15 extern "C" __global__ void OffsetNoTranspose(const CeedInt *__restrict__ indices, const CeedScalar …  in OffsetNoTranspose()
16 …for (CeedInt node = blockIdx.x * blockDim.x + threadIdx.x; node < RSTR_NUM_ELEM * RSTR_ELEM_SIZE; …  in OffsetNoTranspose()
17     const CeedInt ind      = indices[node];  in OffsetNoTranspose()
18     const CeedInt loc_node = node % RSTR_ELEM_SIZE;  in OffsetNoTranspose()
19     const CeedInt elem     = node / RSTR_ELEM_SIZE;  in OffsetNoTranspose()
21     for (CeedInt comp = 0; comp < RSTR_NUM_COMP; comp++) {  in OffsetNoTranspose()
31 extern "C" __global__ void OffsetTranspose(const CeedInt *__restrict__ indices, const CeedScalar *_…  in OffsetTranspose()
32 …for (CeedInt node = blockIdx.x * blockDim.x + threadIdx.x; node < RSTR_NUM_ELEM * RSTR_ELEM_SIZE; …  in OffsetTranspose()
33     const CeedInt ind      = indices[node];  in OffsetTranspose()
34     const CeedInt loc_node = node % RSTR_ELEM_SIZE;  in OffsetTranspose()
35     const CeedInt elem     = node / RSTR_ELEM_SIZE;  in OffsetTranspose()
37     for (CeedInt comp = 0; comp < RSTR_NUM_COMP; comp++) {  in OffsetTranspose()
43 extern "C" __global__ void OffsetTranspose(const CeedInt *__restrict__ l_vec_indices, const CeedInt…  in OffsetTranspose()
44 …const CeedInt *__restrict__ t_offsets, const CeedScalar *__restrict__ u, CeedScalar *__restrict__ …  in OffsetTranspose()
47 …for (CeedInt i = blockIdx.x * blockDim.x + threadIdx.x; i < RSTR_NUM_NODES; i += blockDim.x * grid…  in OffsetTranspose()
48     const CeedInt ind     = l_vec_indices[i];  in OffsetTranspose()
49     const CeedInt range_1 = t_offsets[i];  in OffsetTranspose()
50     const CeedInt range_N = t_offsets[i + 1];  in OffsetTranspose()
52     for (CeedInt comp = 0; comp < RSTR_NUM_COMP; comp++) value[comp] = 0.0;  in OffsetTranspose()
54     for (CeedInt j = range_1; j < range_N; j++) {  in OffsetTranspose()
55       const CeedInt t_ind    = t_indices[j];  in OffsetTranspose()
56       const CeedInt loc_node = t_ind % RSTR_ELEM_SIZE;  in OffsetTranspose()
57       const CeedInt elem     = t_ind / RSTR_ELEM_SIZE;  in OffsetTranspose()
59       for (CeedInt comp = 0; comp < RSTR_NUM_COMP; comp++) {  in OffsetTranspose()
64 …for (CeedInt comp = 0; comp < RSTR_NUM_COMP; comp++) v[ind + comp * RSTR_COMP_STRIDE] += value[com…  in OffsetTranspose()