Neko 0.9.99
A portable framework for high-order spectral element flow simulations
Loading...
Searching...
No Matches
fdm_kernel.h
Go to the documentation of this file.
1#ifndef __MATH_FDM_KERNEL_H__
2#define __MATH_FDM_KERNEL_H__
3/*
4 Copyright (c) 2021-2022, The Neko Authors
5 All rights reserved.
6
7 Redistribution and use in source and binary forms, with or without
8 modification, are permitted provided that the following conditions
9 are met:
10
11 * Redistributions of source code must retain the above copyright
12 notice, this list of conditions and the following disclaimer.
13
14 * Redistributions in binary form must reproduce the above
15 copyright notice, this list of conditions and the following
16 disclaimer in the documentation and/or other materials provided
17 with the distribution.
18
19 * Neither the name of the authors nor the names of its
20 contributors may be used to endorse or promote products derived
21 from this software without specific prior written permission.
22
23 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
24 "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
25 LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
26 FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
27 COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
28 INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
29 BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
30 LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
31 CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
32 LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
33 ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
34 POSSIBILITY OF SUCH DAMAGE.
35*/
36
37template< typename T, const int NL >
38__global__ void fdm_do_fast_kernel(T * __restrict__ e,
39 T * __restrict__ r,
40 T * __restrict__ s,
41 T * __restrict__ d) {
42 __shared__ T shwork[NL*NL*NL];
43 __shared__ T shwork2[NL*NL*NL];
44 __shared__ T A[NL*NL];
45 __shared__ T Bt[NL*NL];
46 __shared__ T Ct[NL*NL];
47
48 const int idx = threadIdx.x;
49 const int str = blockDim.x;
50 const int el = blockIdx.x;
51 if( idx < NL*NL){
52 A[idx] = s[idx+NL*NL+el*NL*NL*3*2];
53 Bt[idx] = s[idx+2*NL*NL+el*NL*NL*3*2];
54 Ct[idx] = s[idx+2*2*NL*NL+el*NL*NL*3*2];
55 }
56 __syncthreads();
57
58 for (int ii = idx; ii< NL*NL*NL; ii += str){
59 T tmp = 0.0;
60 int j = ii/NL;
61 int i = ii - j*NL;
62 for( int l = 0; l < NL; l++){
63 tmp += A[i+l*NL]*r[l+NL*j+el*NL*NL*NL];
64 }
65 shwork[ii] = tmp;
66 }
67 __syncthreads();
68 for (int ijk = idx; ijk< NL*NL*NL; ijk += str){
69 const int jk = ijk / NL;
70 const int i = ijk - jk * NL;
71 const int k = jk / NL;
72 const int j = jk - k * NL;
73 T tmp = 0.0;
74 const int ik2 = i + k*NL*NL;
75 for( int l = 0; l < NL; l++){
76 tmp += Bt[l+j*NL]*shwork[l*NL+ik2];
77 }
78 shwork2[ijk] = tmp;
79 }
80 __syncthreads();
81 for (int ijk = idx; ijk< NL*NL*NL; ijk += str){
82 const int jk = ijk / NL;
83 const int i = ijk - jk * NL;
84 const int k = jk / NL;
85 const int j = jk - k * NL;
86 T tmp = 0.0;
87 const int ij2 = i + j*NL;
88 for( int l = 0; l < NL; l++){
89 tmp += Ct[l+k*NL]*shwork2[ij2 + l*NL*NL];
90 }
91 r[ijk+el*NL*NL*NL] = tmp*d[ijk+el*NL*NL*NL];
92 }
93 __syncthreads();
94 if( idx < NL*NL){
95 A[idx] = s[idx+el*NL*NL*3*2];
96 Bt[idx] = s[idx+NL*NL+2*NL*NL+el*NL*NL*3*2];
97 Ct[idx] = s[idx+NL*NL+2*2*NL*NL+el*NL*NL*3*2];
98 }
99 __syncthreads();
100
101 for (int ii = idx; ii< NL*NL*NL; ii += str){
102 T tmp = 0.0;
103 int j = ii/NL;
104 int i = ii - j*NL;
105 for( int l = 0; l < NL; l++){
106 tmp += A[i+l*NL]*r[l+NL*j+el*NL*NL*NL];
107 }
108 shwork[ii] = tmp;
109 }
110 __syncthreads();
111 for (int ijk = idx; ijk< NL*NL*NL; ijk += str){
112 const int jk = ijk / NL;
113 const int i = ijk - jk * NL;
114 const int k = jk / NL;
115 const int j = jk - k * NL;
116 T tmp = 0.0;
117 const int ik2 = i + k*NL*NL;
118 for( int l = 0; l < NL; l++){
119 tmp += Bt[l+j*NL]*shwork[l*NL+ik2];
120 }
121 shwork2[ijk] = tmp;
122 }
123 __syncthreads();
124 for (int ijk = idx; ijk< NL*NL*NL; ijk += str){
125 const int jk = ijk / NL;
126 const int i = ijk - jk * NL;
127 const int k = jk / NL;
128 const int j = jk - k * NL;
129 T tmp = 0.0;
130 const int ij2 = i + j*NL;
131 for( int l = 0; l < NL; l++){
132 tmp += Ct[l+k*NL]*shwork2[ij2 + l*NL*NL];
133 }
134 e[ijk+el*NL*NL*NL] = tmp;
135 }
136
137}
138
139
140
141#endif // __MATH_FDM_KERNEL_H__
i
const int i
Definition ax_helm_full_kernel.h:95
e
const int e
Definition ax_helm_full_kernel.h:93
j
const int j
Definition ax_helm_full_kernel.h:94
__syncthreads
__syncthreads()
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
fdm_do_fast_kernel
__global__ void fdm_do_fast_kernel(T *__restrict__ e, T *__restrict__ r, T *__restrict__ s, T *__restrict__ d)
Definition fdm_kernel.h:38