1 // Copyright (c) 2017-2018, Lawrence Livermore National Security, LLC. 2 // Produced at the Lawrence Livermore National Laboratory. LLNL-CODE-734707. 3 // All Rights reserved. See files LICENSE and NOTICE for details. 4 // 5 // This file is part of CEED, a collection of benchmarks, miniapps, software 6 // libraries and APIs for efficient high-order finite element and spectral 7 // element discretizations for exascale applications. For more information and 8 // source code availability see http://github.com/ceed. 9 // 10 // The CEED research is supported by the Exascale Computing Project 17-SC-20-SC, 11 // a collaborative effort of two U.S. Department of Energy organizations (Office 12 // of Science and the National Nuclear Security Administration) responsible for 13 // the planning and preparation of a capable exascale ecosystem, including 14 // software, applications, hardware, advanced system engineering and early 15 // testbed platforms, in support of the nation's exascale computing imperative. 16 17 #define CEED_DEBUG_COLOR 12 18 19 #include <ceed/ceed.h> 20 #include <ceed/backend.h> 21 #include <cuda_runtime.h> 22 #include <iostream> 23 #include <sstream> 24 #include "ceed-cuda-gen.h" 25 #include "../cuda/ceed-cuda-compile.h" 26 #include "../cuda-ref/ceed-cuda-ref.h" 27 #include "../cuda-shared/ceed-cuda-shared.h" 28 29 static const char *atomicAdd = QUOTE( 30 //------------------------------------------------------------------------------ 31 // Atomic add, for older CUDA 32 //------------------------------------------------------------------------------ 33 __device__ CeedScalar atomicAdd(CeedScalar *address, CeedScalar val) { 34 unsigned long long int *address_as_ull = (unsigned long long int *)address; 35 unsigned long long int old = *address_as_ull, assumed; 36 do { 37 assumed = old; 38 old = 39 atomicCAS(address_as_ull, assumed, 40 __double_as_longlong(val + 41 __longlong_as_double(assumed))); 42 // Note: uses integer comparison to avoid hang in case of NaN 43 // (since NaN != NaN) 44 } while (assumed != old); 45 return __longlong_as_double(old); 46 } 47 ); 48 49 static const char *deviceFunctions = QUOTE( 50 51 //------------------------------------------------------------------------------ 52 // Typedefs 53 //------------------------------------------------------------------------------ 54 typedef struct { const CeedScalar* in[16]; CeedScalar* out[16]; } CudaFields; 55 typedef struct { CeedInt* in[16]; CeedInt* out[16]; } CudaFieldsInt; 56 57 typedef struct { 58 CeedInt tidx; 59 CeedInt tidy; 60 CeedInt tidz; 61 CeedInt tid; 62 CeedScalar* slice; 63 } BackendData; 64 65 //------------------------------------------------------------------------------ 66 // Load matrices for basis actions 67 //------------------------------------------------------------------------------ 68 template <int P, int Q> 69 inline __device__ void loadMatrix(BackendData &data, const CeedScalar *__restrict__ d_B, CeedScalar *B) { 70 for (CeedInt i = data.tid; i < P*Q; i += blockDim.x*blockDim.y*blockDim.z) 71 B[i] = d_B[i]; 72 } 73 74 //------------------------------------------------------------------------------ 75 // 1D 76 //------------------------------------------------------------------------------ 77 78 //------------------------------------------------------------------------------ 79 // L-vector -> E-vector, offsets provided 80 //------------------------------------------------------------------------------ 81 template <int NCOMP, int COMPSTRIDE, int P1d> 82 inline __device__ void readDofsOffset1d(BackendData &data, const CeedInt nnodes, const CeedInt elem, const CeedInt *__restrict__ indices, const CeedScalar *__restrict__ d_u, CeedScalar *r_u) { 83 if (data.tidx < P1d) { 84 const CeedInt node = data.tidx; 85 const CeedInt ind = indices[node + elem * P1d]; 86 for (CeedInt comp = 0; comp < NCOMP; ++comp) 87 r_u[comp] = d_u[ind + COMPSTRIDE * comp]; 88 } 89 } 90 91 //------------------------------------------------------------------------------ 92 // L-vector -> E-vector, strided 93 //------------------------------------------------------------------------------ 94 template <int NCOMP, int P1d, int STRIDES_NODE, int STRIDES_COMP, int STRIDES_ELEM> 95 inline __device__ void readDofsStrided1d(BackendData &data, const CeedInt elem, const CeedScalar *__restrict__ d_u, CeedScalar *r_u) { 96 if (data.tidx < P1d) { 97 const CeedInt node = data.tidx; 98 const CeedInt ind = node * STRIDES_NODE + elem * STRIDES_ELEM; 99 for (CeedInt comp = 0; comp < NCOMP; ++comp) 100 r_u[comp] = d_u[ind + comp * STRIDES_COMP]; 101 } 102 } 103 104 //------------------------------------------------------------------------------ 105 // E-vector -> L-vector, offsets provided 106 //------------------------------------------------------------------------------ 107 template <int NCOMP, int COMPSTRIDE, int P1d> 108 inline __device__ void writeDofsOffset1d(BackendData &data, const CeedInt nnodes, const CeedInt elem, const CeedInt *__restrict__ indices, const CeedScalar *r_v, CeedScalar *d_v) { 109 if (data.tidx < P1d) { 110 const CeedInt node = data.tidx; 111 const CeedInt ind = indices[node + elem * P1d]; 112 for (CeedInt comp = 0; comp < NCOMP; ++comp) 113 atomicAdd(&d_v[ind + COMPSTRIDE * comp], r_v[comp]); 114 } 115 } 116 117 //------------------------------------------------------------------------------ 118 // E-vector -> L-vector, strided 119 //------------------------------------------------------------------------------ 120 template <int NCOMP, int P1d, int STRIDES_NODE, int STRIDES_COMP, int STRIDES_ELEM> 121 inline __device__ void writeDofsStrided1d(BackendData &data, const CeedInt elem, const CeedScalar *r_v, CeedScalar *d_v) { 122 if (data.tidx < P1d) { 123 const CeedInt node = data.tidx; 124 const CeedInt ind = node * STRIDES_NODE + elem * STRIDES_ELEM; 125 for (CeedInt comp = 0; comp < NCOMP; ++comp) 126 d_v[ind + comp * STRIDES_COMP] += r_v[comp]; 127 } 128 } 129 130 //------------------------------------------------------------------------------ 131 // 1D tensor contraction x 132 //------------------------------------------------------------------------------ 133 template <int NCOMP, int P1d, int Q1d> 134 inline __device__ void ContractX1d(BackendData &data, const CeedScalar *U, const CeedScalar *B, CeedScalar *V) { 135 data.slice[data.tidx] = *U; 136 __syncthreads(); 137 *V = 0.0; 138 if (data.tidx < Q1d) 139 for (CeedInt i = 0; i < P1d; ++i) 140 *V += B[i + data.tidx*P1d] * data.slice[i]; // Contract x direction 141 __syncthreads(); 142 } 143 144 //------------------------------------------------------------------------------ 145 // 1D transpose tensor contraction x 146 //------------------------------------------------------------------------------ 147 template <int NCOMP, int P1d, int Q1d> 148 inline __device__ void ContractTransposeX1d(BackendData &data, const CeedScalar *U, const CeedScalar *B, CeedScalar *V) { 149 data.slice[data.tidx] = *U; 150 __syncthreads(); 151 *V = 0.0; 152 if (data.tidx < P1d) 153 for (CeedInt i = 0; i < Q1d; ++i) 154 *V += B[data.tidx + i*P1d] * data.slice[i]; // Contract x direction 155 __syncthreads(); 156 } 157 158 //------------------------------------------------------------------------------ 159 // 1D interpolate to quadrature points 160 //------------------------------------------------------------------------------ 161 template <int NCOMP, int P1d, int Q1d> 162 inline __device__ void interp1d(BackendData &data, const CeedScalar *__restrict__ r_U, const CeedScalar *c_B, CeedScalar *__restrict__ r_V) { 163 for (CeedInt comp = 0; comp < NCOMP; comp++) 164 ContractX1d<NCOMP, P1d, Q1d>(data, r_U + comp, c_B, r_V + comp); 165 } 166 167 //------------------------------------------------------------------------------ 168 // 1D interpolate transpose 169 //------------------------------------------------------------------------------ 170 template <int NCOMP, int P1d, int Q1d> 171 inline __device__ void interpTranspose1d(BackendData &data, const CeedScalar *__restrict__ r_U, const CeedScalar *c_B, CeedScalar *__restrict__ r_V) { 172 for (CeedInt comp=0; comp<NCOMP; comp++) 173 ContractTransposeX1d<NCOMP, P1d, Q1d>(data, r_U + comp, c_B, r_V + comp); 174 } 175 176 //------------------------------------------------------------------------------ 177 // 1D derivatives at quadrature points 178 //------------------------------------------------------------------------------ 179 template <int NCOMP, int P1d, int Q1d> 180 inline __device__ void grad1d(BackendData &data, const CeedScalar *__restrict__ r_U, const CeedScalar *c_B, const CeedScalar *c_G, CeedScalar *__restrict__ r_V) { 181 for (CeedInt comp = 0; comp < NCOMP; comp++) 182 ContractX1d<NCOMP, P1d, Q1d>(data, r_U + comp, c_G, r_V + comp); 183 } 184 185 //------------------------------------------------------------------------------ 186 // 1D derivatives transpose 187 //------------------------------------------------------------------------------ 188 template <int NCOMP, int P1d, int Q1d> 189 inline __device__ void gradTranspose1d(BackendData &data, const CeedScalar *__restrict__ r_U, const CeedScalar *c_B, const CeedScalar *c_G, CeedScalar *__restrict__ r_V) { 190 for (CeedInt comp = 0; comp < NCOMP; comp++) 191 ContractTransposeX1d<NCOMP, P1d, Q1d>(data, r_U + comp, c_G, r_V + comp); 192 } 193 194 //------------------------------------------------------------------------------ 195 // 2D 196 //------------------------------------------------------------------------------ 197 198 //------------------------------------------------------------------------------ 199 // L-vector -> E-vector, offsets provided 200 //------------------------------------------------------------------------------ 201 template <int NCOMP, int COMPSTRIDE, int P1d> 202 inline __device__ void readDofsOffset2d(BackendData &data, const CeedInt nnodes, const CeedInt elem, const CeedInt *__restrict__ indices, const CeedScalar *__restrict__ d_u, CeedScalar *r_u) { 203 if (data.tidx < P1d && data.tidy < P1d) { 204 const CeedInt node = data.tidx + data.tidy*P1d; 205 const CeedInt ind = indices[node + elem * P1d*P1d]; 206 for (CeedInt comp = 0; comp < NCOMP; ++comp) 207 r_u[comp] = d_u[ind + COMPSTRIDE * comp]; 208 } 209 } 210 211 //------------------------------------------------------------------------------ 212 // L-vector -> E-vector, strided 213 //------------------------------------------------------------------------------ 214 template <int NCOMP, int P1d, int STRIDES_NODE, int STRIDES_COMP, int STRIDES_ELEM> 215 inline __device__ void readDofsStrided2d(BackendData &data, const CeedInt elem, const CeedScalar *__restrict__ d_u, CeedScalar *r_u) { 216 if (data.tidx < P1d && data.tidy < P1d) { 217 const CeedInt node = data.tidx + data.tidy*P1d; 218 const CeedInt ind = node * STRIDES_NODE + elem * STRIDES_ELEM; 219 for (CeedInt comp = 0; comp < NCOMP; ++comp) 220 r_u[comp] = d_u[ind + comp * STRIDES_COMP]; 221 } 222 } 223 224 //------------------------------------------------------------------------------ 225 // E-vector -> L-vector, offsets provided 226 //------------------------------------------------------------------------------ 227 template <int NCOMP, int COMPSTRIDE, int P1d> 228 inline __device__ void writeDofsOffset2d(BackendData &data, const CeedInt nnodes, const CeedInt elem, const CeedInt *__restrict__ indices, const CeedScalar *r_v, CeedScalar *d_v) { 229 if (data.tidx < P1d && data.tidy < P1d) { 230 const CeedInt node = data.tidx + data.tidy*P1d; 231 const CeedInt ind = indices[node + elem * P1d*P1d]; 232 for (CeedInt comp = 0; comp < NCOMP; ++comp) 233 atomicAdd(&d_v[ind + COMPSTRIDE * comp], r_v[comp]); 234 } 235 } 236 237 //------------------------------------------------------------------------------ 238 // E-vector -> L-vector, strided 239 //------------------------------------------------------------------------------ 240 template <int NCOMP, int P1d, int STRIDES_NODE, int STRIDES_COMP, int STRIDES_ELEM> 241 inline __device__ void writeDofsStrided2d(BackendData &data, const CeedInt elem, const CeedScalar *r_v, CeedScalar *d_v) { 242 if (data.tidx < P1d && data.tidy < P1d) { 243 const CeedInt node = data.tidx + data.tidy*P1d; 244 const CeedInt ind = node * STRIDES_NODE + elem * STRIDES_ELEM; 245 for (CeedInt comp = 0; comp < NCOMP; ++comp) 246 d_v[ind + comp * STRIDES_COMP] += r_v[comp]; 247 } 248 } 249 250 //------------------------------------------------------------------------------ 251 // 2D tensor contraction x 252 //------------------------------------------------------------------------------ 253 template <int NCOMP, int P1d, int Q1d> 254 inline __device__ void ContractX2d(BackendData &data, const CeedScalar *U, const CeedScalar *B, CeedScalar *V) { 255 data.slice[data.tidx+data.tidy*T1d] = *U; 256 __syncthreads(); 257 *V = 0.0; 258 if (data.tidx < Q1d && data.tidy < P1d) 259 for (CeedInt i = 0; i < P1d; ++i) 260 *V += B[i + data.tidx*P1d] * data.slice[i + data.tidy*T1d]; // Contract x direction 261 __syncthreads(); 262 } 263 264 //------------------------------------------------------------------------------ 265 // 2D tensor contract y 266 //------------------------------------------------------------------------------ 267 template <int NCOMP, int P1d, int Q1d> 268 inline __device__ void ContractY2d(BackendData &data, const CeedScalar *U, const CeedScalar *B, CeedScalar *V) { 269 data.slice[data.tidx+data.tidy*T1d] = *U; 270 __syncthreads(); 271 *V = 0.0; 272 if (data.tidx < Q1d && data.tidy < Q1d) 273 for (CeedInt i = 0; i < P1d; ++i) 274 *V += B[i + data.tidy*P1d] * data.slice[data.tidx + i*T1d]; // Contract y direction 275 __syncthreads(); 276 } 277 278 //------------------------------------------------------------------------------ 279 // 2D transpose tensor contract y 280 //------------------------------------------------------------------------------ 281 template <int NCOMP, int P1d, int Q1d> 282 inline __device__ void ContractYTranspose2d(BackendData &data, const CeedScalar *U, const CeedScalar *B, CeedScalar *V) { 283 data.slice[data.tidx+data.tidy*T1d] = *U; 284 __syncthreads(); 285 *V = 0.0; 286 if (data.tidx < Q1d && data.tidy < P1d) 287 for (CeedInt i = 0; i < Q1d; ++i) 288 *V += B[data.tidy + i*P1d] * data.slice[data.tidx + i*T1d]; // Contract y direction 289 __syncthreads(); 290 } 291 292 //------------------------------------------------------------------------------ 293 // 2D transpose tensor contract x 294 //------------------------------------------------------------------------------ 295 template <int NCOMP, int P1d, int Q1d> 296 inline __device__ void ContractXTranspose2d(BackendData &data, const CeedScalar *U, const CeedScalar *B, CeedScalar *V) { 297 data.slice[data.tidx+data.tidy*T1d] = *U; 298 __syncthreads(); 299 *V = 0.0; 300 if (data.tidx < P1d && data.tidy < P1d) 301 for (CeedInt i = 0; i < Q1d; ++i) 302 *V += B[data.tidx + i*P1d] * data.slice[i + data.tidy*T1d]; // Contract x direction 303 __syncthreads(); 304 } 305 306 //------------------------------------------------------------------------------ 307 // 2D transpose tensor contract and add x 308 //------------------------------------------------------------------------------ 309 template <int NCOMP, int P1d, int Q1d> 310 inline __device__ void ContractXTransposeAdd2d(BackendData &data, const CeedScalar *U, const CeedScalar *B, CeedScalar *V) { 311 data.slice[data.tidx+data.tidy*T1d] = *U; 312 __syncthreads(); 313 if (data.tidx < P1d && data.tidy < P1d) 314 for (CeedInt i = 0; i < Q1d; ++i) 315 *V += B[data.tidx + i*P1d] * data.slice[i + data.tidy*T1d]; // Contract x direction 316 __syncthreads(); 317 } 318 319 //------------------------------------------------------------------------------ 320 // 2D interpolate to quadrature points 321 //------------------------------------------------------------------------------ 322 template <int NCOMP, int P1d, int Q1d> 323 inline __device__ void interp2d(BackendData &data, const CeedScalar *__restrict__ r_U, const CeedScalar *c_B, CeedScalar *__restrict__ r_V) { 324 CeedScalar r_t[1]; 325 for (CeedInt comp = 0; comp < NCOMP; comp++) { 326 ContractX2d<NCOMP, P1d, Q1d>(data, r_U + comp, c_B, r_t); 327 ContractY2d<NCOMP, P1d, Q1d>(data, r_t, c_B, r_V + comp); 328 } 329 } 330 331 //------------------------------------------------------------------------------ 332 // 2D interpolate transpose 333 //------------------------------------------------------------------------------ 334 template <int NCOMP, int P1d, int Q1d> 335 inline __device__ void interpTranspose2d(BackendData &data, const CeedScalar *__restrict__ r_U, const CeedScalar *c_B, CeedScalar *__restrict__ r_V) { 336 CeedScalar r_t[1]; 337 for (CeedInt comp = 0; comp < NCOMP; comp++) { 338 ContractYTranspose2d<NCOMP, P1d, Q1d>(data, r_U + comp, c_B, r_t); 339 ContractXTranspose2d<NCOMP, P1d, Q1d>(data, r_t, c_B, r_V + comp); 340 } 341 } 342 343 //------------------------------------------------------------------------------ 344 // 2D derivatives at quadrature points 345 //------------------------------------------------------------------------------ 346 template <int NCOMP, int P1d, int Q1d> 347 inline __device__ void grad2d(BackendData &data, const CeedScalar *__restrict__ r_U, const CeedScalar *c_B, const CeedScalar *c_G, CeedScalar *__restrict__ r_V) { 348 CeedScalar r_t[1]; 349 for (CeedInt comp = 0; comp < NCOMP; comp++) { 350 ContractX2d<NCOMP, P1d, Q1d>(data, r_U + comp, c_G, r_t); 351 ContractY2d<NCOMP, P1d, Q1d>(data, r_t, c_B, r_V + comp + 0*NCOMP); 352 ContractX2d<NCOMP, P1d, Q1d>(data, r_U + comp, c_B, r_t); 353 ContractY2d<NCOMP, P1d, Q1d>(data, r_t, c_G, r_V + comp + 1*NCOMP); 354 } 355 } 356 357 //------------------------------------------------------------------------------ 358 // 2D derivatives transpose 359 //------------------------------------------------------------------------------ 360 template <int NCOMP, int P1d, int Q1d> 361 inline __device__ void gradTranspose2d(BackendData &data, const CeedScalar *__restrict__ r_U, const CeedScalar *c_B, const CeedScalar *c_G, CeedScalar *__restrict__ r_V) { 362 CeedScalar r_t[1]; 363 for (CeedInt comp = 0; comp < NCOMP; comp++) { 364 ContractYTranspose2d<NCOMP, P1d, Q1d>(data, r_U + comp + 0*NCOMP, c_B, r_t); 365 ContractXTranspose2d<NCOMP, P1d, Q1d>(data, r_t, c_G, r_V + comp); 366 ContractYTranspose2d<NCOMP, P1d, Q1d>(data, r_U + comp + 1*NCOMP, c_G, r_t); 367 ContractXTransposeAdd2d<NCOMP, P1d, Q1d>(data, r_t, c_B, r_V + comp); 368 } 369 } 370 371 //------------------------------------------------------------------------------ 372 // 3D 373 //------------------------------------------------------------------------------ 374 375 //------------------------------------------------------------------------------ 376 // L-vector -> E-vector, offsets provided 377 //------------------------------------------------------------------------------ 378 template <int NCOMP, int COMPSTRIDE, int P1d> 379 inline __device__ void readDofsOffset3d(BackendData &data, const CeedInt nnodes, const CeedInt elem, const CeedInt *__restrict__ indices, const CeedScalar *__restrict__ d_u, CeedScalar *r_u) { 380 if (data.tidx < P1d && data.tidy < P1d) 381 for (CeedInt z = 0; z < P1d; ++z) { 382 const CeedInt node = data.tidx + data.tidy*P1d + z*P1d*P1d; 383 const CeedInt ind = indices[node + elem * P1d*P1d*P1d]; 384 for (CeedInt comp = 0; comp < NCOMP; ++comp) 385 r_u[z+comp*P1d] = d_u[ind + COMPSTRIDE * comp]; 386 } 387 } 388 389 //------------------------------------------------------------------------------ 390 // L-vector -> E-vector, strided 391 //------------------------------------------------------------------------------ 392 template <int NCOMP, int P1d, int STRIDES_NODE, int STRIDES_COMP, int STRIDES_ELEM> 393 inline __device__ void readDofsStrided3d(BackendData &data, const CeedInt elem, const CeedScalar *__restrict__ d_u, CeedScalar *r_u) { 394 if (data.tidx < P1d && data.tidy < P1d) 395 for (CeedInt z = 0; z < P1d; ++z) { 396 const CeedInt node = data.tidx + data.tidy*P1d + z*P1d*P1d; 397 const CeedInt ind = node * STRIDES_NODE + elem * STRIDES_ELEM; 398 for (CeedInt comp = 0; comp < NCOMP; ++comp) 399 r_u[z+comp*P1d] = d_u[ind + comp * STRIDES_COMP]; 400 } 401 } 402 403 //------------------------------------------------------------------------------ 404 // E-vector -> Q-vector, offests provided 405 //------------------------------------------------------------------------------ 406 template <int NCOMP, int COMPSTRIDE, int Q1d> 407 inline __device__ void readSliceQuadsOffset3d(BackendData &data, const CeedInt nquads, const CeedInt elem, const CeedInt q, const CeedInt *__restrict__ indices, const CeedScalar *__restrict__ d_u, CeedScalar *r_u) { 408 if (data.tidx < Q1d && data.tidy < Q1d) { 409 const CeedInt node = data.tidx + data.tidy*Q1d + q*Q1d*Q1d; 410 const CeedInt ind = indices[node + elem * Q1d*Q1d*Q1d];; 411 for (CeedInt comp = 0; comp < NCOMP; ++comp) 412 r_u[comp] = d_u[ind + COMPSTRIDE * comp]; 413 } 414 } 415 416 //------------------------------------------------------------------------------ 417 // E-vector -> Q-vector, strided 418 //------------------------------------------------------------------------------ 419 template <int NCOMP, int Q1d, int STRIDES_NODE, int STRIDES_COMP, int STRIDES_ELEM> 420 inline __device__ void readSliceQuadsStrided3d(BackendData &data, const CeedInt elem, const CeedInt q, const CeedScalar *__restrict__ d_u, CeedScalar *r_u) { 421 if (data.tidx < Q1d && data.tidy < Q1d) { 422 const CeedInt node = data.tidx + data.tidy*Q1d + q*Q1d*Q1d; 423 const CeedInt ind = node * STRIDES_NODE + elem * STRIDES_ELEM; 424 for (CeedInt comp = 0; comp < NCOMP; ++comp) 425 r_u[comp] = d_u[ind + comp * STRIDES_COMP]; 426 } 427 } 428 429 //------------------------------------------------------------------------------ 430 // E-vector -> L-vector, offsets provided 431 //------------------------------------------------------------------------------ 432 template <int NCOMP, int COMPSTRIDE, int P1d> 433 inline __device__ void writeDofsOffset3d(BackendData &data, const CeedInt nnodes, const CeedInt elem, const CeedInt *__restrict__ indices, const CeedScalar *r_v, CeedScalar *d_v) { 434 if (data.tidx < P1d && data.tidy < P1d) 435 for (CeedInt z = 0; z < P1d; ++z) { 436 const CeedInt node = data.tidx + data.tidy*P1d + z*P1d*P1d; 437 const CeedInt ind = indices[node + elem * P1d*P1d*P1d]; 438 for (CeedInt comp = 0; comp < NCOMP; ++comp) 439 atomicAdd(&d_v[ind + COMPSTRIDE * comp], r_v[z+comp*P1d]); 440 } 441 } 442 443 //------------------------------------------------------------------------------ 444 // E-vector -> L-vector, strided 445 //------------------------------------------------------------------------------ 446 template <int NCOMP, int P1d, int STRIDES_NODE, int STRIDES_COMP, int STRIDES_ELEM> 447 inline __device__ void writeDofsStrided3d(BackendData &data, const CeedInt elem, const CeedScalar *r_v, CeedScalar *d_v) { 448 if (data.tidx < P1d && data.tidy < P1d) 449 for (CeedInt z = 0; z < P1d; ++z) { 450 const CeedInt node = data.tidx + data.tidy*P1d + z*P1d*P1d; 451 const CeedInt ind = node * STRIDES_NODE + elem * STRIDES_ELEM; 452 for (CeedInt comp = 0; comp < NCOMP; ++comp) 453 d_v[ind + comp * STRIDES_COMP] += r_v[z+comp*P1d]; 454 } 455 } 456 457 //------------------------------------------------------------------------------ 458 // 3D tensor contract x 459 //------------------------------------------------------------------------------ 460 template <int NCOMP, int P1d, int Q1d> 461 inline __device__ void ContractX3d(BackendData &data, const CeedScalar *U, const CeedScalar *B, CeedScalar *V) { 462 CeedScalar r_B[P1d]; 463 for (CeedInt i = 0; i < P1d; ++i) 464 r_B[i] = B[i + data.tidx*P1d]; 465 466 for (CeedInt k = 0; k < P1d; ++k) { 467 data.slice[data.tidx+data.tidy*T1d] = U[k]; 468 __syncthreads(); 469 V[k] = 0.0; 470 if (data.tidx < Q1d && data.tidy < P1d) 471 for (CeedInt i = 0; i < P1d; ++i) 472 V[k] += r_B[i] * data.slice[i + data.tidy*T1d]; // Contract x direction 473 __syncthreads(); 474 } 475 } 476 477 //------------------------------------------------------------------------------ 478 // 3D tensor contract y 479 //------------------------------------------------------------------------------ 480 template <int NCOMP, int P1d, int Q1d> 481 inline __device__ void ContractY3d(BackendData &data, const CeedScalar *U, const CeedScalar *B, CeedScalar *V) { 482 CeedScalar r_B[P1d]; 483 for (CeedInt i = 0; i < P1d; ++i) 484 r_B[i] = B[i + data.tidy*P1d]; 485 486 for (CeedInt k = 0; k < P1d; ++k) { 487 data.slice[data.tidx+data.tidy*T1d] = U[k]; 488 __syncthreads(); 489 V[k] = 0.0; 490 if (data.tidx < Q1d && data.tidy < Q1d) 491 for (CeedInt i = 0; i < P1d; ++i) 492 V[k] += r_B[i] * data.slice[data.tidx + i*T1d]; // Contract y direction 493 __syncthreads(); 494 } 495 } 496 497 //------------------------------------------------------------------------------ 498 // 3D tensor contract z 499 //------------------------------------------------------------------------------ 500 template <int NCOMP, int P1d, int Q1d> 501 inline __device__ void ContractZ3d(BackendData &data, const CeedScalar *U, const CeedScalar *B, CeedScalar *V) { 502 for (CeedInt k = 0; k < Q1d; ++k) { 503 V[k] = 0.0; 504 if (data.tidx < Q1d && data.tidy < Q1d) 505 for (CeedInt i = 0; i < P1d; ++i) 506 V[k] += B[i + k*P1d] * U[i]; // Contract z direction 507 } 508 } 509 510 //------------------------------------------------------------------------------ 511 // 3D transpose tensor contract z 512 //------------------------------------------------------------------------------ 513 template <int NCOMP, int P1d, int Q1d> 514 inline __device__ void ContractTransposeZ3d(BackendData &data, const CeedScalar *U, const CeedScalar *B, CeedScalar *V) { 515 for (CeedInt k = 0; k < P1d; ++k) { 516 V[k] = 0.0; 517 if (data.tidx < Q1d && data.tidy < Q1d) 518 for (CeedInt i = 0; i < Q1d; ++i) 519 V[k] += B[k + i*P1d] * U[i]; // Contract z direction 520 } 521 } 522 523 //------------------------------------------------------------------------------ 524 // 3D transpose tensor contract y 525 //------------------------------------------------------------------------------ 526 template <int NCOMP, int P1d, int Q1d> 527 inline __device__ void ContractTransposeY3d(BackendData &data, const CeedScalar *U, const CeedScalar *B, CeedScalar *V) { 528 CeedScalar r_B[Q1d]; 529 for (CeedInt i = 0; i < Q1d; ++i) 530 r_B[i] = B[data.tidy + i*P1d]; 531 532 for (CeedInt k = 0; k < P1d; ++k) { 533 data.slice[data.tidx+data.tidy*T1d] = U[k]; 534 __syncthreads(); 535 V[k] = 0.0; 536 if (data.tidx < Q1d && data.tidy < P1d) 537 for (CeedInt i = 0; i < Q1d; ++i) 538 V[k] += r_B[i] * data.slice[data.tidx + i*T1d]; // Contract y direction 539 __syncthreads(); 540 } 541 } 542 543 //------------------------------------------------------------------------------ 544 // 3D transpose tensor contract add y 545 //------------------------------------------------------------------------------ 546 template <int NCOMP, int P1d, int Q1d> 547 inline __device__ void ContractTransposeAddY3d(BackendData &data, const CeedScalar *U, const CeedScalar *B, CeedScalar *V) { 548 CeedScalar r_B[Q1d]; 549 for (CeedInt i = 0; i < Q1d; ++i) 550 r_B[i] = B[data.tidy + i*P1d]; 551 552 for (CeedInt k = 0; k < P1d; ++k) { 553 data.slice[data.tidx+data.tidy*T1d] = U[k]; 554 __syncthreads(); 555 if (data.tidx < Q1d && data.tidy < P1d) 556 for (CeedInt i = 0; i < Q1d; ++i) 557 V[k] += r_B[i] * data.slice[data.tidx + i*T1d]; // Contract y direction 558 __syncthreads(); 559 } 560 } 561 562 //------------------------------------------------------------------------------ 563 // 3D transpose tensor contract x 564 //------------------------------------------------------------------------------ 565 template <int NCOMP, int P1d, int Q1d> 566 inline __device__ void ContractTransposeX3d(BackendData &data, const CeedScalar *U, const CeedScalar *B, CeedScalar *V) { 567 CeedScalar r_B[Q1d]; 568 for (CeedInt i = 0; i < Q1d; ++i) 569 r_B[i] = B[data.tidx + i*P1d]; 570 571 for (CeedInt k = 0; k < P1d; ++k) { 572 data.slice[data.tidx+data.tidy*T1d] = U[k]; 573 __syncthreads(); 574 V[k] = 0.0; 575 if (data.tidx < P1d && data.tidy < P1d) 576 for (CeedInt i = 0; i < Q1d; ++i) 577 V[k] += r_B[i] * data.slice[i + data.tidy*T1d]; // Contract x direction 578 __syncthreads(); 579 } 580 } 581 582 //------------------------------------------------------------------------------ 583 // 3D transpose tensor contract add x 584 //------------------------------------------------------------------------------ 585 template <int NCOMP, int P1d, int Q1d> 586 inline __device__ void ContractTransposeAddX3d(BackendData &data, const CeedScalar *U, const CeedScalar *B, CeedScalar *V) { 587 CeedScalar r_B[Q1d]; 588 for (CeedInt i = 0; i < Q1d; ++i) 589 r_B[i] = B[data.tidx + i*P1d]; 590 591 for (CeedInt k = 0; k < P1d; ++k) { 592 data.slice[data.tidx+data.tidy*T1d] = U[k]; 593 __syncthreads(); 594 if (data.tidx < P1d && data.tidy < P1d) 595 for (CeedInt i = 0; i < Q1d; ++i) 596 V[k] += r_B[i] * data.slice[i + data.tidy*T1d]; // Contract x direction 597 __syncthreads(); 598 } 599 } 600 601 //------------------------------------------------------------------------------ 602 // 3D interpolate to quadrature points 603 //------------------------------------------------------------------------------ 604 template <int NCOMP, int P1d, int Q1d> 605 inline __device__ void interp3d(BackendData &data, const CeedScalar *__restrict__ r_U, const CeedScalar *c_B, CeedScalar *__restrict__ r_V) { 606 CeedScalar r_t1[T1d]; 607 CeedScalar r_t2[T1d]; 608 for (CeedInt comp = 0; comp < NCOMP; comp++) { 609 ContractX3d<NCOMP, P1d, Q1d>(data, r_U + comp*P1d, c_B, r_t1); 610 ContractY3d<NCOMP, P1d, Q1d>(data, r_t1, c_B, r_t2); 611 ContractZ3d<NCOMP, P1d, Q1d>(data, r_t2, c_B, r_V + comp*Q1d); 612 } 613 } 614 615 //------------------------------------------------------------------------------ 616 // 3D interpolate transpose 617 //------------------------------------------------------------------------------ 618 template <int NCOMP, int P1d, int Q1d> 619 inline __device__ void interpTranspose3d(BackendData &data, const CeedScalar *__restrict__ r_U, const CeedScalar *c_B, CeedScalar *__restrict__ r_V) { 620 CeedScalar r_t1[T1d]; 621 CeedScalar r_t2[T1d]; 622 for (CeedInt comp = 0; comp < NCOMP; comp++) { 623 ContractTransposeZ3d<NCOMP, P1d, Q1d>(data, r_U + comp*Q1d, c_B, r_t1); 624 ContractTransposeY3d<NCOMP, P1d, Q1d>(data, r_t1, c_B, r_t2); 625 ContractTransposeX3d<NCOMP, P1d, Q1d>(data, r_t2, c_B, r_V + comp*P1d); 626 } 627 } 628 629 //------------------------------------------------------------------------------ 630 // 3D derivatives at quadrature points 631 //------------------------------------------------------------------------------ 632 template <int NCOMP, int P1d, int Q1d> 633 inline __device__ void grad3d(BackendData &data, const CeedScalar *__restrict__ r_U, const CeedScalar *c_B, const CeedScalar *c_G, CeedScalar *__restrict__ r_V) { 634 CeedScalar r_t1[T1d]; 635 CeedScalar r_t2[T1d]; 636 for (CeedInt comp = 0; comp < NCOMP; comp++) { 637 ContractX3d<NCOMP, P1d, Q1d>(data, r_U + comp*P1d, c_G, r_t1); 638 ContractY3d<NCOMP, P1d, Q1d>(data, r_t1, c_B, r_t2); 639 ContractZ3d<NCOMP, P1d, Q1d>(data, r_t2, c_B, r_V + comp*Q1d + 0*NCOMP*Q1d); 640 ContractX3d<NCOMP, P1d, Q1d>(data, r_U + comp*P1d, c_B, r_t1); 641 ContractY3d<NCOMP, P1d, Q1d>(data, r_t1, c_G, r_t2); 642 ContractZ3d<NCOMP, P1d, Q1d>(data, r_t2, c_B, r_V + comp*Q1d + 1*NCOMP*Q1d); 643 ContractX3d<NCOMP, P1d, Q1d>(data, r_U + comp*P1d, c_B, r_t1); 644 ContractY3d<NCOMP, P1d, Q1d>(data, r_t1, c_B, r_t2); 645 ContractZ3d<NCOMP, P1d, Q1d>(data, r_t2, c_G, r_V + comp*Q1d + 2*NCOMP*Q1d); 646 } 647 } 648 649 //------------------------------------------------------------------------------ 650 // 3D derivatives transpose 651 //------------------------------------------------------------------------------ 652 template <int NCOMP, int P1d, int Q1d> 653 inline __device__ void gradTranspose3d(BackendData &data, const CeedScalar *__restrict__ r_U, const CeedScalar *c_B, const CeedScalar *c_G, CeedScalar *__restrict__ r_V) { 654 CeedScalar r_t1[T1d]; 655 CeedScalar r_t2[T1d]; 656 for (CeedInt comp = 0; comp < NCOMP; comp++) { 657 ContractTransposeZ3d<NCOMP, P1d, Q1d>(data, r_U + comp*Q1d + 0*NCOMP*Q1d, c_B, r_t1); 658 ContractTransposeY3d<NCOMP, P1d, Q1d>(data, r_t1, c_B, r_t2); 659 ContractTransposeX3d<NCOMP, P1d, Q1d>(data, r_t2, c_G, r_V + comp*P1d); 660 ContractTransposeZ3d<NCOMP, P1d, Q1d>(data, r_U + comp*Q1d + 1*NCOMP*Q1d, c_B, r_t1); 661 ContractTransposeY3d<NCOMP, P1d, Q1d>(data, r_t1, c_G, r_t2); 662 ContractTransposeAddX3d<NCOMP,P1d, Q1d>(data, r_t2, c_B, r_V + comp*P1d); 663 ContractTransposeZ3d<NCOMP, P1d, Q1d>(data, r_U + comp*Q1d + 2*NCOMP*Q1d, c_G, r_t1); 664 ContractTransposeY3d<NCOMP, P1d, Q1d>(data, r_t1, c_B, r_t2); 665 ContractTransposeAddX3d<NCOMP, P1d, Q1d>(data, r_t2, c_B, r_V + comp*P1d); 666 } 667 } 668 669 //------------------------------------------------------------------------------ 670 // 3D collocated derivatives computation 671 //------------------------------------------------------------------------------ 672 template <int NCOMP, int Q1d> 673 inline __device__ void gradCollo3d(BackendData &data, const CeedInt q, const CeedScalar *__restrict__ r_U, const CeedScalar *c_G, CeedScalar *__restrict__ r_V) { 674 if (data.tidx < Q1d && data.tidy < Q1d) { 675 for (CeedInt comp = 0; comp < NCOMP; ++comp) { 676 data.slice[data.tidx + data.tidy*T1d] = r_U[q + comp*Q1d]; 677 __syncthreads(); 678 // X derivative 679 r_V[comp+0*NCOMP] = 0.0; 680 for (CeedInt i = 0; i < Q1d; ++i) 681 r_V[comp+0*NCOMP] += c_G[i + data.tidx*Q1d] * data.slice[i + data.tidy*T1d]; // Contract x direction (X derivative) 682 // Y derivative 683 r_V[comp+1*NCOMP] = 0.0; 684 for (CeedInt i = 0; i < Q1d; ++i) 685 r_V[comp+1*NCOMP] += c_G[i + data.tidy*Q1d] * data.slice[data.tidx + i*T1d]; // Contract y direction (Y derivative) 686 // Z derivative 687 r_V[comp+2*NCOMP] = 0.0; 688 for (CeedInt i = 0; i < Q1d; ++i) 689 r_V[comp+2*NCOMP] += c_G[i + q*Q1d] * r_U[i + comp*Q1d]; // Contract z direction (Z derivative) 690 __syncthreads(); 691 } 692 } 693 } 694 695 //------------------------------------------------------------------------------ 696 // 3D collocated derivatives transpose 697 //------------------------------------------------------------------------------ 698 template <int NCOMP, int Q1d> 699 inline __device__ void gradColloTranspose3d(BackendData &data, const CeedInt q, const CeedScalar *__restrict__ r_U, const CeedScalar *c_G, CeedScalar *__restrict__ r_V) { 700 if (data.tidx < Q1d && data.tidy < Q1d) { 701 for (CeedInt comp = 0; comp < NCOMP; ++comp) { 702 // X derivative 703 data.slice[data.tidx + data.tidy*T1d] = r_U[comp + 0*NCOMP]; 704 __syncthreads(); 705 for (CeedInt i = 0; i < Q1d; ++i) 706 r_V[q+comp*Q1d] += c_G[data.tidx + i*Q1d] * data.slice[i + data.tidy*T1d]; // Contract x direction (X derivative) 707 __syncthreads(); 708 // Y derivative 709 data.slice[data.tidx + data.tidy*T1d] = r_U[comp + 1*NCOMP]; 710 __syncthreads(); 711 for (CeedInt i = 0; i < Q1d; ++i) 712 r_V[q+comp*Q1d] += c_G[data.tidy + i*Q1d] * data.slice[data.tidx + i*T1d]; // Contract y direction (Y derivative) 713 __syncthreads(); 714 // Z derivative 715 for (CeedInt i = 0; i < Q1d; ++i) 716 r_V[i+comp*Q1d] += c_G[i + q*Q1d] * r_U[comp + 2*NCOMP]; // PARTIAL contract z direction (Z derivative) 717 } 718 } 719 } 720 721 //------------------------------------------------------------------------------ 722 // 1D quadrature weights 723 //------------------------------------------------------------------------------ 724 template <int Q1d> 725 inline __device__ void weight1d(BackendData &data, const CeedScalar *__restrict__ qweight1d, CeedScalar *w) { 726 *w = (data.tidx < Q1d) ? qweight1d[data.tidx] : 0.0; 727 } 728 729 //------------------------------------------------------------------------------ 730 // 2D quadrature weights 731 //------------------------------------------------------------------------------ 732 template <int Q1d> 733 inline __device__ void weight2d(BackendData &data, const CeedScalar *__restrict__ qweight1d, CeedScalar *w) { 734 *w = (data.tidx < Q1d && data.tidy < Q1d) ? 735 qweight1d[data.tidx]*qweight1d[data.tidy] : 0.0; 736 } 737 738 //------------------------------------------------------------------------------ 739 // 3D quadrature weights 740 //------------------------------------------------------------------------------ 741 template <int Q1d> 742 inline __device__ void weight3d(BackendData &data, const CeedScalar *__restrict__ qweight1d, CeedScalar *w) { 743 const bool quad = (data.tidx < Q1d && data.tidy < Q1d); 744 const CeedScalar pw = quad ? qweight1d[data.tidx]*qweight1d[data.tidy] : 0.0; 745 for (CeedInt z = 0; z < Q1d; ++z) 746 w[z] = quad ? pw*qweight1d[z] : 0.0; 747 } 748 749 ); 750 //------------------------------------------------------------------------------ 751 // Build singe operator kernel 752 //------------------------------------------------------------------------------ 753 extern "C" int CeedCudaGenOperatorBuild(CeedOperator op) { 754 755 using std::ostringstream; 756 using std::string; 757 int ierr; 758 bool setupdone; 759 ierr = CeedOperatorIsSetupDone(op, &setupdone); CeedChkBackend(ierr); 760 if (setupdone) return CEED_ERROR_SUCCESS; 761 Ceed ceed; 762 ierr = CeedOperatorGetCeed(op, &ceed); CeedChkBackend(ierr); 763 CeedOperator_Cuda_gen *data; 764 ierr = CeedOperatorGetData(op, &data); CeedChkBackend(ierr); 765 CeedQFunction qf; 766 CeedQFunction_Cuda_gen *qf_data; 767 ierr = CeedOperatorGetQFunction(op, &qf); CeedChkBackend(ierr); 768 ierr = CeedQFunctionGetData(qf, &qf_data); CeedChkBackend(ierr); 769 CeedInt Q, P1d = 0, Q1d = 0, numelements, elemsize, numinputfields, 770 numoutputfields, ncomp, dim = 0, lsize; 771 ierr = CeedOperatorGetNumQuadraturePoints(op, &Q); CeedChkBackend(ierr); 772 ierr = CeedOperatorGetNumElements(op, &numelements); CeedChkBackend(ierr); 773 CeedOperatorField *opinputfields, *opoutputfields; 774 ierr = CeedOperatorGetFields(op, &numinputfields, &opinputfields, &numoutputfields, &opoutputfields); 775 CeedChkBackend(ierr); 776 CeedQFunctionField *qfinputfields, *qfoutputfields; 777 ierr = CeedQFunctionGetFields(qf, NULL, &qfinputfields, NULL, &qfoutputfields); 778 CeedChkBackend(ierr); 779 CeedEvalMode emode; 780 CeedBasis basis; 781 CeedBasis_Cuda_shared *basis_data; 782 CeedElemRestriction Erestrict; 783 CeedElemRestriction_Cuda *restr_data; 784 785 // Check for restriction only identity operator 786 bool is_identity_qf; 787 ierr = CeedQFunctionIsIdentity(qf, &is_identity_qf); CeedChkBackend(ierr); 788 if (is_identity_qf) { 789 CeedEvalMode emodein, emodeout; 790 ierr = CeedQFunctionFieldGetEvalMode(qfinputfields[0], &emodein); CeedChkBackend(ierr); 791 ierr = CeedQFunctionFieldGetEvalMode(qfoutputfields[0], &emodeout); CeedChkBackend(ierr); 792 if (emodein == CEED_EVAL_NONE && emodeout == CEED_EVAL_NONE) 793 // LCOV_EXCL_START 794 return CeedError(ceed, CEED_ERROR_BACKEND, 795 "Backend does not implement restriction only identity operators"); 796 // LCOV_EXCL_STOP 797 } 798 799 ostringstream code; 800 string devFunctions(deviceFunctions); 801 802 // Add atomicAdd function for old NVidia architectures 803 struct cudaDeviceProp prop; 804 Ceed_Cuda *ceed_data; 805 ierr = CeedGetData(ceed, &ceed_data); CeedChkBackend(ierr); CeedChkBackend(ierr); 806 ierr = cudaGetDeviceProperties(&prop, ceed_data->device_id); CeedChkBackend(ierr); 807 if ((prop.major < 6) && (CEED_SCALAR_TYPE != CEED_SCALAR_FP32)){ 808 code << atomicAdd; 809 } 810 811 code << devFunctions; 812 813 string qFunction(qf_data->qFunctionSource); 814 string qFunctionName(qf_data->qFunctionName); 815 string oper; 816 oper = "CeedKernel_Cuda_gen_" + qFunctionName; 817 818 code << "\n#define CEED_QFUNCTION(name) inline __device__ int name\n"; 819 code << "#define CEED_QFUNCTION_HELPER inline __device__\n"; 820 code << "#define CeedPragmaSIMD\n"; 821 code << "#define CEED_ERROR_SUCCESS 0\n\n"; 822 823 // Find dim and Q1d 824 bool useCollograd = true; 825 data->maxP1d = 0; 826 for (CeedInt i = 0; i < numinputfields; i++) { 827 ierr = CeedOperatorFieldGetBasis(opinputfields[i], &basis); CeedChkBackend(ierr); 828 if (basis != CEED_BASIS_COLLOCATED) { 829 ierr = CeedBasisGetData(basis, &basis_data); CeedChkBackend(ierr); 830 ierr = CeedQFunctionFieldGetEvalMode(qfinputfields[i], &emode); 831 CeedChkBackend(ierr); 832 833 // Check for collocated gradient 834 useCollograd = useCollograd && basis_data->d_collograd1d; 835 836 // Collect dim and Q1d 837 ierr = CeedBasisGetDimension(basis, &dim); CeedChkBackend(ierr); 838 bool isTensor; 839 ierr = CeedBasisIsTensor(basis, &isTensor); CeedChkBackend(ierr); 840 if (isTensor) { 841 ierr = CeedBasisGetNumQuadraturePoints1D(basis, &Q1d); CeedChkBackend(ierr); 842 ierr = CeedBasisGetNumNodes1D(basis, &P1d); CeedChkBackend(ierr); 843 if (P1d>data->maxP1d) data->maxP1d = P1d; 844 } else { 845 // LCOV_EXCL_START 846 return CeedError(ceed, CEED_ERROR_BACKEND, "Backend does not implement operators with non-tensor basis"); 847 // LCOV_EXCL_STOP 848 } 849 } 850 } 851 // Check output bases for Q1d, dim as well 852 // The only imput basis might be CEED_BASIS_COLLOCATED 853 for (CeedInt i = 0; i < numoutputfields; i++) { 854 ierr = CeedOperatorFieldGetBasis(opoutputfields[i], &basis); CeedChkBackend(ierr); 855 856 if (basis != CEED_BASIS_COLLOCATED) { 857 ierr = CeedBasisGetData(basis, &basis_data); CeedChkBackend(ierr); 858 ierr = CeedQFunctionFieldGetEvalMode(qfoutputfields[i], &emode); 859 CeedChkBackend(ierr); 860 861 // Collect dim and Q1d 862 ierr = CeedBasisGetDimension(basis, &dim); CeedChkBackend(ierr); 863 bool isTensor; 864 ierr = CeedBasisIsTensor(basis, &isTensor); CeedChkBackend(ierr); 865 if (isTensor) { 866 ierr = CeedBasisGetNumQuadraturePoints1D(basis, &Q1d); CeedChkBackend(ierr); 867 } else { 868 // LCOV_EXCL_START 869 return CeedError(ceed, CEED_ERROR_BACKEND, "Backend does not implement operators with non-tensor basis"); 870 // LCOV_EXCL_STOP 871 } 872 873 // Check for collocated gradient 874 useCollograd = useCollograd && basis_data->d_collograd1d; 875 } 876 } 877 data->dim = dim; 878 data->Q1d = Q1d; 879 880 // Define CEED_Q_VLA 881 if (dim != 3 || useCollograd) { 882 code << "\n#define CEED_Q_VLA 1\n\n"; 883 } else { 884 code << "\n#define CEED_Q_VLA "<<Q1d<<"\n\n"; 885 } 886 887 code << qFunction; 888 889 // Setup 890 code << "\n// -----------------------------------------------------------------------------\n"; 891 code << "\nextern \"C\" __global__ void "<<oper<<"(CeedInt nelem, void* ctx, CudaFieldsInt indices, CudaFields fields, CudaFields B, CudaFields G, CeedScalar* W) {\n"; 892 for (CeedInt i = 0; i < numinputfields; i++) { 893 ierr = CeedQFunctionFieldGetEvalMode(qfinputfields[i], &emode); 894 CeedChkBackend(ierr); 895 if (emode != CEED_EVAL_WEIGHT) { // Skip CEED_EVAL_WEIGHT 896 code << " const CeedScalar* d_u" <<i<<" = fields.in["<<i<<"];\n"; 897 } 898 } 899 900 for (CeedInt i = 0; i < numoutputfields; i++) { 901 code << " CeedScalar* d_v"<<i<<" = fields.out["<<i<<"];\n"; 902 } 903 904 code << " const CeedInt Dim = "<<dim<<";\n"; 905 code << " const CeedInt Q1d = "<<Q1d<<";\n"; 906 907 code << " extern __shared__ CeedScalar slice[];\n"; 908 code << " BackendData data;\n"; 909 code << " data.tidx = threadIdx.x;\n"; 910 code << " data.tidy = threadIdx.y;\n"; 911 code << " data.tidz = threadIdx.z;\n"; 912 code << " data.tid = threadIdx.x + threadIdx.y*blockDim.x + threadIdx.z*blockDim.y*blockDim.x;\n"; 913 code << " data.slice = slice+data.tidz*T1d"<<(dim>1?"*T1d":"")<<";\n"; 914 915 code << "\n // -- Input field constants and basis data --\n"; 916 //Initialize constants, and matrices B and G 917 for (CeedInt i = 0; i < numinputfields; i++) { 918 code << " // ---- Input field "<<i<<" ----\n"; 919 // Get elemsize, emode, ncomp 920 ierr = CeedOperatorFieldGetElemRestriction(opinputfields[i], &Erestrict); 921 CeedChkBackend(ierr); 922 ierr = CeedElemRestrictionGetElementSize(Erestrict, &elemsize); 923 CeedChkBackend(ierr); 924 ierr = CeedQFunctionFieldGetEvalMode(qfinputfields[i], &emode); 925 CeedChkBackend(ierr); 926 ierr = CeedElemRestrictionGetNumComponents(Erestrict, &ncomp); 927 CeedChkBackend(ierr); 928 929 // Set field constants 930 if (emode != CEED_EVAL_WEIGHT) { 931 ierr = CeedOperatorFieldGetBasis(opinputfields[i], &basis); CeedChkBackend(ierr); 932 if (basis != CEED_BASIS_COLLOCATED) { 933 ierr = CeedBasisGetNumNodes1D(basis, &P1d); CeedChkBackend(ierr); 934 code << " const CeedInt P_in_"<<i<<" = "<<P1d<<";\n"; 935 } else { 936 code << " const CeedInt P_in_"<<i<<" = "<<Q1d<<";\n"; 937 } 938 code << " const CeedInt ncomp_in_"<<i<<" = "<<ncomp<<";\n"; 939 } 940 941 // Load basis data 942 code << " // EvalMode: "<<CeedEvalModes[emode]<<"\n"; 943 switch (emode) { 944 case CEED_EVAL_NONE: 945 break; 946 case CEED_EVAL_INTERP: 947 ierr = CeedBasisGetData(basis, &basis_data); CeedChkBackend(ierr); 948 data->B.in[i] = basis_data->d_interp1d; 949 code << " __shared__ CeedScalar s_B_in_"<<i<<"["<<P1d*Q1d<<"];\n"; 950 code << " loadMatrix<P_in_"<<i<<",Q1d>(data, B.in["<<i<<"], s_B_in_"<<i<<");\n"; 951 break; 952 case CEED_EVAL_GRAD: 953 ierr = CeedBasisGetData(basis, &basis_data); CeedChkBackend(ierr); 954 data->B.in[i] = basis_data->d_interp1d; 955 code << " __shared__ CeedScalar s_B_in_"<<i<<"["<<P1d*Q1d<<"];\n"; 956 code << " loadMatrix<P_in_"<<i<<",Q1d>(data, B.in["<<i<<"], s_B_in_"<<i<<");\n"; 957 if (useCollograd) { 958 data->G.in[i] = basis_data->d_collograd1d; 959 code << " __shared__ CeedScalar s_G_in_"<<i<<"["<<Q1d*Q1d<<"];\n"; 960 code << " loadMatrix<Q1d,Q1d>(data, G.in["<<i<<"], s_G_in_"<<i<<");\n"; 961 } else { 962 data->G.in[i] = basis_data->d_grad1d; 963 code << " __shared__ CeedScalar s_G_in_"<<i<<"["<<P1d*Q1d<<"];\n"; 964 code << " loadMatrix<P_in_"<<i<<",Q1d>(data, G.in["<<i<<"], s_G_in_"<<i<<");\n"; 965 } 966 break; 967 case CEED_EVAL_WEIGHT: 968 break; // No action 969 case CEED_EVAL_DIV: 970 break; // TODO: Not implemented 971 case CEED_EVAL_CURL: 972 break; // TODO: Not implemented 973 } 974 } 975 976 code << "\n // -- Output field constants and basis data --\n"; 977 for (CeedInt i = 0; i < numoutputfields; i++) { 978 code << " // ---- Output field "<<i<<" ----\n"; 979 // Get elemsize, emode, ncomp 980 ierr = CeedOperatorFieldGetElemRestriction(opoutputfields[i], &Erestrict); 981 CeedChkBackend(ierr); 982 ierr = CeedElemRestrictionGetElementSize(Erestrict, &elemsize); 983 CeedChkBackend(ierr); 984 ierr = CeedQFunctionFieldGetEvalMode(qfoutputfields[i], &emode); 985 CeedChkBackend(ierr); 986 ierr = CeedElemRestrictionGetNumComponents(Erestrict, &ncomp); 987 CeedChkBackend(ierr); 988 989 // Set field constants 990 ierr = CeedOperatorFieldGetBasis(opoutputfields[i], &basis); CeedChkBackend(ierr); 991 if (basis != CEED_BASIS_COLLOCATED) { 992 ierr = CeedBasisGetNumNodes1D(basis, &P1d); CeedChkBackend(ierr); 993 code << " const CeedInt P_out_"<<i<<" = "<<P1d<<";\n"; 994 } else { 995 code << " const CeedInt P_out_"<<i<<" = "<<Q1d<<";\n"; 996 } 997 code << " const CeedInt ncomp_out_"<<i<<" = "<<ncomp<<";\n"; 998 999 // Load basis data 1000 code << " // EvalMode: "<<CeedEvalModes[emode]<<"\n"; 1001 switch (emode) { 1002 case CEED_EVAL_NONE: 1003 break; // No action 1004 case CEED_EVAL_INTERP: 1005 ierr = CeedBasisGetData(basis, &basis_data); CeedChkBackend(ierr); 1006 data->B.out[i] = basis_data->d_interp1d; 1007 code << " __shared__ CeedScalar s_B_out_"<<i<<"["<<P1d*Q1d<<"];\n"; 1008 code << " loadMatrix<P_out_"<<i<<",Q1d>(data, B.out["<<i<<"], s_B_out_"<<i<<");\n"; 1009 break; 1010 case CEED_EVAL_GRAD: 1011 ierr = CeedBasisGetData(basis, &basis_data); CeedChkBackend(ierr); 1012 data->B.out[i] = basis_data->d_interp1d; 1013 code << " __shared__ CeedScalar s_B_out_"<<i<<"["<<P1d*Q1d<<"];\n"; 1014 code << " loadMatrix<P_out_"<<i<<",Q1d>(data, B.out["<<i<<"], s_B_out_"<<i<<");\n"; 1015 if (useCollograd) { 1016 data->G.out[i] = basis_data->d_collograd1d; 1017 code << " __shared__ CeedScalar s_G_out_"<<i<<"["<<Q1d*Q1d<<"];\n"; 1018 code << " loadMatrix<Q1d,Q1d>(data, G.out["<<i<<"], s_G_out_"<<i<<");\n"; 1019 } else { 1020 data->G.out[i] = basis_data->d_grad1d; 1021 code << " __shared__ CeedScalar s_G_out_"<<i<<"["<<P1d*Q1d<<"];\n"; 1022 code << " loadMatrix<P_out_"<<i<<",Q1d>(data, G.out["<<i<<"], s_G_out_"<<i<<");\n"; 1023 } 1024 break; 1025 // LCOV_EXCL_START 1026 case CEED_EVAL_WEIGHT: { 1027 Ceed ceed; 1028 ierr = CeedOperatorGetCeed(op, &ceed); CeedChkBackend(ierr); 1029 return CeedError(ceed, CEED_ERROR_BACKEND, 1030 "CEED_EVAL_WEIGHT cannot be an output evaluation mode"); 1031 break; // Should not occur 1032 } 1033 case CEED_EVAL_DIV: 1034 break; // TODO: Not implemented 1035 case CEED_EVAL_CURL: 1036 break; // TODO: Not implemented 1037 // LCOV_EXCL_STOP 1038 } 1039 } 1040 code << "\n // -- Element loop --\n"; 1041 code << " __syncthreads();\n"; 1042 code << " for (CeedInt elem = blockIdx.x*blockDim.z + threadIdx.z; elem < nelem; elem += gridDim.x*blockDim.z) {\n"; 1043 // Input basis apply if needed 1044 // Generate the correct eval mode code for each input 1045 code << " // -- Input field restrictions and basis actions --\n"; 1046 for (CeedInt i = 0; i < numinputfields; i++) { 1047 code << " // ---- Input field "<<i<<" ----\n"; 1048 // Get elemsize, emode, ncomp 1049 ierr = CeedOperatorFieldGetElemRestriction(opinputfields[i], &Erestrict); 1050 CeedChkBackend(ierr); 1051 ierr = CeedElemRestrictionGetElementSize(Erestrict, &elemsize); 1052 CeedChkBackend(ierr); 1053 ierr = CeedQFunctionFieldGetEvalMode(qfinputfields[i], &emode); 1054 CeedChkBackend(ierr); 1055 ierr = CeedElemRestrictionGetNumComponents(Erestrict, &ncomp); 1056 CeedChkBackend(ierr); 1057 1058 // Restriction 1059 if (emode != CEED_EVAL_WEIGHT && 1060 !((emode == CEED_EVAL_NONE) && useCollograd)) { 1061 code << " CeedScalar r_u"<<i<<"[ncomp_in_"<<i<<"*P_in_"<<i<<"];\n"; 1062 1063 bool isStrided; 1064 ierr = CeedElemRestrictionIsStrided(Erestrict, &isStrided); CeedChkBackend(ierr); 1065 if (!isStrided) { 1066 ierr = CeedElemRestrictionGetLVectorSize(Erestrict, &lsize); 1067 CeedChkBackend(ierr); 1068 code << " const CeedInt lsize_in_"<<i<<" = "<<lsize<<";\n"; 1069 CeedInt compstride; 1070 ierr = CeedElemRestrictionGetCompStride(Erestrict, &compstride); CeedChkBackend(ierr); 1071 code << " // CompStride: "<<compstride<<"\n"; 1072 ierr = CeedElemRestrictionGetData(Erestrict, &restr_data); CeedChkBackend(ierr); 1073 data->indices.in[i] = restr_data->d_ind; 1074 code << " readDofsOffset"<<dim<<"d<ncomp_in_"<<i<<", "<<compstride<<", P_in_"<<i<<">(data, lsize_in_"<<i<<", elem, indices.in["<<i<<"], d_u"<<i<<", r_u"<<i<<");\n"; 1075 } else { 1076 bool backendstrides; 1077 ierr = CeedElemRestrictionHasBackendStrides(Erestrict, &backendstrides); 1078 CeedChkBackend(ierr); 1079 CeedInt nelem; 1080 ierr = CeedElemRestrictionGetNumElements(Erestrict, &nelem); 1081 CeedChkBackend(ierr); 1082 CeedInt strides[3] = {1, elemsize*nelem, elemsize}; 1083 if (!backendstrides) { 1084 ierr = CeedElemRestrictionGetStrides(Erestrict, &strides); 1085 CeedChkBackend(ierr); 1086 } 1087 code << " // Strides: {"<<strides[0]<<", "<<strides[1]<<", "<<strides[2]<<"}\n"; 1088 code << " readDofsStrided"<<dim<<"d<ncomp_in_"<<i<<",P_in_"<<i<<","<<strides[0]<<","<<strides[1]<<","<<strides[2]<<">(data, elem, d_u"<<i<<", r_u"<<i<<");\n"; 1089 } 1090 } 1091 1092 // Basis action 1093 code << " // EvalMode: "<<CeedEvalModes[emode]<<"\n"; 1094 switch (emode) { 1095 case CEED_EVAL_NONE: 1096 if (!useCollograd) { 1097 code << " CeedScalar* r_t"<<i<<" = r_u"<<i<<";\n"; 1098 } 1099 break; 1100 case CEED_EVAL_INTERP: 1101 code << " CeedScalar r_t"<<i<<"[ncomp_in_"<<i<<"*Q1d];\n"; 1102 code << " interp"<<dim<<"d<ncomp_in_"<<i<<",P_in_"<<i<<",Q1d>(data, r_u"<<i<<", s_B_in_"<<i<<", r_t"<<i<<");\n"; 1103 break; 1104 case CEED_EVAL_GRAD: 1105 if (useCollograd) { 1106 code << " CeedScalar r_t"<<i<<"[ncomp_in_"<<i<<"*Q1d];\n"; 1107 code << " interp"<<dim<<"d<ncomp_in_"<<i<<",P_in_"<<i<<",Q1d>(data, r_u"<<i<<", s_B_in_"<<i<<", r_t"<<i<<");\n"; 1108 } else { 1109 code << " CeedScalar r_t"<<i<<"[ncomp_in_"<<i<<"*Dim*Q1d];\n"; 1110 code << " grad"<<dim<<"d<ncomp_in_"<<i<<",P_in_"<<i<<",Q1d>(data, r_u"<<i<<", s_B_in_"<<i<<", s_G_in_"<<i<<", r_t"<<i<<");\n"; 1111 } 1112 break; 1113 case CEED_EVAL_WEIGHT: 1114 code << " CeedScalar r_t"<<i<<"[Q1d];\n"; 1115 ierr = CeedOperatorFieldGetBasis(opinputfields[i], &basis); CeedChkBackend(ierr); 1116 ierr = CeedBasisGetData(basis, &basis_data); CeedChkBackend(ierr); 1117 data->W = basis_data->d_qweight1d; 1118 code << " weight"<<dim<<"d<Q1d>(data, W, r_t"<<i<<");\n"; 1119 break; // No action 1120 case CEED_EVAL_DIV: 1121 break; // TODO: Not implemented 1122 case CEED_EVAL_CURL: 1123 break; // TODO: Not implemented 1124 } 1125 } 1126 1127 // Q function 1128 code << "\n // -- Output field setup --\n"; 1129 for (CeedInt i = 0; i < numoutputfields; i++) { 1130 code << "\n // ---- Output field "<<i<<" ----\n"; 1131 ierr = CeedQFunctionFieldGetEvalMode(qfoutputfields[i], &emode); 1132 CeedChkBackend(ierr); 1133 if (emode==CEED_EVAL_GRAD) 1134 { 1135 if (useCollograd) { 1136 //Accumulator for gradient slices 1137 code << " CeedScalar r_tt"<<i<<"[ncomp_out_"<<i<<"*Q1d];\n"; 1138 code << " for (CeedInt i = 0; i < ncomp_out_"<<i<<"; ++i) {\n"; 1139 code << " for (CeedInt j = 0; j < Q1d; ++j) {\n"; 1140 code << " r_tt"<<i<<"[j + i*Q1d] = 0.0;\n"; 1141 code << " }\n"; 1142 code << " }\n"; 1143 } else { 1144 code << " CeedScalar r_tt"<<i<<"[ncomp_out_"<<i<<"*Dim*Q1d];\n"; 1145 } 1146 } 1147 if (emode==CEED_EVAL_NONE || emode==CEED_EVAL_INTERP) 1148 { 1149 code << " CeedScalar r_tt"<<i<<"[ncomp_out_"<<i<<"*Q1d];\n"; 1150 } 1151 } 1152 // We treat quadrature points per slice in 3d to save registers 1153 if (useCollograd) { 1154 code << "\n // Note: Collocated Gradient\n"; 1155 code << "#pragma unroll\n"; 1156 code << " for (CeedInt q=0; q<Q1d; q++) {\n"; 1157 code << " // -- Input fields --\n"; 1158 for (CeedInt i = 0; i < numinputfields; i++) { 1159 code << " // ---- Input field "<<i<<" ----\n"; 1160 // Get elemsize, emode, ncomp 1161 ierr = CeedQFunctionFieldGetEvalMode(qfinputfields[i], &emode); 1162 CeedChkBackend(ierr); 1163 // Basis action 1164 code << " // EvalMode: "<<CeedEvalModes[emode]<<"\n"; 1165 switch (emode) { 1166 case CEED_EVAL_NONE: 1167 code << " CeedScalar r_q"<<i<<"[ncomp_in_"<<i<<"];\n"; 1168 1169 bool isStrided; 1170 ierr = CeedOperatorFieldGetElemRestriction(opinputfields[i], &Erestrict); CeedChkBackend(ierr); 1171 ierr = CeedElemRestrictionGetElementSize(Erestrict, &elemsize); CeedChkBackend(ierr); 1172 ierr = CeedElemRestrictionIsStrided(Erestrict, &isStrided); CeedChkBackend(ierr); 1173 if (!isStrided) { 1174 ierr = CeedElemRestrictionGetLVectorSize(Erestrict, &lsize); 1175 CeedChkBackend(ierr); 1176 code << " const CeedInt lsize_in_"<<i<<" = "<<lsize<<";\n"; 1177 CeedInt compstride; 1178 ierr = CeedElemRestrictionGetCompStride(Erestrict, &compstride); CeedChkBackend(ierr); 1179 code << " // CompStride: "<<compstride<<"\n"; 1180 ierr = CeedElemRestrictionGetData(Erestrict, &restr_data); CeedChkBackend(ierr); 1181 data->indices.in[i] = restr_data->d_ind; 1182 code << " readSliceQuadsOffset"<<"3d<ncomp_in_"<<i<<", "<<compstride<<", Q1d>(data, lsize_in_"<<i<<", elem, q, indices.in["<<i<<"], d_u"<<i<<", r_q"<<i<<");\n"; 1183 } else { 1184 bool backendstrides; 1185 ierr = CeedElemRestrictionHasBackendStrides(Erestrict, &backendstrides); 1186 CeedChkBackend(ierr); 1187 CeedInt nelem; 1188 ierr = CeedElemRestrictionGetNumElements(Erestrict, &nelem); 1189 CeedChkBackend(ierr); 1190 CeedInt strides[3] = {1, elemsize*nelem, elemsize}; 1191 if (!backendstrides) { 1192 ierr = CeedElemRestrictionGetStrides(Erestrict, &strides); 1193 CeedChkBackend(ierr); 1194 } 1195 code << " // Strides: {"<<strides[0]<<", "<<strides[1]<<", "<<strides[2]<<"}\n"; 1196 code << " readSliceQuadsStrided"<<"3d<ncomp_in_"<<i<<",Q1d"","<<strides[0]<<","<<strides[1]<<","<<strides[2]<<">(data, elem, q, d_u"<<i<<", r_q"<<i<<");\n"; 1197 } 1198 break; 1199 case CEED_EVAL_INTERP: 1200 code << " CeedScalar r_q"<<i<<"[ncomp_in_"<<i<<"];\n"; 1201 code << " for (CeedInt j = 0; j < ncomp_in_"<<i<<" ; ++j) {\n"; 1202 code << " r_q"<<i<<"[j] = r_t"<<i<<"[q + j*Q1d];\n"; 1203 code << " }\n"; 1204 break; 1205 case CEED_EVAL_GRAD: 1206 code << " CeedScalar r_q"<<i<<"[ncomp_in_"<<i<<"*Dim];\n"; 1207 code << " gradCollo3d<ncomp_in_"<<i<<",Q1d>(data, q, r_t"<<i<<", s_G_in_"<<i<<", r_q"<<i<<");\n"; 1208 break; 1209 case CEED_EVAL_WEIGHT: 1210 code << " CeedScalar r_q"<<i<<"[1];\n"; 1211 code << " r_q"<<i<<"[0] = r_t"<<i<<"[q];\n"; 1212 break; // No action 1213 case CEED_EVAL_DIV: 1214 break; // TODO: Not implemented 1215 case CEED_EVAL_CURL: 1216 break; // TODO: Not implemented 1217 } 1218 } 1219 code << "\n // -- Output fields --\n"; 1220 for (CeedInt i = 0; i < numoutputfields; i++) { 1221 code << " // ---- Output field "<<i<<" ----\n"; 1222 ierr = CeedQFunctionFieldGetEvalMode(qfoutputfields[i], &emode); 1223 CeedChkBackend(ierr); 1224 // Basis action 1225 switch (emode) { 1226 case CEED_EVAL_NONE: 1227 code << " CeedScalar r_qq"<<i<<"[ncomp_out_"<<i<<"];\n"; 1228 break; // No action 1229 case CEED_EVAL_INTERP: 1230 code << " CeedScalar r_qq"<<i<<"[ncomp_out_"<<i<<"];\n"; 1231 break; 1232 case CEED_EVAL_GRAD: 1233 code << " CeedScalar r_qq"<<i<<"[ncomp_out_"<<i<<"*Dim];\n"; 1234 break; 1235 case CEED_EVAL_WEIGHT: 1236 break; // Should not occur 1237 case CEED_EVAL_DIV: 1238 break; // TODO: Not implemented 1239 case CEED_EVAL_CURL: 1240 break; // TODO: Not implemented 1241 } 1242 } 1243 } else { 1244 code << "\n // Note: No Collocated Gradient\n"; 1245 code << " // -- Input fields --\n"; 1246 for (CeedInt i = 0; i < numinputfields; i++) { 1247 code << " // ---- Input field "<<i<<" ----\n"; 1248 code << " CeedScalar* r_q"<<i<<" = r_t"<<i<<";\n"; 1249 } 1250 code << " // -- Output fields --\n"; 1251 for (CeedInt i = 0; i < numoutputfields; i++) { 1252 code << " // ---- Output field "<<i<<" ----\n"; 1253 code << " CeedScalar* r_qq"<<i<<" = r_tt"<<i<<";\n"; 1254 } 1255 } 1256 code << "\n // -- QFunction Inputs and outputs --\n"; 1257 code << " CeedScalar* in["<<numinputfields<<"];\n"; 1258 for (CeedInt i = 0; i < numinputfields; i++) { 1259 code << " // ---- Input field "<<i<<" ----\n"; 1260 code << " in["<<i<<"] = r_q"<<i<<";\n"; 1261 } 1262 code << " CeedScalar* out["<<numoutputfields<<"];\n"; 1263 for (CeedInt i = 0; i < numoutputfields; i++) { 1264 code << " // ---- Output field "<<i<<" ----\n"; 1265 code << " out["<<i<<"] = r_qq"<<i<<";\n"; 1266 } 1267 code << "\n // -- Apply QFunction --\n"; 1268 code << " "<<qFunctionName<<"(ctx, "; 1269 if (dim != 3 || useCollograd) { 1270 code << "1"; 1271 } else { 1272 code << "Q1d"; 1273 } 1274 code << ", in, out);\n"; 1275 if (useCollograd) { 1276 code << "\n // Note: Collocated Gradient\n"; 1277 code << " // -- Output fields --\n"; 1278 for (CeedInt i = 0; i < numoutputfields; i++) { 1279 code << " // ---- Output field "<<i<<" ----\n"; 1280 ierr = CeedQFunctionFieldGetEvalMode(qfoutputfields[i], &emode); 1281 CeedChkBackend(ierr); 1282 // Basis action 1283 code << " // EvalMode: "<<CeedEvalModes[emode]<<"\n"; 1284 switch (emode) { 1285 case CEED_EVAL_NONE: 1286 code << " for (CeedInt j = 0; j < ncomp_out_"<<i<<" ; ++j) {\n"; 1287 code << " r_tt"<<i<<"[q + j*Q1d] = r_qq"<<i<<"[j];\n"; 1288 code << " }\n"; 1289 break; // No action 1290 case CEED_EVAL_INTERP: 1291 code << " for (CeedInt j = 0; j < ncomp_out_"<<i<<" ; ++j) {\n"; 1292 code << " r_tt"<<i<<"[q + j*Q1d] = r_qq"<<i<<"[j];\n"; 1293 code << " }\n"; 1294 break; 1295 case CEED_EVAL_GRAD: 1296 code << " gradColloTranspose3d<ncomp_out_"<<i<<",Q1d>(data, q, r_qq"<<i<<", s_G_out_"<<i<<", r_tt"<<i<<");\n"; 1297 break; 1298 case CEED_EVAL_WEIGHT: 1299 break; // Should not occur 1300 case CEED_EVAL_DIV: 1301 break; // TODO: Not implemented 1302 case CEED_EVAL_CURL: 1303 break; // TODO: Not implemented 1304 } 1305 } 1306 code << " }\n"; 1307 } 1308 1309 // Output basis apply if needed 1310 // Generate the correct eval mode code for each output 1311 code << "\n // -- Output field basis action and restrictions --\n"; 1312 for (CeedInt i = 0; i < numoutputfields; i++) { 1313 code << " // ---- Output field "<<i<<" ----\n"; 1314 // Get elemsize, emode, ncomp 1315 ierr = CeedOperatorFieldGetElemRestriction(opoutputfields[i], &Erestrict); 1316 CeedChkBackend(ierr); 1317 ierr = CeedElemRestrictionGetElementSize(Erestrict, &elemsize); 1318 CeedChkBackend(ierr); 1319 ierr = CeedQFunctionFieldGetEvalMode(qfoutputfields[i], &emode); 1320 CeedChkBackend(ierr); 1321 ierr = CeedElemRestrictionGetNumComponents(Erestrict, &ncomp); 1322 CeedChkBackend(ierr); 1323 // Basis action 1324 code << " // EvalMode: "<<CeedEvalModes[emode]<<"\n"; 1325 switch (emode) { 1326 case CEED_EVAL_NONE: 1327 code << " CeedScalar* r_v"<<i<<" = r_tt"<<i<<";\n"; 1328 break; // No action 1329 case CEED_EVAL_INTERP: 1330 code << " CeedScalar r_v"<<i<<"[ncomp_out_"<<i<<"*P_out_"<<i<<"];\n"; 1331 code << " interpTranspose"<<dim<<"d<ncomp_out_"<<i<<",P_out_"<<i<<",Q1d>(data, r_tt"<<i<<", s_B_out_"<<i<<", r_v"<<i<<");\n"; 1332 break; 1333 case CEED_EVAL_GRAD: 1334 code << " CeedScalar r_v"<<i<<"[ncomp_out_"<<i<<"*P_out_"<<i<<"];\n"; 1335 if (useCollograd) { 1336 code << " interpTranspose"<<dim<<"d<ncomp_out_"<<i<<",P_out_"<<i<<",Q1d>(data, r_tt"<<i<<", s_B_out_"<<i<<", r_v"<<i<<");\n"; 1337 } else { 1338 code << " gradTranspose"<<dim<<"d<ncomp_out_"<<i<<",P_out_"<<i<<",Q1d>(data, r_tt"<<i<<", s_B_out_"<<i<<", s_G_out_"<<i<<", r_v"<<i<<");\n"; 1339 } 1340 break; 1341 // LCOV_EXCL_START 1342 case CEED_EVAL_WEIGHT: { 1343 Ceed ceed; 1344 ierr = CeedOperatorGetCeed(op, &ceed); CeedChkBackend(ierr); 1345 return CeedError(ceed, CEED_ERROR_BACKEND, 1346 "CEED_EVAL_WEIGHT cannot be an output evaluation mode"); 1347 break; // Should not occur 1348 } 1349 case CEED_EVAL_DIV: 1350 break; // TODO: Not implemented 1351 case CEED_EVAL_CURL: 1352 break; // TODO: Not implemented 1353 // LCOV_EXCL_STOP 1354 } 1355 // Restriction 1356 bool isStrided; 1357 ierr = CeedElemRestrictionIsStrided(Erestrict, &isStrided); CeedChkBackend(ierr); 1358 if (!isStrided) { 1359 ierr = CeedElemRestrictionGetLVectorSize(Erestrict, &lsize); 1360 CeedChkBackend(ierr); 1361 code << " const CeedInt lsize_out_"<<i<<" = "<<lsize<<";\n"; 1362 CeedInt compstride; 1363 ierr = CeedElemRestrictionGetCompStride(Erestrict, &compstride); CeedChkBackend(ierr); 1364 code << " // CompStride: "<<compstride<<"\n"; 1365 ierr = CeedElemRestrictionGetData(Erestrict, &restr_data); CeedChkBackend(ierr); 1366 data->indices.out[i] = restr_data->d_ind; 1367 code << " writeDofsOffset"<<dim<<"d<ncomp_out_"<<i<<", "<<compstride<<", P_out_"<<i<<">(data, lsize_out_"<<i<<", elem, indices.out["<<i<<"], r_v"<<i<<", d_v"<<i<<");\n"; 1368 } else { 1369 bool backendstrides; 1370 ierr = CeedElemRestrictionHasBackendStrides(Erestrict, &backendstrides); 1371 CeedChkBackend(ierr); 1372 CeedInt nelem; 1373 ierr = CeedElemRestrictionGetNumElements(Erestrict, &nelem); 1374 CeedChkBackend(ierr); 1375 CeedInt strides[3] = {1, elemsize*nelem, elemsize}; 1376 if (!backendstrides) { 1377 ierr = CeedElemRestrictionGetStrides(Erestrict, &strides); 1378 CeedChkBackend(ierr); 1379 } 1380 code << " // Strides: {"<<strides[0]<<", "<<strides[1]<<", "<<strides[2]<<"}\n"; 1381 code << " writeDofsStrided"<<dim<<"d<ncomp_out_"<<i<<",P_out_"<<i<<","<<strides[0]<<","<<strides[1]<<","<<strides[2]<<">(data, elem, r_v"<<i<<", d_v"<<i<<");\n"; 1382 } 1383 } 1384 1385 code << " }\n"; 1386 code << "}\n"; 1387 code << "// -----------------------------------------------------------------------------\n\n"; 1388 1389 // View kernel for debugging 1390 CeedDebug(ceed, code.str().c_str()); 1391 1392 ierr = CeedCompileCuda(ceed, code.str().c_str(), &data->module, 1, 1393 "T1d", CeedIntMax(Q1d, data->maxP1d)); 1394 CeedChkBackend(ierr); 1395 ierr = CeedGetKernelCuda(ceed, data->module, oper.c_str(), &data->op); 1396 CeedChkBackend(ierr); 1397 1398 ierr = CeedOperatorSetSetupDone(op); CeedChkBackend(ierr); 1399 return CEED_ERROR_SUCCESS; 1400 } 1401 //------------------------------------------------------------------------------ 1402