernel = clCreateKernel(schwarz_program, \
66  STR(schwarz_extrude_kernel_nx##NX), \
67  &err); \
68  CL_CHECK(err); \
69  \
70  CL_CHECK(clSetKernelArg(kernel, 0, sizeof(cl_mem), (void *) &arr1)); \
71  CL_CHECK(clSetKernelArg(kernel, 1, sizeof(int), l1)); \
72  CL_CHECK(clSetKernelArg(kernel, 2, sizeof(real), f1)); \
73  CL_CHECK(clSetKernelArg(kernel, 3, sizeof(cl_mem), (void *) &arr2)); \
74  CL_CHECK(clSetKernelArg(kernel, 4, sizeof(int), l2)); \
75  CL_CHECK(clSetKernelArg(kernel, 5, sizeof(real), f2)); \
76  \
77  CL_CHECK(clEnqueueNDRangeKernel((cl_command_queue) glb_cmd_queue, \
78  kernel, 1, NULL, &global_item_size, \
79  &local_item_size, 0, NULL, NULL)); \
80  } \
81  break
82 
83  switch(*nx) {
84  CASE(2);
85  CASE(3);
86  CASE(4);
87  CASE(5);
88  CASE(6);
89  CASE(7);
90  CASE(8);
91  CASE(9);
92  CASE(10);
93  CASE(11);
94  CASE(12);
95  CASE(13);
96  CASE(14);
97  CASE(15);
98  CASE(16);
99  }
100 }
101 
102 void opencl_schwarz_toext3d(void *a, void *b,int * nx, int * nel) {
103  cl_int err;
104 
105  if (schwarz_program == NULL)
106  opencl_kernel_jit(schwarz_kernel, (cl_program *) &schwarz_program);
107 
108  cl_kernel kernel = clCreateKernel(schwarz_program,
109  "schwarz_toext3d_kernel", &err);
110  CL_CHECK(err);
111 
112  CL_CHECK(clSetKernelArg(kernel, 0, sizeof(cl_mem), (void *) &a));
113  CL_CHECK(clSetKernelArg(kernel, 1, sizeof(cl_mem), (void *) &b));
114  CL_CHECK(clSetKernelArg(kernel, 2, sizeof(int), nx));
115 
116  const size_t global_item_size = 256 * (*nel);
117  const size_t local_item_size = 256;
118 
119  CL_CHECK(clEnqueueNDRangeKernel((cl_command_queue) glb_cmd_queue, kernel, 1,
120  NULL, &global_item_size, &local_item_size,
121  0, NULL, NULL));
122 }
123 
124 void opencl_schwarz_toreg3d(void *b, void *a,int * nx, int * nel) {
125  cl_int err;
126 
127  if (schwarz_program == NULL)
128  opencl_kernel_jit(schwarz_kernel, (cl_program *) &schwarz_program);
129 
130  cl_kernel kernel = clCreateKernel(schwarz_program,
131  "schwarz_toreg3d_kernel", &err);
132  CL_CHECK(err);
133 
134  CL_CHECK(clSetKernelArg(kernel, 0, sizeof(cl_mem), (void *) &b));
135  CL_CHECK(clSetKernelArg(kernel, 1, sizeof(cl_mem), (void *) &a));
136  CL_CHECK(clSetKernelArg(kernel, 2, sizeof(int), nx));
137 
138  const size_t global_item_size = 256 * (*nel);
139  const size_t local_item_size = 256;
140 
141  CL_CHECK(clEnqueueNDRangeKernel((cl_command_queue) glb_cmd_queue, kernel, 1,
142  NULL, &global_item_size, &local_item_size,
143  0, NULL, NULL));
144 }
145 
146 
double real
Definition: device_config.h:12
void * glb_cmd_queue
void opencl_kernel_jit(const char *kernel, cl_program *program)
Definition: jit.c:50
#define CL_CHECK(err)
Definition: check.h:12
void * schwarz_program
void opencl_schwarz_toreg3d(void *b, void *a, int *nx, int *nel)
Definition: schwarz.c:124
void opencl_schwarz_toext3d(void *a, void *b, int *nx, int *nel)
Definition: schwarz.c:102
#define CASE(NX)
void opencl_schwarz_extrude(void *arr1, int *l1, real *f1, void *arr2, int *l2, real *f2, int *nx, int *nel)
Definition: schwarz.c:49