27 template <int NUM_COMP, int COMP_STRIDE, int NUM_PTS>
32   for (CeedInt comp = 0; comp < NUM_COMP; comp++) {  in ReadPoint()
40 template <int NUM_COMP, int COMP_STRIDE, int NUM_PTS>
46     for (CeedInt comp = 0; comp < NUM_COMP; comp++) {  in WritePoint()
59 template <int NUM_COMP, int P_1D>
72 template <int NUM_COMP, int COMP_STRIDE, int P_1D>
79     for (CeedInt comp = 0; comp < NUM_COMP; comp++) r_u[comp] = d_u[ind + COMP_STRIDE * comp];  in ReadLVecStandard1d()
86 template <int NUM_COMP, int P_1D, int STRIDES_NODE, int STRIDES_COMP, int STRIDES_ELEM>
92     for (CeedInt comp = 0; comp < NUM_COMP; comp++) r_u[comp] = d_u[ind + comp * STRIDES_COMP];  in ReadLVecStrided1d()
99 template <int NUM_COMP, int COMP_STRIDE, int P_1D>
106 …for (CeedInt comp = 0; comp < NUM_COMP; comp++) atomicAdd(&d_v[ind + COMP_STRIDE * comp], r_v[comp…  in WriteLVecStandard1d()
110 template <int NUM_COMP, int COMP_STRIDE, int P_1D>
127 template <int NUM_COMP, int COMP_STRIDE, int P_1D>
132   const CeedInt e_vec_size = P_1D * NUM_COMP;  in WriteLVecStandard1d_Assembly()
137     for (CeedInt comp = 0; comp < NUM_COMP; comp++) {  in WriteLVecStandard1d_Assembly()
138 …d_v[elem * e_vec_size * e_vec_size + (in_comp * NUM_COMP + comp) * P_1D * P_1D + out_node * P_1D +…  in WriteLVecStandard1d_Assembly()
161 template <int NUM_COMP, int P_1D, int STRIDES_NODE, int STRIDES_COMP, int STRIDES_ELEM>
168     for (CeedInt comp = 0; comp < NUM_COMP; comp++) d_v[ind + comp * STRIDES_COMP] += r_v[comp];  in WriteLVecStrided1d()
179 template <int NUM_COMP, int P_1D>
193 template <int NUM_COMP, int COMP_STRIDE, int P_1D>
200     for (CeedInt comp = 0; comp < NUM_COMP; comp++) r_u[comp] = d_u[ind + COMP_STRIDE * comp];  in ReadLVecStandard2d()
207 template <int NUM_COMP, int P_1D, int STRIDES_NODE, int STRIDES_COMP, int STRIDES_ELEM>
213     for (CeedInt comp = 0; comp < NUM_COMP; comp++) r_u[comp] = d_u[ind + comp * STRIDES_COMP];  in ReadLVecStrided2d()
220 template <int NUM_COMP, int COMP_STRIDE, int P_1D>
227 …for (CeedInt comp = 0; comp < NUM_COMP; comp++) atomicAdd(&d_v[ind + COMP_STRIDE * comp], r_v[comp…  in WriteLVecStandard2d()
231 template <int NUM_COMP, int COMP_STRIDE, int P_1D>
250 template <int NUM_COMP, int COMP_STRIDE, int P_1D>
257   const CeedInt e_vec_size = elem_size * NUM_COMP;  in WriteLVecStandard2d_Assembly()
263     for (CeedInt comp = 0; comp < NUM_COMP; comp++) {  in WriteLVecStandard2d_Assembly()
264 …const CeedInt index = (in_comp * NUM_COMP + comp) * elem_size * elem_size + out_node * elem_size +…  in WriteLVecStandard2d_Assembly()
289 template <int NUM_COMP, int P_1D, int STRIDES_NODE, int STRIDES_COMP, int STRIDES_ELEM>
296     for (CeedInt comp = 0; comp < NUM_COMP; comp++) d_v[ind + comp * STRIDES_COMP] += r_v[comp];  in WriteLVecStrided2d()
307 template <int NUM_COMP, int P_1D>
322 template <int NUM_COMP, int COMP_STRIDE, int P_1D>
330 …for (CeedInt comp = 0; comp < NUM_COMP; comp++) r_u[z + comp * P_1D] = d_u[ind + COMP_STRIDE * com…  in ReadLVecStandard3d()
338 template <int NUM_COMP, int P_1D, int STRIDES_NODE, int STRIDES_COMP, int STRIDES_ELEM>
345 …for (CeedInt comp = 0; comp < NUM_COMP; comp++) r_u[z + comp * P_1D] = d_u[ind + comp * STRIDES_CO…  in ReadLVecStrided3d()
353 template <int NUM_COMP, int COMP_STRIDE, int Q_1D>
361     for (CeedInt comp = 0; comp < NUM_COMP; comp++) r_u[comp] = d_u[ind + COMP_STRIDE * comp];  in ReadEVecSliceStandard3d()
368 template <int NUM_COMP, int Q_1D, int STRIDES_NODE, int STRIDES_COMP, int STRIDES_ELEM>
375     for (CeedInt comp = 0; comp < NUM_COMP; comp++) r_u[comp] = d_u[ind + comp * STRIDES_COMP];  in ReadEVecSliceStrided3d()
382 template <int NUM_COMP, int COMP_STRIDE, int P_1D>
390 …for (CeedInt comp = 0; comp < NUM_COMP; comp++) atomicAdd(&d_v[ind + COMP_STRIDE * comp], r_v[z + …  in WriteLVecStandard3d()
395 template <int NUM_COMP, int COMP_STRIDE, int P_1D>
415 template <int NUM_COMP, int COMP_STRIDE, int P_1D>
423   const CeedInt e_vec_size = elem_size * NUM_COMP;  in WriteLVecStandard3d_Assembly()
430       for (CeedInt comp = 0; comp < NUM_COMP; comp++) {  in WriteLVecStandard3d_Assembly()
431 …const CeedInt index = (in_comp * NUM_COMP + comp) * elem_size * elem_size + out_node * elem_size +…  in WriteLVecStandard3d_Assembly()
459 template <int NUM_COMP, int P_1D, int STRIDES_NODE, int STRIDES_COMP, int STRIDES_ELEM>
467 …for (CeedInt comp = 0; comp < NUM_COMP; comp++) d_v[ind + comp * STRIDES_COMP] += r_v[z + comp * P…  in WriteLVecStrided3d()
475 template <int NUM_COMP, int Q_1D, int T_1D>
479     for (CeedInt comp = 0; comp < NUM_COMP; comp++) {  in GradColloSlice3d()
484       r_V[comp + 0 * NUM_COMP] = 0.0;  in GradColloSlice3d()
486 …      r_V[comp + 0 * NUM_COMP] += c_G[i + data.t_id_x * Q_1D] * data.slice[i + data.t_id_y * T_1D];  in GradColloSlice3d()
489       r_V[comp + 1 * NUM_COMP] = 0.0;  in GradColloSlice3d()
491 …      r_V[comp + 1 * NUM_COMP] += c_G[i + data.t_id_y * Q_1D] * data.slice[data.t_id_x + i * T_1D];  in GradColloSlice3d()
494       r_V[comp + 2 * NUM_COMP] = 0.0;  in GradColloSlice3d()
496         r_V[comp + 2 * NUM_COMP] += c_G[i + q * Q_1D] * r_U[i + comp * Q_1D];  in GradColloSlice3d()
505 template <int NUM_COMP, int Q_1D, int T_1D>
509     for (CeedInt comp = 0; comp < NUM_COMP; comp++) {  in GradColloSliceTranspose3d()
512       data.slice[data.t_id_x + data.t_id_y * T_1D] = r_U[comp + 0 * NUM_COMP];  in GradColloSliceTranspose3d()
519       data.slice[data.t_id_x + data.t_id_y * T_1D] = r_U[comp + 1 * NUM_COMP];  in GradColloSliceTranspose3d()
526         r_V[i + comp * Q_1D] += c_G[i + q * Q_1D] * r_U[comp + 2 * NUM_COMP];  in GradColloSliceTranspose3d()