19 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
26   if (t_id_x < Q_1D && t_id_y < P_1D) {  in ContractX2dFlattened()
27     for (CeedInt i = 0; i < P_1D; i++) {  in ContractX2dFlattened()
28       *V += B[i + t_id_x * P_1D] * data.slice[i + t_id_y * T_1D];  // Contract x direction  in ContractX2dFlattened()
36 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
44     for (CeedInt i = 0; i < P_1D; i++) {  in ContractY2dFlattened()
45       *V += B[i + t_id_y * P_1D] * data.slice[t_id_x + i * T_1D];  // Contract y direction  in ContractY2dFlattened()
53 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
60   if (t_id_x < Q_1D && t_id_y < P_1D) {  in ContractTransposeY2dFlattened()
62       *V += B[t_id_y + i * P_1D] * data.slice[t_id_x + i * T_1D];  // Contract y direction  in ContractTransposeY2dFlattened()
70 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
77   if (t_id_x < P_1D && t_id_y < P_1D) {  in ContractTransposeX2dFlattened()
79       *V += B[t_id_x + i * P_1D] * data.slice[i + t_id_y * T_1D];  // Contract x direction  in ContractTransposeX2dFlattened()
87 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
93   if (t_id_x < P_1D && t_id_y < P_1D) {  in ContractTransposeAddX2dFlattened()
95       *V += B[t_id_x + i * P_1D] * data.slice[i + t_id_y * T_1D];  // Contract x direction  in ContractTransposeAddX2dFlattened()
130 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
136   if (P_1D != T_1D) QUnpack2d<NUM_COMP, P_1D, T_1D>(data, t_id_x, t_id_y, r_U);  in InterpTensor2dFlattened()
138     ContractX2dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, &r_U[comp], c_B, r_t);  in InterpTensor2dFlattened()
139     ContractY2dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, r_t, c_B, &r_V[comp]);  in InterpTensor2dFlattened()
142   if (P_1D != T_1D) QPack2d<NUM_COMP, P_1D, T_1D>(data, t_id_x, t_id_y, r_U);  in InterpTensor2dFlattened()
149 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
157 …ContractTransposeY2dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, &r_U[comp], c_B, r…  in InterpTransposeTensor2dFlattened()
158 …ContractTransposeX2dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, r_t, c_B, &r_V[com…  in InterpTransposeTensor2dFlattened()
161   if (P_1D != T_1D) QPack2d<NUM_COMP, P_1D, T_1D>(data, t_id_x, t_id_y, r_V);  in InterpTransposeTensor2dFlattened()
167 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
172   if (P_1D != T_1D) QUnpack2d<NUM_COMP, P_1D, T_1D>(data, t_id_x, t_id_y, r_U);  in InterpTensorCollocatedNodes2dFlattened()
177   if (P_1D != T_1D) QPack2d<NUM_COMP, P_1D, T_1D>(data, t_id_x, t_id_y, r_U);  in InterpTensorCollocatedNodes2dFlattened()
184 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
194   if (P_1D != T_1D) QPack2d<NUM_COMP, P_1D, T_1D>(data, t_id_x, t_id_y, r_V);  in InterpTransposeTensorCollocatedNodes2dFlattened()
200 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
206   if (P_1D != T_1D) QUnpack2d<NUM_COMP, P_1D, T_1D>(data, t_id_x, t_id_y, r_U);  in GradTensor2dFlattened()
208     ContractX2dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, &r_U[comp], c_G, r_t);  in GradTensor2dFlattened()
209 …ContractY2dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, r_t, c_B, &r_V[comp + 0 * N…  in GradTensor2dFlattened()
210     ContractX2dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, &r_U[comp], c_B, r_t);  in GradTensor2dFlattened()
211 …ContractY2dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, r_t, c_G, &r_V[comp + 1 * N…  in GradTensor2dFlattened()
214   if (P_1D != T_1D) QPack2d<NUM_COMP, P_1D, T_1D>(data, t_id_x, t_id_y, r_U);  in GradTensor2dFlattened()
221 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
229 …ContractTransposeY2dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, &r_U[comp + 0 * NU…  in GradTransposeTensor2dFlattened()
230 …ContractTransposeX2dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, r_t, c_G, &r_V[com…  in GradTransposeTensor2dFlattened()
231 …ContractTransposeY2dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, &r_U[comp + 1 * NU…  in GradTransposeTensor2dFlattened()
232 …ContractTransposeAddX2dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, r_t, c_B, &r_V[…  in GradTransposeTensor2dFlattened()
235   if (P_1D != T_1D) QPack2d<NUM_COMP, P_1D, T_1D>(data, t_id_x, t_id_y, r_V);  in GradTransposeTensor2dFlattened()
241 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
246   if (P_1D != T_1D) QUnpack2d<NUM_COMP, P_1D, T_1D>(data, t_id_x, t_id_y, r_U);  in GradTensorCollocatedNodes2dFlattened()
248 …ContractX2dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, &r_U[comp], c_G, &r_V[comp …  in GradTensorCollocatedNodes2dFlattened()
249 …ContractY2dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, &r_U[comp], c_G, &r_V[comp …  in GradTensorCollocatedNodes2dFlattened()
252   if (P_1D != T_1D) QPack2d<NUM_COMP, P_1D, T_1D>(data, t_id_x, t_id_y, r_U);  in GradTensorCollocatedNodes2dFlattened()
259 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
266 …ContractTransposeY2dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, &r_U[comp + 1 * NU…  in GradTransposeTensorCollocatedNodes2dFlattened()
267 …ContractTransposeAddX2dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, &r_U[comp + 0 *…  in GradTransposeTensorCollocatedNodes2dFlattened()
270   if (P_1D != T_1D) QPack2d<NUM_COMP, P_1D, T_1D>(data, t_id_x, t_id_y, r_V);  in GradTransposeTensorCollocatedNodes2dFlattened()
276 template <int P_1D, int Q_1D>
290 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
297   if (t_id_x < Q_1D && t_id_y < P_1D && t_id_z < P_1D) {  in ContractX3dFlattened()
298     for (CeedInt i = 0; i < P_1D; i++) {  in ContractX3dFlattened()
299 …*V += B[i + t_id_x * P_1D] * data.slice[i + t_id_y * T_1D + t_id_z * T_1D * T_1D];  // Contract x …  in ContractX3dFlattened()
307 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
314   if (t_id_x < Q_1D && t_id_y < Q_1D && t_id_z < P_1D) {  in ContractY3dFlattened()
315     for (CeedInt i = 0; i < P_1D; i++) {  in ContractY3dFlattened()
316 …*V += B[i + t_id_y * P_1D] * data.slice[t_id_x + i * T_1D + t_id_z * T_1D * T_1D];  // Contract y …  in ContractY3dFlattened()
324 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
332     for (CeedInt i = 0; i < P_1D; i++) {  in ContractZ3dFlattened()
333 …*V += B[i + t_id_z * P_1D] * data.slice[t_id_x + t_id_y * T_1D + i * T_1D * T_1D];  // Contract z …  in ContractZ3dFlattened()
341 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
348   if (t_id_x < Q_1D && t_id_y < Q_1D && t_id_z < P_1D) {  in ContractTransposeZ3dFlattened()
350 …*V += B[t_id_z + i * P_1D] * data.slice[t_id_x + t_id_y * T_1D + i * T_1D * T_1D];  // Contract z …  in ContractTransposeZ3dFlattened()
358 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
364   if (t_id_x < Q_1D && t_id_y < Q_1D && t_id_z < P_1D) {  in ContractTransposeAddZ3dFlattened()
366 …*V += B[t_id_z + i * P_1D] * data.slice[t_id_x + t_id_y * T_1D + i * T_1D * T_1D];  // Contract z …  in ContractTransposeAddZ3dFlattened()
374 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
381   if (t_id_x < Q_1D && t_id_y < P_1D && t_id_z < P_1D) {  in ContractTransposeY3dFlattened()
383 …*V += B[t_id_y + i * P_1D] * data.slice[t_id_x + i * T_1D + t_id_z * T_1D * T_1D];  // Contract y …  in ContractTransposeY3dFlattened()
391 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
397   if (t_id_x < Q_1D && t_id_y < P_1D && t_id_z < P_1D) {  in ContractTransposeAddY3dFlattened()
399 …*V += B[t_id_y + i * P_1D] * data.slice[t_id_x + i * T_1D + t_id_z * T_1D * T_1D];  // Contract y …  in ContractTransposeAddY3dFlattened()
407 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
414   if (t_id_x < P_1D && t_id_y < P_1D && t_id_z < P_1D) {  in ContractTransposeX3dFlattened()
416 …*V += B[t_id_x + i * P_1D] * data.slice[i + t_id_y * T_1D + t_id_z * T_1D * T_1D];  // Contract x …  in ContractTransposeX3dFlattened()
424 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
430   if (t_id_x < P_1D && t_id_y < P_1D && t_id_z < P_1D) {  in ContractTransposeAddX3dFlattened()
432 …*V += B[t_id_x + i * P_1D] * data.slice[i + t_id_y * T_1D + t_id_z * T_1D * T_1D];  // Contract x …  in ContractTransposeAddX3dFlattened()
467 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
473   if (P_1D != T_1D) QUnpack3d<NUM_COMP, P_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, r_U);  in InterpTensor3dFlattened()
475 …ContractX3dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, &r_U[comp], c_B, r_…  in InterpTensor3dFlattened()
476     ContractY3dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, r_t1, c_B, r_t2);  in InterpTensor3dFlattened()
477 …ContractZ3dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, r_t2, c_B, &r_V[com…  in InterpTensor3dFlattened()
480   if (P_1D != T_1D) QPack3d<NUM_COMP, P_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, r_U);  in InterpTensor3dFlattened()
487 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
495 …ContractTransposeZ3dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, &r_U[comp]…  in InterpTransposeTensor3dFlattened()
496 …ContractTransposeY3dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, r_t1, c_B,…  in InterpTransposeTensor3dFlattened()
497 …ContractTransposeX3dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, r_t2, c_B,…  in InterpTransposeTensor3dFlattened()
500   if (P_1D != T_1D) QPack3d<NUM_COMP, P_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, r_V);  in InterpTransposeTensor3dFlattened()
506 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
511   if (P_1D != T_1D) QUnpack3d<NUM_COMP, P_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, r_U);  in InterpTensorCollocatedNodes3dFlattened()
516   if (P_1D != T_1D) QPack3d<NUM_COMP, P_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, r_U);  in InterpTensorCollocatedNodes3dFlattened()
523 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
533   if (P_1D != T_1D) QPack3d<NUM_COMP, P_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, r_V);  in InterpTransposeTensorCollocatedNodes3dFlattened()
539 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
545   if (P_1D != T_1D) QUnpack3d<NUM_COMP, P_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, r_U);  in GradTensor3dFlattened()
547 …ContractX3dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, &r_U[comp], c_G, r_…  in GradTensor3dFlattened()
548     ContractY3dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, r_t1, c_B, r_t2);  in GradTensor3dFlattened()
549 …ContractZ3dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, r_t2, c_B, &r_V[com…  in GradTensor3dFlattened()
550 …ContractX3dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, &r_U[comp], c_B, r_…  in GradTensor3dFlattened()
551     ContractY3dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, r_t1, c_G, r_t2);  in GradTensor3dFlattened()
552 …ContractZ3dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, r_t2, c_B, &r_V[com…  in GradTensor3dFlattened()
553 …ContractX3dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, &r_U[comp], c_B, r_…  in GradTensor3dFlattened()
554     ContractY3dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, r_t1, c_B, r_t2);  in GradTensor3dFlattened()
555 …ContractZ3dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, r_t2, c_G, &r_V[com…  in GradTensor3dFlattened()
558   if (P_1D != T_1D) QPack3d<NUM_COMP, P_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, r_U);  in GradTensor3dFlattened()
565 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
573 …ContractTransposeZ3dFlattened<NUM_COMP, t_id_x, t_id_y, t_id_z, P_1D, Q_1D, T_1D>(data, &r_U[comp …  in GradTransposeTensor3dFlattened()
574 …ContractTransposeY3dFlattened<NUM_COMP, t_id_x, t_id_y, t_id_z, P_1D, Q_1D, T_1D>(data, r_t1, c_B,…  in GradTransposeTensor3dFlattened()
575 …ContractTransposeX3dFlattened<NUM_COMP, t_id_x, t_id_y, t_id_z, P_1D, Q_1D, T_1D>(data, r_t2, c_G,…  in GradTransposeTensor3dFlattened()
576 …ContractTransposeZ3dFlattened<NUM_COMP, t_id_x, t_id_y, t_id_z, P_1D, Q_1D, T_1D>(data, &r_U[comp …  in GradTransposeTensor3dFlattened()
577 …ContractTransposeY3dFlattened<NUM_COMP, t_id_x, t_id_y, t_id_z, P_1D, Q_1D, T_1D>(data, r_t1, c_G,…  in GradTransposeTensor3dFlattened()
578 …ContractTransposeAddX3dFlattened<NUM_COMP, t_id_x, t_id_y, t_id_z, P_1D, Q_1D, T_1D>(data, r_t2, c…  in GradTransposeTensor3dFlattened()
579 …ContractTransposeZ3dFlattened<NUM_COMP, t_id_x, t_id_y, t_id_z, P_1D, Q_1D, T_1D>(data, &r_U[comp …  in GradTransposeTensor3dFlattened()
580 …ContractTransposeY3dFlattened<NUM_COMP, t_id_x, t_id_y, t_id_z, P_1D, Q_1D, T_1D>(data, r_t1, c_B,…  in GradTransposeTensor3dFlattened()
581 …ContractTransposeAddX3dFlattened<NUM_COMP, t_id_x, t_id_y, t_id_z, P_1D, Q_1D, T_1D>(data, r_t2, c…  in GradTransposeTensor3dFlattened()
584   if (P_1D != T_1D) QPack3d<NUM_COMP, P_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, r_V);  in GradTransposeTensor3dFlattened()
590 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
596   if (P_1D != T_1D) QUnpack3d<NUM_COMP, P_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, r_U);  in GradTensorCollocated3dFlattened()
598 …ContractX3dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, &r_U[comp], c_B, r_…  in GradTensorCollocated3dFlattened()
599     ContractY3dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, r_t1, c_B, r_t2);  in GradTensorCollocated3dFlattened()
600     ContractZ3dFlattened<NUM_COMP, P_1D, Q_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, r_t2, c_B, r_t1);  in GradTensorCollocated3dFlattened()
606   if (P_1D != T_1D) QPack3d<NUM_COMP, P_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, r_U);  in GradTensorCollocated3dFlattened()
613 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
621 …ContractTransposeZ3dFlattened<NUM_COMP, t_id_x, t_id_y, t_id_z, P_1D, Q_1D, T_1D>(data, &r_U[comp …  in GradTransposeTensor3dFlattened()
622 …ContractTransposeY3dFlattened<NUM_COMP, t_id_x, t_id_y, t_id_z, P_1D, Q_1D, T_1D>(data, r_t1, c_B,…  in GradTransposeTensor3dFlattened()
623 …ContractTransposeX3dFlattened<NUM_COMP, t_id_x, t_id_y, t_id_z, P_1D, Q_1D, T_1D>(data, r_t2, c_G,…  in GradTransposeTensor3dFlattened()
624 …ContractTransposeZ3dFlattened<NUM_COMP, t_id_x, t_id_y, t_id_z, P_1D, Q_1D, T_1D>(data, &r_U[comp …  in GradTransposeTensor3dFlattened()
625 …ContractTransposeY3dFlattened<NUM_COMP, t_id_x, t_id_y, t_id_z, P_1D, Q_1D, T_1D>(data, r_t1, c_G,…  in GradTransposeTensor3dFlattened()
626 …ContractTransposeAddX3dFlattened<NUM_COMP, t_id_x, t_id_y, t_id_z, P_1D, Q_1D, T_1D>(data, r_t2, c…  in GradTransposeTensor3dFlattened()
627 …ContractTransposeZ3dFlattened<NUM_COMP, t_id_x, t_id_y, t_id_z, P_1D, Q_1D, T_1D>(data, &r_U[comp …  in GradTransposeTensor3dFlattened()
628 …ContractTransposeY3dFlattened<NUM_COMP, t_id_x, t_id_y, t_id_z, P_1D, Q_1D, T_1D>(data, r_t1, c_B,…  in GradTransposeTensor3dFlattened()
629 …ContractTransposeAddX3dFlattened<NUM_COMP, t_id_x, t_id_y, t_id_z, P_1D, Q_1D, T_1D>(data, r_t2, c…  in GradTransposeTensor3dFlattened()
632   if (P_1D != T_1D) QPack3d<NUM_COMP, P_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, r_V);  in GradTransposeTensor3dFlattened()
638 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
643   if (P_1D != T_1D) QUnpack3d<NUM_COMP, P_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, r_U);  in GradTensorCollocatedNodes3dFlattened()
650   if (P_1D != T_1D) QPack3d<NUM_COMP, P_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, r_U);  in GradTensorCollocatedNodes3dFlattened()
657 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
669   if (P_1D != T_1D) QPack3d<NUM_COMP, P_1D, T_1D>(data, t_id_x, t_id_y, t_id_z, r_V);  in GradTransposeTensorCollocatedNodes3dFlattened()
675 template <int P_1D, int Q_1D>