xref: /libCEED/backends/cuda-gen/ceed-cuda-gen-operator-build.cpp (revision 204bfdd766dd3467cc980d6ab2360b045ce7fa2c) !

// Copyright (c) 2017-2022, Lawrence Livermore National Security, LLC and other CEED contributors.
// All Rights Reserved. See the top-level LICENSE and NOTICE files for details.
//
// SPDX-License-Identifier: BSD-2-Clause
//
// This file is part of CEED:  http://github.com/ceed

#define CEED_DEBUG_COLOR 12

#include <ceed/ceed.h>
#include <ceed/backend.h>
#include <ceed/jit-tools.h>
#include <cuda_runtime.h>
#include <iostream>
#include <sstream>
#include "ceed-cuda-gen.h"
#include "../cuda/ceed-cuda-compile.h"
#include "../cuda-ref/ceed-cuda-ref.h"
#include "../cuda-shared/ceed-cuda-shared.h"

//------------------------------------------------------------------------------
// Build singe operator kernel
//------------------------------------------------------------------------------
extern "C" int CeedCudaGenOperatorBuild(CeedOperator op) {

  using std::ostringstream;
  using std::string;
  int ierr;
  bool is_setup_done;
  ierr = CeedOperatorIsSetupDone(op, &is_setup_done); CeedChkBackend(ierr);
  if (is_setup_done) return CEED_ERROR_SUCCESS;
  Ceed ceed;
  ierr = CeedOperatorGetCeed(op, &ceed); CeedChkBackend(ierr);
  CeedOperator_Cuda_gen *data;
  ierr = CeedOperatorGetData(op, &data); CeedChkBackend(ierr);
  CeedQFunction qf;
  CeedQFunction_Cuda_gen *qf_data;
  ierr = CeedOperatorGetQFunction(op, &qf); CeedChkBackend(ierr);
  ierr = CeedQFunctionGetData(qf, &qf_data); CeedChkBackend(ierr);
  CeedSize lsize;
  CeedInt Q, P_1d = 0, Q_1d = 0, elem_size, num_input_fields,
          num_output_fields, num_comp, dim = 1;
  ierr = CeedOperatorGetNumQuadraturePoints(op, &Q); CeedChkBackend(ierr);
  Q_1d = Q;
  CeedOperatorField *op_input_fields, *op_output_fields;
  ierr = CeedOperatorGetFields(op, &num_input_fields, &op_input_fields, &num_output_fields, &op_output_fields);
  CeedChkBackend(ierr);
  CeedQFunctionField *qf_input_fields, *qf_output_fields;
  ierr = CeedQFunctionGetFields(qf, NULL, &qf_input_fields, NULL, &qf_output_fields);
  CeedChkBackend(ierr);
  CeedEvalMode eval_mode;
  CeedBasis basis;
  CeedBasis_Cuda_shared *basis_data;
  CeedElemRestriction Erestrict;
  CeedElemRestriction_Cuda *restr_data;

  // TODO: put in a function?
  // Check for restriction only identity operator
  bool is_identity_qf;
  ierr = CeedQFunctionIsIdentity(qf, &is_identity_qf); CeedChkBackend(ierr);
  if (is_identity_qf) {
    CeedEvalMode eval_mode_in, eval_mode_out;
    ierr = CeedQFunctionFieldGetEvalMode(qf_input_fields[0], &eval_mode_in);  CeedChkBackend(ierr);
    ierr = CeedQFunctionFieldGetEvalMode(qf_output_fields[0], &eval_mode_out);  CeedChkBackend(ierr);
    if (eval_mode_in == CEED_EVAL_NONE && eval_mode_out == CEED_EVAL_NONE)
      // LCOV_EXCL_START
      return CeedError(ceed, CEED_ERROR_BACKEND,
                       "Backend does not implement restriction only identity operators");
    // LCOV_EXCL_STOP
  }

  ostringstream code;

  // TODO: put in a function?
  // Add atomicAdd function for old NVidia architectures
  struct cudaDeviceProp prop;
  Ceed_Cuda *ceed_data;
  ierr = CeedGetData(ceed, &ceed_data); CeedChkBackend(ierr); CeedChkBackend(ierr);
  ierr = cudaGetDeviceProperties(&prop, ceed_data->device_id); CeedChkBackend(ierr);
  if ((prop.major < 6) && (CEED_SCALAR_TYPE != CEED_SCALAR_FP32)){
    char *atomic_add_path, *atomic_add_source;
    ierr = CeedGetJitAbsolutePath(ceed,
                                  "ceed/jit-source/cuda/cuda-atomic-add-fallback.h",
                                  &atomic_add_path); CeedChkBackend(ierr);
    CeedDebug256(ceed, 2, "----- Loading Atomic Add Source -----\n");
    ierr = CeedLoadSourceToBuffer(ceed, atomic_add_path, &atomic_add_source);
    CeedChkBackend(ierr);
    code << atomic_add_source;
    ierr = CeedFree(&atomic_add_path); CeedChkBackend(ierr);
    ierr = CeedFree(&atomic_add_source); CeedChkBackend(ierr);
  }

  // Load basis source files
  // TODO: generalize to accept different device functions?
  {
    char *tensor_basis_kernel_path, *tensor_basis_kernel_source;
    ierr = CeedGetJitAbsolutePath(ceed,
                                  "ceed/jit-source/cuda/cuda-shared-basis-tensor-templates.h",
                                  &tensor_basis_kernel_path); CeedChkBackend(ierr);
    CeedDebug256(ceed, 2, "----- Loading Tensor Basis Kernel Source -----\n");
    ierr = CeedLoadSourceToBuffer(ceed, tensor_basis_kernel_path, &tensor_basis_kernel_source);
    CeedChkBackend(ierr);
    code << tensor_basis_kernel_source;
    ierr = CeedFree(&tensor_basis_kernel_path); CeedChkBackend(ierr);
    ierr = CeedFree(&tensor_basis_kernel_source); CeedChkBackend(ierr);
  }
  {
    char *cuda_gen_template_path, *cuda_gen_template_source;
    ierr = CeedGetJitAbsolutePath(ceed,
                                  "ceed/jit-source/cuda/cuda-gen-templates.h",
                                  &cuda_gen_template_path); CeedChkBackend(ierr);
    CeedDebug256(ceed, 2, "----- Loading Cuda-Gen Template Source -----\n");
    ierr = CeedLoadSourceToBuffer(ceed, cuda_gen_template_path, &cuda_gen_template_source);
    CeedChkBackend(ierr);
    code << cuda_gen_template_source;
    ierr = CeedFree(&cuda_gen_template_path); CeedChkBackend(ierr);
    ierr = CeedFree(&cuda_gen_template_source); CeedChkBackend(ierr);
  }

  // Get QFunction source and name
  string q_function_source(qf_data->q_function_source);
  string q_function_name(qf_data->q_function_name);
  string operator_name;
  operator_name = "CeedKernelCudaGenOperator_" + q_function_name;

  // Find dim, P_1d, Q_1d
  data->max_P_1d = 0;
  for (CeedInt i = 0; i < num_input_fields; i++) {
    ierr = CeedOperatorFieldGetBasis(op_input_fields[i], &basis); CeedChkBackend(ierr);
    if (basis != CEED_BASIS_COLLOCATED) {
      ierr = CeedBasisGetData(basis, &basis_data); CeedChkBackend(ierr);
      ierr = CeedQFunctionFieldGetEvalMode(qf_input_fields[i], &eval_mode);
      CeedChkBackend(ierr);

      // Collect dim, P_1d, and Q_1d
      ierr = CeedBasisGetDimension(basis, &dim); CeedChkBackend(ierr);
      bool isTensor;
      ierr = CeedBasisIsTensor(basis, &isTensor); CeedChkBackend(ierr);
      if (isTensor) {
        ierr = CeedBasisGetNumQuadraturePoints1D(basis, &Q_1d); CeedChkBackend(ierr);
        ierr = CeedBasisGetNumNodes1D(basis, &P_1d); CeedChkBackend(ierr);
        if (P_1d > data->max_P_1d) data->max_P_1d = P_1d;
      } else {
        // LCOV_EXCL_START
        return CeedError(ceed, CEED_ERROR_BACKEND, "Backend does not implement operators with non-tensor basis");
        // LCOV_EXCL_STOP
      }
    }
  }
  // Check output bases for Q_1d, dim as well
  //   The only input basis might be CEED_BASIS_COLLOCATED
  for (CeedInt i = 0; i < num_output_fields; i++) {
    ierr = CeedOperatorFieldGetBasis(op_output_fields[i], &basis); CeedChkBackend(ierr);

    if (basis != CEED_BASIS_COLLOCATED) {
      ierr = CeedBasisGetData(basis, &basis_data); CeedChkBackend(ierr);
      ierr = CeedQFunctionFieldGetEvalMode(qf_output_fields[i], &eval_mode);
      CeedChkBackend(ierr);

      // Collect Q_1d
      ierr = CeedBasisGetDimension(basis, &dim); CeedChkBackend(ierr);
      bool isTensor;
      ierr = CeedBasisIsTensor(basis, &isTensor); CeedChkBackend(ierr);
      if (isTensor) {
        ierr = CeedBasisGetNumQuadraturePoints1D(basis, &Q_1d); CeedChkBackend(ierr);
      } else {
        // LCOV_EXCL_START
        return CeedError(ceed, CEED_ERROR_BACKEND, "Backend does not implement operators with non-tensor basis");
        // LCOV_EXCL_STOP
      }
    }
  }
  data->dim = dim;
  data->Q_1d = Q_1d;

  // Only use 3D collocated gradient parallelization strategy when gradient is computed
  // TODO: put in a function?
  bool use_collograd_parallelization = false;
  if (dim == 3) {
    bool was_grad_found = false;
    for (CeedInt i = 0; i < num_input_fields; i++) {
      ierr = CeedQFunctionFieldGetEvalMode(qf_input_fields[i], &eval_mode);
      if (eval_mode == CEED_EVAL_GRAD) {
        ierr = CeedOperatorFieldGetBasis(op_input_fields[i], &basis); CeedChkBackend(ierr);
        ierr = CeedBasisGetData(basis, &basis_data); CeedChkBackend(ierr);
        use_collograd_parallelization = !!basis_data->d_collo_grad_1d && (was_grad_found ? use_collograd_parallelization : true);
        was_grad_found = true;
      }
    }
    for (CeedInt i = 0; i < num_output_fields; i++) {
      ierr = CeedQFunctionFieldGetEvalMode(qf_output_fields[i], &eval_mode);
      if (eval_mode == CEED_EVAL_GRAD) {
        ierr = CeedOperatorFieldGetBasis(op_output_fields[i], &basis); CeedChkBackend(ierr);
        ierr = CeedBasisGetData(basis, &basis_data); CeedChkBackend(ierr);
        use_collograd_parallelization = !!basis_data->d_collo_grad_1d && (was_grad_found ? use_collograd_parallelization : true);
        was_grad_found = true;
      }
    }
  }

  // Define CEED_Q_VLA
  code << "\n#undef CEED_Q_VLA\n";
  if (dim != 3 || use_collograd_parallelization) {
    code << "#define CEED_Q_VLA 1\n\n";
  } else {
    code << "#define CEED_Q_VLA "<<Q_1d<<"\n\n";
  }

  code << q_function_source;

  // Setup
  code << "\n// -----------------------------------------------------------------------------\n";
  code << "\nextern \"C\" __global__ void "<<operator_name<<"(CeedInt num_elem, void* ctx, FieldsInt_Cuda indices, Fields_Cuda fields, Fields_Cuda B, Fields_Cuda G, CeedScalar* W) {\n";
  for (CeedInt i = 0; i < num_input_fields; i++) {
    ierr = CeedQFunctionFieldGetEvalMode(qf_input_fields[i], &eval_mode);
    CeedChkBackend(ierr);
    if (eval_mode != CEED_EVAL_WEIGHT) { // Skip CEED_EVAL_WEIGHT
      code << "  const CeedScalar* d_u_" <<i<<" = fields.inputs["<<i<<"];\n";
    }
  }

  for (CeedInt i = 0; i < num_output_fields; i++) {
    code << "  CeedScalar* d_v_"<<i<<" = fields.outputs["<<i<<"];\n";
  }

  code << "  const CeedInt dim = "<<dim<<";\n";
  code << "  const CeedInt Q_1d = "<<Q_1d<<";\n";

  code << "  extern __shared__ CeedScalar slice[];\n";
  // TODO put in a function? InitSharedData_Cuda?
  code << "  SharedData_Cuda data;\n";
  code << "  data.t_id_x = threadIdx.x;\n";
  code << "  data.t_id_y = threadIdx.y;\n";
  code << "  data.t_id_z = threadIdx.z;\n";
  code << "  data.t_id  = threadIdx.x + threadIdx.y*blockDim.x + threadIdx.z*blockDim.y*blockDim.x;\n";
  code << "  data.slice = slice+data.t_id_z*T_1D"<<(dim>1?"*T_1D":"")<<";\n";

  code << "\n  // -- Input field constants and basis data --\n";
  // TODO: Put in a function?
  //Initialize constants, and matrices B and G
  for (CeedInt i = 0; i < num_input_fields; i++) {
    code << "  // ---- Input field "<<i<<" ----\n";
    // Get elem_size, eval_mode, num_comp
    ierr = CeedOperatorFieldGetElemRestriction(op_input_fields[i], &Erestrict);
    CeedChkBackend(ierr);
    ierr = CeedElemRestrictionGetElementSize(Erestrict, &elem_size);
    CeedChkBackend(ierr);
    ierr = CeedQFunctionFieldGetEvalMode(qf_input_fields[i], &eval_mode);
    CeedChkBackend(ierr);
    ierr = CeedElemRestrictionGetNumComponents(Erestrict, &num_comp);
    CeedChkBackend(ierr);

    // Set field constants
    if (eval_mode != CEED_EVAL_WEIGHT) {
      ierr = CeedOperatorFieldGetBasis(op_input_fields[i], &basis); CeedChkBackend(ierr);
      if (basis != CEED_BASIS_COLLOCATED) {
        ierr = CeedBasisGetNumNodes1D(basis, &P_1d); CeedChkBackend(ierr);
        code << "  const CeedInt P_in_"<<i<<" = "<<P_1d<<";\n";
      } else {
        code << "  const CeedInt P_in_"<<i<<" = "<<Q_1d<<";\n";
      }
      code << "  const CeedInt num_comp_in_"<<i<<" = "<<num_comp<<";\n";
    }

    // Load basis data
    code << "  // EvalMode: "<<CeedEvalModes[eval_mode]<<"\n";
    switch (eval_mode) {
    case CEED_EVAL_NONE:
      break;
    case CEED_EVAL_INTERP:
      ierr = CeedBasisGetData(basis, &basis_data); CeedChkBackend(ierr);
      data->B.inputs[i] = basis_data->d_interp_1d;
      code << "  __shared__ CeedScalar s_B_in_"<<i<<"["<<P_1d*Q_1d<<"];\n";
      code << "  loadMatrix<P_in_"<<i<<",Q_1d>(data, B.inputs["<<i<<"], s_B_in_"<<i<<");\n";
      break;
    case CEED_EVAL_GRAD:
      ierr = CeedBasisGetData(basis, &basis_data); CeedChkBackend(ierr);
      data->B.inputs[i] = basis_data->d_interp_1d;
      code << "  __shared__ CeedScalar s_B_in_"<<i<<"["<<P_1d*Q_1d<<"];\n";
      code << "  loadMatrix<P_in_"<<i<<",Q_1d>(data, B.inputs["<<i<<"], s_B_in_"<<i<<");\n";
      if (use_collograd_parallelization) {
        data->G.inputs[i] = basis_data->d_collo_grad_1d;
        code << "  __shared__ CeedScalar s_G_in_"<<i<<"["<<Q_1d*Q_1d<<"];\n";
        code << "  loadMatrix<Q_1d,Q_1d>(data, G.inputs["<<i<<"], s_G_in_"<<i<<");\n";
      } else {
        bool has_collo_grad = !!basis_data->d_collo_grad_1d;
        data->G.inputs[i] = has_collo_grad ? basis_data->d_collo_grad_1d : basis_data->d_grad_1d;
        code << "  __shared__ CeedScalar s_G_in_"<<i<<"["<<Q_1d*(has_collo_grad?Q_1d:P_1d)<<"];\n";
        code << "  loadMatrix<"<<(has_collo_grad?"Q_1d":("P_in_"+std::to_string(i)))<<",Q_1d>(data, G.inputs["<<i<<"], s_G_in_"<<i<<");\n";
      }
      break;
    case CEED_EVAL_WEIGHT:
      break; // No action
    case CEED_EVAL_DIV:
      break; // TODO: Not implemented
    case CEED_EVAL_CURL:
      break; // TODO: Not implemented
    }
  }

  code << "\n  // -- Output field constants and basis data --\n";
  for (CeedInt i = 0; i < num_output_fields; i++) {
    code << "  // ---- Output field "<<i<<" ----\n";
    // Get elem_size, eval_mode, num_comp
    ierr = CeedOperatorFieldGetElemRestriction(op_output_fields[i], &Erestrict);
    CeedChkBackend(ierr);
    ierr = CeedElemRestrictionGetElementSize(Erestrict, &elem_size);
    CeedChkBackend(ierr);
    ierr = CeedQFunctionFieldGetEvalMode(qf_output_fields[i], &eval_mode);
    CeedChkBackend(ierr);
    ierr = CeedElemRestrictionGetNumComponents(Erestrict, &num_comp);
    CeedChkBackend(ierr);

    // Set field constants
    ierr = CeedOperatorFieldGetBasis(op_output_fields[i], &basis); CeedChkBackend(ierr);
    if (basis != CEED_BASIS_COLLOCATED) {
      ierr = CeedBasisGetNumNodes1D(basis, &P_1d); CeedChkBackend(ierr);
      code << "  const CeedInt P_out_"<<i<<" = "<<P_1d<<";\n";
    } else {
      code << "  const CeedInt P_out_"<<i<<" = "<<Q_1d<<";\n";
    }
    code << "  const CeedInt num_comp_out_"<<i<<" = "<<num_comp<<";\n";

    // Load basis data
    code << "  // EvalMode: "<<CeedEvalModes[eval_mode]<<"\n";
    switch (eval_mode) {
    case CEED_EVAL_NONE:
      break; // No action
    case CEED_EVAL_INTERP:
      ierr = CeedBasisGetData(basis, &basis_data); CeedChkBackend(ierr);
      data->B.outputs[i] = basis_data->d_interp_1d;
      code << "  __shared__ CeedScalar s_B_out_"<<i<<"["<<P_1d*Q_1d<<"];\n";
      code << "  loadMatrix<P_out_"<<i<<",Q_1d>(data, B.outputs["<<i<<"], s_B_out_"<<i<<");\n";
      break;
    case CEED_EVAL_GRAD:
      ierr = CeedBasisGetData(basis, &basis_data); CeedChkBackend(ierr);
      data->B.outputs[i] = basis_data->d_interp_1d;
      code << "  __shared__ CeedScalar s_B_out_"<<i<<"["<<P_1d*Q_1d<<"];\n";
      code << "  loadMatrix<P_out_"<<i<<",Q_1d>(data, B.outputs["<<i<<"], s_B_out_"<<i<<");\n";
      if (use_collograd_parallelization) {
        data->G.outputs[i] = basis_data->d_collo_grad_1d;
        code << "  __shared__ CeedScalar s_G_out_"<<i<<"["<<Q_1d*Q_1d<<"];\n";
        code << "  loadMatrix<Q_1d,Q_1d>(data, G.outputs["<<i<<"], s_G_out_"<<i<<");\n";
      } else {
        bool has_collo_grad = !!basis_data->d_collo_grad_1d;
        data->G.outputs[i] = has_collo_grad ? basis_data->d_collo_grad_1d : basis_data->d_grad_1d;
        code << "  __shared__ CeedScalar s_G_out_"<<i<<"["<<Q_1d*(has_collo_grad?Q_1d:P_1d)<<"];\n";
        code << "  loadMatrix<"<<(has_collo_grad?"Q_1d":("P_out_"+std::to_string(i)))<<",Q_1d>(data, G.outputs["<<i<<"], s_G_out_"<<i<<");\n";
      }
      break;
    // LCOV_EXCL_START
    case CEED_EVAL_WEIGHT: {
      Ceed ceed;
      ierr = CeedOperatorGetCeed(op, &ceed); CeedChkBackend(ierr);
      return CeedError(ceed, CEED_ERROR_BACKEND,
                       "CEED_EVAL_WEIGHT cannot be an output evaluation mode");
      break; // Should not occur
    }
    case CEED_EVAL_DIV:
      break; // TODO: Not implemented
    case CEED_EVAL_CURL:
      break; // TODO: Not implemented
      // LCOV_EXCL_STOP
    }
  }
  code << "\n  // -- Element loop --\n";
  code << "  __syncthreads();\n";
  code << "  for (CeedInt elem = blockIdx.x*blockDim.z + threadIdx.z; elem < num_elem; elem += gridDim.x*blockDim.z) {\n";
  // Input basis apply if needed
  // Generate the correct eval mode code for each input
  code << "    // -- Input field restrictions and basis actions --\n";
  for (CeedInt i = 0; i < num_input_fields; i++) {
    code << "    // ---- Input field "<<i<<" ----\n";
    // Get elem_size, eval_mode, num_comp
    ierr = CeedOperatorFieldGetElemRestriction(op_input_fields[i], &Erestrict);
    CeedChkBackend(ierr);
    ierr = CeedElemRestrictionGetElementSize(Erestrict, &elem_size);
    CeedChkBackend(ierr);
    ierr = CeedQFunctionFieldGetEvalMode(qf_input_fields[i], &eval_mode);
    CeedChkBackend(ierr);
    ierr = CeedElemRestrictionGetNumComponents(Erestrict, &num_comp);
    CeedChkBackend(ierr);

    // TODO: put in a function?
    // Restriction
    if (eval_mode != CEED_EVAL_WEIGHT &&
        !((eval_mode == CEED_EVAL_NONE) && use_collograd_parallelization)) {
      code << "    CeedScalar r_u_"<<i<<"[num_comp_in_"<<i<<"*P_in_"<<i<<"];\n";

      bool is_strided;
      ierr = CeedElemRestrictionIsStrided(Erestrict, &is_strided); CeedChkBackend(ierr);
      if (!is_strided) {
        ierr = CeedElemRestrictionGetLVectorSize(Erestrict, &lsize);
        CeedChkBackend(ierr);
        code << "    const CeedInt lsize_in_"<<i<<" = "<<lsize<<";\n";
        CeedInt comp_stride;
        ierr = CeedElemRestrictionGetCompStride(Erestrict, &comp_stride); CeedChkBackend(ierr);
        code << "    // CompStride: "<<comp_stride<<"\n";
        ierr = CeedElemRestrictionGetData(Erestrict, &restr_data); CeedChkBackend(ierr);
        data->indices.inputs[i] = restr_data->d_ind;
        code << "    readDofsOffset"<<dim<<"d<num_comp_in_"<<i<<", "<<comp_stride<<", P_in_"<<i<<">(data, lsize_in_"<<i<<", elem, indices.inputs["<<i<<"], d_u_"<<i<<", r_u_"<<i<<");\n";
      } else {
        bool backendstrides;
        ierr = CeedElemRestrictionHasBackendStrides(Erestrict, &backendstrides);
        CeedChkBackend(ierr);
        CeedInt num_elem;
        ierr = CeedElemRestrictionGetNumElements(Erestrict, &num_elem);
        CeedChkBackend(ierr);
        CeedInt strides[3] = {1, elem_size*num_elem, elem_size};
        if (!backendstrides) {
          ierr = CeedElemRestrictionGetStrides(Erestrict, &strides);
          CeedChkBackend(ierr);
        }
        code << "    // Strides: {"<<strides[0]<<", "<<strides[1]<<", "<<strides[2]<<"}\n";
        code << "    readDofsStrided"<<dim<<"d<num_comp_in_"<<i<<",P_in_"<<i<<","<<strides[0]<<","<<strides[1]<<","<<strides[2]<<">(data, elem, d_u_"<<i<<", r_u_"<<i<<");\n";
      }
    }

    // TODO: put in a function?
    // Basis action
    code << "    // EvalMode: "<<CeedEvalModes[eval_mode]<<"\n";
    switch (eval_mode) {
    case CEED_EVAL_NONE:
      if (!use_collograd_parallelization) {
        code << "    CeedScalar* r_t_"<<i<<" = r_u_"<<i<<";\n";
      }
      break;
    case CEED_EVAL_INTERP:
      code << "    CeedScalar r_t_"<<i<<"[num_comp_in_"<<i<<"*Q_1d];\n";
      code << "    Interp"<<(dim>1?"Tensor":"")<<dim<<"d<num_comp_in_"<<i<<",P_in_"<<i<<",Q_1d>(data, r_u_"<<i<<", s_B_in_"<<i<<", r_t_"<<i<<");\n";
      break;
    case CEED_EVAL_GRAD:
      if (use_collograd_parallelization) {
        code << "    CeedScalar r_t_"<<i<<"[num_comp_in_"<<i<<"*Q_1d];\n";
        code << "    Interp"<<(dim>1?"Tensor":"")<<dim<<"d<num_comp_in_"<<i<<",P_in_"<<i<<",Q_1d>(data, r_u_"<<i<<", s_B_in_"<<i<<", r_t_"<<i<<");\n";
      } else {
        CeedInt P_1d;
        ierr = CeedOperatorFieldGetBasis(op_input_fields[i], &basis); CeedChkBackend(ierr);
        ierr = CeedBasisGetNumNodes1D(basis, &P_1d); CeedChkBackend(ierr);
        code << "    CeedScalar r_t_"<<i<<"[num_comp_in_"<<i<<"*dim*Q_1d];\n";
        code << "    Grad"<<(dim>1?"Tensor":"")<<(dim==3&&Q_1d>=P_1d?"Collocated":"")<<dim<<"d<num_comp_in_"<<i<<",P_in_"<<i<<",Q_1d>(data, r_u_"<<i<<", s_B_in_"<<i<<", s_G_in_"<<i<<", r_t_"<<i<<");\n";
      }
      break;
    case CEED_EVAL_WEIGHT:
      code << "    CeedScalar r_t_"<<i<<"[Q_1d];\n";
      ierr = CeedOperatorFieldGetBasis(op_input_fields[i], &basis); CeedChkBackend(ierr);
      ierr = CeedBasisGetData(basis, &basis_data); CeedChkBackend(ierr);
      data->W = basis_data->d_q_weight_1d;
      code << "    Weight"<<(dim>1?"Tensor":"")<<dim<<"d<Q_1d>(data, W, r_t_"<<i<<");\n";
      break; // No action
    case CEED_EVAL_DIV:
      break; // TODO: Not implemented
    case CEED_EVAL_CURL:
      break; // TODO: Not implemented
    }
  }

  // TODO: put in a function + separate colograd logic
  // Q function
  code << "\n    // -- Output field setup --\n";
  for (CeedInt i = 0; i < num_output_fields; i++) {
      code << "\n    // ---- Output field "<<i<<" ----\n";
    ierr = CeedQFunctionFieldGetEvalMode(qf_output_fields[i], &eval_mode);
    CeedChkBackend(ierr);
    if (eval_mode==CEED_EVAL_GRAD)
    {
      if (use_collograd_parallelization) {
        //Accumulator for gradient slices
        code << "    CeedScalar r_tt_"<<i<<"[num_comp_out_"<<i<<"*Q_1d];\n";
        code << "    for (CeedInt i = 0; i < num_comp_out_"<<i<<"; i++) {\n";
        code << "      for (CeedInt j = 0; j < Q_1d; ++j) {\n";
        code << "        r_tt_"<<i<<"[j + i*Q_1d] = 0.0;\n";
        code << "      }\n";
        code << "    }\n";
      } else {
        code << "    CeedScalar r_tt_"<<i<<"[num_comp_out_"<<i<<"*dim*Q_1d];\n";
      }
    }
    if (eval_mode==CEED_EVAL_NONE || eval_mode==CEED_EVAL_INTERP)
    {
      code << "    CeedScalar r_tt_"<<i<<"[num_comp_out_"<<i<<"*Q_1d];\n";
    }
  }
  // We treat quadrature points per slice in 3d to save registers
  if (use_collograd_parallelization) {
    code << "\n    // Note: Using planes of 3D elements\n";
    code << "#pragma unroll\n";
    code << "    for (CeedInt q = 0; q < Q_1d; q++) {\n";
    code << "      // -- Input fields --\n";
    for (CeedInt i = 0; i < num_input_fields; i++) {
      code << "      // ---- Input field "<<i<<" ----\n";
      // Get elem_size, eval_mode, num_comp
      ierr = CeedQFunctionFieldGetEvalMode(qf_input_fields[i], &eval_mode);
      CeedChkBackend(ierr);
      // Basis action
      code << "      // EvalMode: "<<CeedEvalModes[eval_mode]<<"\n";
      switch (eval_mode) {
      case CEED_EVAL_NONE:
        code << "      CeedScalar r_q_"<<i<<"[num_comp_in_"<<i<<"];\n";

        bool is_strided;
        ierr = CeedOperatorFieldGetElemRestriction(op_input_fields[i], &Erestrict); CeedChkBackend(ierr);
        ierr = CeedElemRestrictionIsStrided(Erestrict, &is_strided); CeedChkBackend(ierr);
        if (!is_strided) {
          ierr = CeedElemRestrictionGetLVectorSize(Erestrict, &lsize);
          CeedChkBackend(ierr);
          code << "      const CeedInt lsize_in_"<<i<<" = "<<lsize<<";\n";
          CeedInt comp_stride;
          ierr = CeedElemRestrictionGetCompStride(Erestrict, &comp_stride); CeedChkBackend(ierr);
          code << "      // CompStride: "<<comp_stride<<"\n";
          ierr = CeedElemRestrictionGetData(Erestrict, &restr_data); CeedChkBackend(ierr);
          data->indices.inputs[i] = restr_data->d_ind;
          code << "      readSliceQuadsOffset"<<"3d<num_comp_in_"<<i<<", "<<comp_stride<<", Q_1d>(data, lsize_in_"<<i<<", elem, q, indices.inputs["<<i<<"], d_u_"<<i<<", r_q_"<<i<<");\n";
        } else {
          ierr = CeedElemRestrictionGetElementSize(Erestrict, &elem_size); CeedChkBackend(ierr);
          bool backendstrides;
          ierr = CeedElemRestrictionHasBackendStrides(Erestrict, &backendstrides);
          CeedChkBackend(ierr);
          CeedInt num_elem;
          ierr = CeedElemRestrictionGetNumElements(Erestrict, &num_elem);
          CeedChkBackend(ierr);
          CeedInt strides[3] = {1, elem_size*num_elem, elem_size};
          if (!backendstrides) {
            ierr = CeedElemRestrictionGetStrides(Erestrict, &strides);
            CeedChkBackend(ierr);
          }
          code << "      // Strides: {"<<strides[0]<<", "<<strides[1]<<", "<<strides[2]<<"}\n";
          code << "      readSliceQuadsStrided"<<"3d<num_comp_in_"<<i<<",Q_1d"","<<strides[0]<<","<<strides[1]<<","<<strides[2]<<">(data, elem, q, d_u_"<<i<<", r_q_"<<i<<");\n";
        }
        break;
      case CEED_EVAL_INTERP:
        code << "      CeedScalar r_q_"<<i<<"[num_comp_in_"<<i<<"];\n";
        code << "      for (CeedInt j = 0; j < num_comp_in_"<<i<<" ; ++j) {\n";
        code << "        r_q_"<<i<<"[j] = r_t_"<<i<<"[q + j*Q_1d];\n";
        code << "      }\n";
        break;
      case CEED_EVAL_GRAD:
        code << "      CeedScalar r_q_"<<i<<"[num_comp_in_"<<i<<"*dim];\n";
        code << "      gradCollo3d<num_comp_in_"<<i<<",Q_1d>(data, q, r_t_"<<i<<", s_G_in_"<<i<<", r_q_"<<i<<");\n";
        break;
      case CEED_EVAL_WEIGHT:
        code << "      CeedScalar r_q_"<<i<<"[1];\n";
        code << "      r_q_"<<i<<"[0] = r_t_"<<i<<"[q];\n";
        break; // No action
      case CEED_EVAL_DIV:
        break; // TODO: Not implemented
      case CEED_EVAL_CURL:
        break; // TODO: Not implemented
      }
    }
    code << "\n      // -- Output fields --\n";
    for (CeedInt i = 0; i < num_output_fields; i++) {
      code << "      // ---- Output field "<<i<<" ----\n";
      ierr = CeedQFunctionFieldGetEvalMode(qf_output_fields[i], &eval_mode);
      CeedChkBackend(ierr);
      // Basis action
      switch (eval_mode) {
      case CEED_EVAL_NONE:
        code << "      CeedScalar r_qq_"<<i<<"[num_comp_out_"<<i<<"];\n";
        break; // No action
      case CEED_EVAL_INTERP:
        code << "      CeedScalar r_qq_"<<i<<"[num_comp_out_"<<i<<"];\n";
        break;
      case CEED_EVAL_GRAD:
        code << "      CeedScalar r_qq_"<<i<<"[num_comp_out_"<<i<<"*dim];\n";
        break;
      case CEED_EVAL_WEIGHT:
        break; // Should not occur
      case CEED_EVAL_DIV:
        break; // TODO: Not implemented
      case CEED_EVAL_CURL:
        break; // TODO: Not implemented
      }
    }
  } else {
    code << "\n      // Note: Using full elements\n";
    code << "      // -- Input fields --\n";
    for (CeedInt i = 0; i < num_input_fields; i++) {
      code << "      // ---- Input field "<<i<<" ----\n";
      code << "      CeedScalar* r_q_"<<i<<" = r_t_"<<i<<";\n";
    }
    code << "      // -- Output fields --\n";
    for (CeedInt i = 0; i < num_output_fields; i++) {
      code << "      // ---- Output field "<<i<<" ----\n";
      code << "      CeedScalar* r_qq_"<<i<<" = r_tt_"<<i<<";\n";
    }
  }
  code << "\n      // -- QFunction Inputs and outputs --\n";
  code << "      CeedScalar* in["<<num_input_fields<<"];\n";
  for (CeedInt i = 0; i < num_input_fields; i++) {
    code << "      // ---- Input field "<<i<<" ----\n";
    code << "      in["<<i<<"] = r_q_"<<i<<";\n";
  }
  code << "      CeedScalar* out["<<num_output_fields<<"];\n";
  for (CeedInt i = 0; i < num_output_fields; i++) {
    code << "      // ---- Output field "<<i<<" ----\n";
    code << "      out["<<i<<"] = r_qq_"<<i<<";\n";
  }
  code << "\n      // -- Apply QFunction --\n";
  code << "      "<<q_function_name<<"(ctx, ";
  if (dim != 3 || use_collograd_parallelization) {
    code << "1";
  } else {
    code << "Q_1d";
  }
  code << ", in, out);\n";
  if (use_collograd_parallelization) {
    code << "      // -- Output fields --\n";
    for (CeedInt i = 0; i < num_output_fields; i++) {
      code << "      // ---- Output field "<<i<<" ----\n";
      ierr = CeedQFunctionFieldGetEvalMode(qf_output_fields[i], &eval_mode);
      CeedChkBackend(ierr);
      // Basis action
      code << "      // EvalMode: "<<CeedEvalModes[eval_mode]<<"\n";
      switch (eval_mode) {
      case CEED_EVAL_NONE:
        code << "      for (CeedInt j = 0; j < num_comp_out_"<<i<<" ; ++j) {\n";
        code << "        r_tt_"<<i<<"[q + j*Q_1d] = r_qq_"<<i<<"[j];\n";
        code << "      }\n";
        break; // No action
      case CEED_EVAL_INTERP:
        code << "      for (CeedInt j = 0; j < num_comp_out_"<<i<<" ; ++j) {\n";
        code << "        r_tt_"<<i<<"[q + j*Q_1d] = r_qq_"<<i<<"[j];\n";
        code << "      }\n";
        break;
      case CEED_EVAL_GRAD:
        code << "      gradColloTranspose3d<num_comp_out_"<<i<<",Q_1d>(data, q, r_qq_"<<i<<", s_G_out_"<<i<<", r_tt_"<<i<<");\n";
        break;
      case CEED_EVAL_WEIGHT:
        break; // Should not occur
      case CEED_EVAL_DIV:
        break; // TODO: Not implemented
      case CEED_EVAL_CURL:
        break; // TODO: Not implemented
      }
    }
    code << "    }\n";
  }

  // Output basis apply if needed
  // Generate the correct eval mode code for each output
  code << "\n    // -- Output field basis action and restrictions --\n";
  for (CeedInt i = 0; i < num_output_fields; i++) {
    code << "    // ---- Output field "<<i<<" ----\n";
    // Get elem_size, eval_mode, num_comp
    ierr = CeedOperatorFieldGetElemRestriction(op_output_fields[i], &Erestrict);
    CeedChkBackend(ierr);
    ierr = CeedElemRestrictionGetElementSize(Erestrict, &elem_size);
    CeedChkBackend(ierr);
    ierr = CeedQFunctionFieldGetEvalMode(qf_output_fields[i], &eval_mode);
    CeedChkBackend(ierr);
    ierr = CeedElemRestrictionGetNumComponents(Erestrict, &num_comp);
    CeedChkBackend(ierr);
    // TODO put in a function
    // Basis action
    code << "    // EvalMode: "<<CeedEvalModes[eval_mode]<<"\n";
    switch (eval_mode) {
    case CEED_EVAL_NONE:
      code << "    CeedScalar* r_v_"<<i<<" = r_tt_"<<i<<";\n";
      break; // No action
    case CEED_EVAL_INTERP:
      code << "    CeedScalar r_v_"<<i<<"[num_comp_out_"<<i<<"*P_out_"<<i<<"];\n";
      code << "    InterpTranspose"<<(dim>1?"Tensor":"")<<dim<<"d<num_comp_out_"<<i<<",P_out_"<<i<<",Q_1d>(data, r_tt_"<<i<<", s_B_out_"<<i<<", r_v_"<<i<<");\n";
      break;
    case CEED_EVAL_GRAD:
      code << "    CeedScalar r_v_"<<i<<"[num_comp_out_"<<i<<"*P_out_"<<i<<"];\n";
      if (use_collograd_parallelization) {
        code << "    InterpTranspose"<<(dim>1?"Tensor":"")<<dim<<"d<num_comp_out_"<<i<<",P_out_"<<i<<",Q_1d>(data, r_tt_"<<i<<", s_B_out_"<<i<<", r_v_"<<i<<");\n";
      } else {
        CeedInt P_1d;
        ierr = CeedOperatorFieldGetBasis(op_output_fields[i], &basis); CeedChkBackend(ierr);
        ierr = CeedBasisGetNumNodes1D(basis, &P_1d); CeedChkBackend(ierr);
        code << "    GradTranspose"<<(dim>1?"Tensor":"")<<(dim==3&&Q_1d>=P_1d?"Collocated":"")<<dim<<"d<num_comp_out_"<<i<<",P_out_"<<i<<",Q_1d>(data, r_tt_"<<i<<", s_B_out_"<<i<<", s_G_out_"<<i<<", r_v_"<<i<<");\n";
      }
      break;
    // LCOV_EXCL_START
    case CEED_EVAL_WEIGHT: {
      Ceed ceed;
      ierr = CeedOperatorGetCeed(op, &ceed); CeedChkBackend(ierr);
      return CeedError(ceed, CEED_ERROR_BACKEND,
                       "CEED_EVAL_WEIGHT cannot be an output evaluation mode");
      break; // Should not occur
    }
    case CEED_EVAL_DIV:
      break; // TODO: Not implemented
    case CEED_EVAL_CURL:
      break; // TODO: Not implemented
      // LCOV_EXCL_STOP
    }
    // TODO put in a function
    // Restriction
      bool is_strided;
      ierr = CeedElemRestrictionIsStrided(Erestrict, &is_strided); CeedChkBackend(ierr);
    if (!is_strided) {
      ierr = CeedElemRestrictionGetLVectorSize(Erestrict, &lsize);
      CeedChkBackend(ierr);
      code << "    const CeedInt lsize_out_"<<i<<" = "<<lsize<<";\n";
      CeedInt comp_stride;
      ierr = CeedElemRestrictionGetCompStride(Erestrict, &comp_stride); CeedChkBackend(ierr);
      code << "    // CompStride: "<<comp_stride<<"\n";
      ierr = CeedElemRestrictionGetData(Erestrict, &restr_data); CeedChkBackend(ierr);
      data->indices.outputs[i] = restr_data->d_ind;
      code << "    writeDofsOffset"<<dim<<"d<num_comp_out_"<<i<<", "<<comp_stride<<", P_out_"<<i<<">(data, lsize_out_"<<i<<", elem, indices.outputs["<<i<<"], r_v_"<<i<<", d_v_"<<i<<");\n";
    } else {
      bool has_backend_strides;
      ierr = CeedElemRestrictionHasBackendStrides(Erestrict, &has_backend_strides);
      CeedChkBackend(ierr);
      CeedInt num_elem;
      ierr = CeedElemRestrictionGetNumElements(Erestrict, &num_elem);
      CeedChkBackend(ierr);
      CeedInt strides[3] = {1, elem_size*num_elem, elem_size};
      if (!has_backend_strides) {
        ierr = CeedElemRestrictionGetStrides(Erestrict, &strides);
        CeedChkBackend(ierr);
      }
      code << "    // Strides: {"<<strides[0]<<", "<<strides[1]<<", "<<strides[2]<<"}\n";
      code << "    writeDofsStrided"<<dim<<"d<num_comp_out_"<<i<<",P_out_"<<i<<","<<strides[0]<<","<<strides[1]<<","<<strides[2]<<">(data, elem, r_v_"<<i<<", d_v_"<<i<<");\n";
    }
  }

  code << "  }\n";
  code << "}\n";
  code << "// -----------------------------------------------------------------------------\n\n";

  // View kernel for debugging
  CeedDebug256(ceed, 2, "Generated Operator Kernels:\n");
  CeedDebug(ceed, code.str().c_str());

  ierr = CeedCompileCuda(ceed, code.str().c_str(), &data->module, 1,
                         "T_1D", CeedIntMax(Q_1d, data->max_P_1d));
  CeedChkBackend(ierr);
  ierr = CeedGetKernelCuda(ceed, data->module, operator_name.c_str(), &data->op);
  CeedChkBackend(ierr);

  ierr = CeedOperatorSetSetupDone(op); CeedChkBackend(ierr);
  return CEED_ERROR_SUCCESS;
}
//------------------------------------------------------------------------------