1 // Copyright (c) 2017-2026, Lawrence Livermore National Security, LLC and other CEED contributors.
2 // All Rights Reserved. See the top-level LICENSE and NOTICE files for details.
3 //
4 // SPDX-License-Identifier: BSD-2-Clause
5 //
6 // This file is part of CEED: http://github.com/ceed
7 #pragma once
8
9 #include <ceed.h>
10 #include "ex-common.h"
11
12 /// libCEED Q-function for building quadrature data for a mass + diffusion operator
build_mass_diff(void * ctx,const CeedInt Q,const CeedScalar * const * in,CeedScalar * const * out)13 CEED_QFUNCTION(build_mass_diff)(void *ctx, const CeedInt Q, const CeedScalar *const *in, CeedScalar *const *out) {
14 // in[0] is Jacobians with shape [dim, dim, Q]
15 // in[1] is quadrature weights, size (Q)
16 const CeedScalar *w = in[1];
17 CeedScalar(*q_data)[CEED_Q_VLA] = (CeedScalar(*)[CEED_Q_VLA])out[0];
18 struct BuildContext *build_data = (struct BuildContext *)ctx;
19
20 // At every quadrature point, compute w/det(J).adj(J).adj(J)^T and store
21 // the symmetric part of the result.
22 switch (build_data->dim + 10 * build_data->space_dim) {
23 case 11: { // dim = 1, space_dim = 1
24 const CeedScalar(*J)[1][CEED_Q_VLA] = (const CeedScalar(*)[1][CEED_Q_VLA])in[0];
25
26 CeedPragmaSIMD for (CeedInt i = 0; i < Q; i++) {
27 // Mass
28 q_data[0][i] = w[i] * J[0][0][i];
29
30 // Diffusion
31 q_data[1][i] = w[i] / J[0][0][i];
32 }
33 } break;
34 case 22: { // dim = 2, space_dim = 2
35 const CeedScalar(*J)[2][CEED_Q_VLA] = (const CeedScalar(*)[2][CEED_Q_VLA])in[0];
36
37 CeedPragmaSIMD for (CeedInt i = 0; i < Q; i++) {
38 // J: 0 2 q_data: 0 2 adj(J): J22 -J12
39 // 1 3 2 1 -J10 J00
40 const CeedScalar J00 = J[0][0][i];
41 const CeedScalar J10 = J[0][1][i];
42 const CeedScalar J01 = J[1][0][i];
43 const CeedScalar J11 = J[1][1][i];
44 const CeedScalar qw = w[i] / (J00 * J11 - J10 * J01);
45
46 // Mass
47 q_data[0][i] = w[i] * (J00 * J11 - J10 * J01);
48
49 // Diffusion
50 q_data[1][i] = qw * (J01 * J01 + J11 * J11);
51 q_data[2][i] = qw * (J00 * J00 + J10 * J10);
52 q_data[3][i] = -qw * (J00 * J01 + J10 * J11);
53 }
54 } break;
55 case 33: { // dim = 3, space_dim = 3
56 const CeedScalar(*J)[3][CEED_Q_VLA] = (const CeedScalar(*)[3][CEED_Q_VLA])in[0];
57
58 CeedPragmaSIMD for (CeedInt i = 0; i < Q; i++) {
59 // Compute the adjoint
60 CeedScalar A[3][3];
61 for (CeedInt j = 0; j < 3; j++) {
62 for (CeedInt k = 0; k < 3; k++) {
63 A[k][j] =
64 J[(k + 1) % 3][(j + 1) % 3][i] * J[(k + 2) % 3][(j + 2) % 3][i] - J[(k + 2) % 3][(j + 1) % 3][i] * J[(k + 1) % 3][(j + 2) % 3][i];
65 }
66 }
67
68 // Compute quadrature weight / det(J)
69 const CeedScalar qw = w[i] / (J[0][0][i] * A[0][0] + J[0][1][i] * A[0][1] + J[0][2][i] * A[0][2]);
70
71 // Mass
72 q_data[0][i] = w[i] * (J[0][0][i] * A[0][0] + J[0][1][i] * A[0][1] + J[0][2][i] * A[0][2]);
73
74 // Diffusion
75 // Stored in Voigt convention
76 // 1 6 5
77 // 6 2 4
78 // 5 4 3
79 q_data[1][i] = qw * (A[0][0] * A[0][0] + A[0][1] * A[0][1] + A[0][2] * A[0][2]);
80 q_data[2][i] = qw * (A[1][0] * A[1][0] + A[1][1] * A[1][1] + A[1][2] * A[1][2]);
81 q_data[3][i] = qw * (A[2][0] * A[2][0] + A[2][1] * A[2][1] + A[2][2] * A[2][2]);
82 q_data[4][i] = qw * (A[1][0] * A[2][0] + A[1][1] * A[2][1] + A[1][2] * A[2][2]);
83 q_data[5][i] = qw * (A[0][0] * A[2][0] + A[0][1] * A[2][1] + A[0][2] * A[2][2]);
84 q_data[6][i] = qw * (A[0][0] * A[1][0] + A[0][1] * A[1][1] + A[0][2] * A[1][2]);
85 }
86 } break;
87 }
88 return CEED_ERROR_SUCCESS;
89 }
90
91 /// libCEED Q-function for applying a mass + diffusion operator
apply_mass_diff(void * ctx,const CeedInt Q,const CeedScalar * const * in,CeedScalar * const * out)92 CEED_QFUNCTION(apply_mass_diff)(void *ctx, const CeedInt Q, const CeedScalar *const *in, CeedScalar *const *out) {
93 struct BuildContext *build_data = (struct BuildContext *)ctx;
94 // in[0], out[0] solution values with shape [1, 1, Q]
95 // in[1], out[1] solution gradients with shape [dim, 1, Q]
96 // in[2] is quadrature data with shape [num_components, Q]
97 const CeedScalar(*q_data)[CEED_Q_VLA] = (const CeedScalar(*)[CEED_Q_VLA])in[2];
98
99 switch (build_data->dim) {
100 case 1: {
101 const CeedScalar *u = in[0], *ug = in[1];
102 CeedScalar *v = out[0], *vg = out[1];
103
104 CeedPragmaSIMD for (CeedInt i = 0; i < Q; i++) {
105 // Mass
106 v[i] = q_data[0][i] * u[i];
107
108 // Diffusion
109 vg[i] = q_data[1][i] * ug[i];
110 }
111 } break;
112 case 2: {
113 const CeedScalar *u = in[0];
114 const CeedScalar(*ug)[CEED_Q_VLA] = (const CeedScalar(*)[CEED_Q_VLA])in[1];
115 CeedScalar *v = out[0];
116 CeedScalar(*vg)[CEED_Q_VLA] = (CeedScalar(*)[CEED_Q_VLA])out[1];
117
118 CeedPragmaSIMD for (CeedInt i = 0; i < Q; i++) {
119 // Mass
120 v[i] = q_data[0][i] * u[i];
121
122 // Diffusion
123 // Read q_data (dXdxdXdx_T symmetric matrix)
124 // Stored in Voigt convention
125 // 1 3
126 // 3 2
127 const CeedScalar dXdxdXdx_T[2][2] = {
128 {q_data[1][i], q_data[3][i]},
129 {q_data[3][i], q_data[2][i]}
130 };
131
132 // j = direction of vg
133 for (int j = 0; j < 2; j++) {
134 vg[j][i] = (ug[0][i] * dXdxdXdx_T[0][j] + ug[1][i] * dXdxdXdx_T[1][j]);
135 }
136 }
137 } break;
138 case 3: {
139 const CeedScalar *u = in[0];
140 const CeedScalar(*ug)[CEED_Q_VLA] = (const CeedScalar(*)[CEED_Q_VLA])in[1];
141 CeedScalar *v = out[0];
142 CeedScalar(*vg)[CEED_Q_VLA] = (CeedScalar(*)[CEED_Q_VLA])out[1];
143
144 CeedPragmaSIMD for (CeedInt i = 0; i < Q; i++) {
145 // Mass
146 v[i] = q_data[0][i] * u[i];
147
148 // Diffusion
149 // Read q_data (dXdxdXdx_T symmetric matrix)
150 // Stored in Voigt convention
151 // 1 6 5
152 // 6 2 4
153 // 5 4 3
154 const CeedScalar dXdxdXdx_T[3][3] = {
155 {q_data[1][i], q_data[6][i], q_data[5][i]},
156 {q_data[6][i], q_data[2][i], q_data[4][i]},
157 {q_data[5][i], q_data[4][i], q_data[3][i]}
158 };
159
160 // j = direction of vg
161 for (int j = 0; j < 3; j++) {
162 vg[j][i] = (ug[0][i] * dXdxdXdx_T[0][j] + ug[1][i] * dXdxdXdx_T[1][j] + ug[2][i] * dXdxdXdx_T[2][j]);
163 }
164 }
165 } break;
166 }
167 return CEED_ERROR_SUCCESS;
168 }
169