libceed-sys 0.12.0

Low-level bindings for libCEED library.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98

// 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

/// @file
/// Internal header for CUDA non-tensor product basis
#ifndef CEED_CUDA_REF_BASIS_NONTENSOR_H
#define CEED_CUDA_REF_BASIS_NONTENSOR_H

#include <ceed.h>

//------------------------------------------------------------------------------
// Non-Tensor Basis Kernels
//------------------------------------------------------------------------------

//------------------------------------------------------------------------------
// Interp
//------------------------------------------------------------------------------
extern "C" __global__ void Interp(const CeedInt num_elem, const CeedInt transpose, const CeedScalar *d_B, const CeedScalar *__restrict__ d_U,
                                  CeedScalar *__restrict__ d_V) {
  const CeedInt     t_id = threadIdx.x;
  const CeedScalar *U;
  CeedScalar        V;
  // TODO load B in shared memory if blockDim.z > 1?

  for (CeedInt elem = blockIdx.x * blockDim.z + threadIdx.z; elem < num_elem; elem += gridDim.x * blockDim.z) {
    for (CeedInt comp = 0; comp < BASIS_NUM_COMP; comp++) {
      if (transpose) {  // run with P threads
        U = d_U + elem * BASIS_Q + comp * num_elem * BASIS_Q;
        V = 0.0;
        for (CeedInt i = 0; i < BASIS_Q; i++) V += d_B[t_id + i * BASIS_P] * U[i];

        d_V[elem * BASIS_P + comp * num_elem * BASIS_P + t_id] = V;
      } else {  // run with Q threads
        U = d_U + elem * BASIS_P + comp * num_elem * BASIS_P;
        V = 0.0;
        for (CeedInt i = 0; i < BASIS_P; i++) V += d_B[i + t_id * BASIS_P] * U[i];

        d_V[elem * BASIS_Q + comp * num_elem * BASIS_Q + t_id] = V;
      }
    }
  }
}

//------------------------------------------------------------------------------
// Grad
//------------------------------------------------------------------------------
extern "C" __global__ void Grad(const CeedInt num_elem, const CeedInt transpose, const CeedScalar *d_G, const CeedScalar *__restrict__ d_U,
                                CeedScalar *__restrict__ d_V) {
  const CeedInt     t_id = threadIdx.x;
  const CeedScalar *U;
  // TODO load G in shared memory if blockDim.z > 1?

  for (CeedInt elem = blockIdx.x * blockDim.z + threadIdx.z; elem < num_elem; elem += gridDim.x * blockDim.z) {
    for (CeedInt comp = 0; comp < BASIS_NUM_COMP; comp++) {
      if (transpose) {  // run with P threads
        CeedScalar V = 0.0;
        for (CeedInt dim = 0; dim < BASIS_DIM; dim++) {
          U = d_U + elem * BASIS_Q + comp * num_elem * BASIS_Q + dim * BASIS_NUM_COMP * num_elem * BASIS_Q;
          for (CeedInt i = 0; i < BASIS_Q; i++) V += d_G[t_id + i * BASIS_P + dim * BASIS_P * BASIS_Q] * U[i];
        }

        d_V[elem * BASIS_P + comp * num_elem * BASIS_P + t_id] = V;
      } else {  // run with Q threads
        CeedScalar V[BASIS_DIM];
        U = d_U + elem * BASIS_P + comp * num_elem * BASIS_P;
        for (CeedInt dim = 0; dim < BASIS_DIM; dim++) V[dim] = 0.0;
        for (CeedInt i = 0; i < BASIS_P; i++) {
          const CeedScalar val = U[i];
          for (CeedInt dim = 0; dim < BASIS_DIM; dim++) V[dim] += d_G[i + t_id * BASIS_P + dim * BASIS_P * BASIS_Q] * val;
        }

        for (CeedInt dim = 0; dim < BASIS_DIM; dim++) {
          d_V[elem * BASIS_Q + comp * num_elem * BASIS_Q + dim * BASIS_NUM_COMP * num_elem * BASIS_Q + t_id] = V[dim];
        }
      }
    }
  }
}

//------------------------------------------------------------------------------
// Weight
//------------------------------------------------------------------------------
extern "C" __global__ void Weight(const CeedInt num_elem, const CeedScalar *__restrict__ q_weight, CeedScalar *__restrict__ d_V) {
  const CeedInt t_id = threadIdx.x;

  // TODO load q_weight in shared memory if blockDim.z > 1?
  for (CeedInt elem = blockIdx.x * blockDim.z + threadIdx.z; elem < num_elem; elem += gridDim.x * blockDim.z) {
    d_V[elem * BASIS_Q + t_id] = q_weight[t_id];
  }
}

//------------------------------------------------------------------------------

#endif  // CEED_CUDA_REF_BASIS_NONTENSOR_H