xref: /libCEED/include/ceed/jit-source/hip/hip-ref-restriction-strided.h (revision bcbe1c99be4be6e5341a66695d3b47170158a1bf)
1 // Copyright (c) 2017-2022, Lawrence Livermore National Security, LLC and other CEED contributors.
2 // All Rights Reserved. See the top-level LICENSE and NOTICE files for details.
3 //
4 // SPDX-License-Identifier: BSD-2-Clause
5 //
6 // This file is part of CEED:  http://github.com/ceed
7 
8 /// @file
9 /// Internal header for HIP strided element restriction kernels
10 #ifndef CEED_HIP_REF_RESTRICTION_STRIDED_H
11 #define CEED_HIP_REF_RESTRICTION_STRIDED_H
12 
13 #include <ceed.h>
14 
15 //------------------------------------------------------------------------------
16 // L-vector -> E-vector, strided
17 //------------------------------------------------------------------------------
18 extern "C" __global__ void StridedNoTranspose(const CeedScalar *__restrict__ u, CeedScalar *__restrict__ v) {
19   for (CeedInt node = blockIdx.x * blockDim.x + threadIdx.x; node < RSTR_NUM_ELEM * RSTR_ELEM_SIZE; node += blockDim.x * gridDim.x) {
20     const CeedInt loc_node = node % RSTR_ELEM_SIZE;
21     const CeedInt elem     = node / RSTR_ELEM_SIZE;
22 
23     for (CeedInt comp = 0; comp < RSTR_NUM_COMP; comp++) {
24       v[loc_node + comp * RSTR_ELEM_SIZE * RSTR_NUM_ELEM + elem * RSTR_ELEM_SIZE] =
25           u[loc_node * RSTR_STRIDE_NODES + comp * RSTR_STRIDE_COMP + elem * RSTR_STRIDE_ELEM];
26     }
27   }
28 }
29 
30 //------------------------------------------------------------------------------
31 // E-vector -> L-vector, strided
32 //------------------------------------------------------------------------------
33 extern "C" __global__ void StridedTranspose(const CeedScalar *__restrict__ u, CeedScalar *__restrict__ v) {
34   for (CeedInt node = blockIdx.x * blockDim.x + threadIdx.x; node < RSTR_NUM_ELEM * RSTR_ELEM_SIZE; node += blockDim.x * gridDim.x) {
35     const CeedInt loc_node = node % RSTR_ELEM_SIZE;
36     const CeedInt elem     = node / RSTR_ELEM_SIZE;
37 
38     for (CeedInt comp = 0; comp < RSTR_NUM_COMP; comp++) {
39       v[loc_node * RSTR_STRIDE_NODES + comp * RSTR_STRIDE_COMP + elem * RSTR_STRIDE_ELEM] +=
40           u[loc_node + comp * RSTR_ELEM_SIZE * RSTR_NUM_ELEM + elem * RSTR_ELEM_SIZE];
41     }
42   }
43 }
44 
45 //------------------------------------------------------------------------------
46 
47 #endif  // CEED_HIP_REF_RESTRICTION_STRIDED_H
48