Home
last modified time | relevance | path

Searched refs:NUM_COMP (Results 1 – 25 of 27) sorted by relevance

12

/libCEED/include/ceed/jit-source/cuda/
H A Dcuda-shared-basis-tensor-templates.h19 template <int NUM_COMP, int P_1D, int Q_1D>
35 template <int NUM_COMP, int P_1D, int Q_1D>
51 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
53 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in Interp1d()
54 ContractX1d<NUM_COMP, P_1D, Q_1D>(data, &r_U[comp], c_B, &r_V[comp]); in Interp1d()
61 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
64 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in InterpTranspose1d()
65 ContractTransposeX1d<NUM_COMP, P_1D, Q_1D>(data, &r_U[comp], c_B, &r_V[comp]); in InterpTranspose1d()
72 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
75 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in InterpCollocatedNodes1d()
[all …]
H A Dcuda-shared-basis-tensor-flattened-templates.h19 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
36 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
53 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
70 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
87 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
103 template <int NUM_COMP, int Q_1D, int T_1D>
107 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in QPack2d()
115 template <int NUM_COMP, int Q_1D, int T_1D>
119 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in QUnpack2d()
130 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
[all …]
H A Dcuda-shared-basis-nontensor-templates.h15 template <int NUM_COMP, int P_1D, int Q_1D>
31 template <int NUM_COMP, int P_1D, int Q_1D>
46 template <int NUM_COMP, int P, int Q, int T_1D>
49 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in InterpNonTensor()
50 Contract1d<NUM_COMP, P, Q>(data, &r_U[comp], c_B, &r_V[comp]); in InterpNonTensor()
57 template <int NUM_COMP, int P, int Q, int T_1D>
60 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in InterpTransposeNonTensor()
62 ContractTranspose1d<NUM_COMP, P, Q>(data, &r_U[comp], c_B, &r_V[comp]); in InterpTransposeNonTensor()
69 template <int NUM_COMP, int DIM, int P, int Q, int T_1D>
72 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in GradNonTensor()
[all …]
H A Dcuda-gen-templates.h27 template <int NUM_COMP, int COMP_STRIDE, int NUM_PTS>
32 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in ReadPoint()
40 template <int NUM_COMP, int COMP_STRIDE, int NUM_PTS>
46 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in WritePoint()
59 template <int NUM_COMP, int P_1D>
72 template <int NUM_COMP, int COMP_STRIDE, int P_1D>
79 for (CeedInt comp = 0; comp < NUM_COMP; comp++) r_u[comp] = d_u[ind + COMP_STRIDE * comp]; in ReadLVecStandard1d()
86 template <int NUM_COMP, int P_1D, int STRIDES_NODE, int STRIDES_COMP, int STRIDES_ELEM>
93 for (CeedInt comp = 0; comp < NUM_COMP; comp++) r_u[comp] = d_u[ind + comp * STRIDES_COMP]; in ReadLVecStrided1d()
100 template <int NUM_COMP, int COMP_STRIDE, int P_1D>
[all …]
H A Dcuda-shared-basis-read-write-templates.h27 template <int NUM_COMP, int P_1D>
34 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in ReadElementStrided1d()
43 template <int NUM_COMP, int P_1D>
50 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in WriteElementStrided1d()
56 template <int NUM_COMP, int P_1D>
63 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in SumElementStrided1d()
76 template <int NUM_COMP, int P_1D>
83 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in ReadElementStrided2d()
92 template <int NUM_COMP, int P_1D>
99 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in WriteElementStrided2d()
[all …]
H A Dcuda-shared-basis-tensor-at-points-templates.h43 template <int NUM_COMP, int NUM_POINTS, int P_1D, int Q_1D>
48 for (CeedInt i = 0; i < NUM_COMP; i++) r_V[i] = 0.0; in InterpAtPoints1d()
50 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in InterpAtPoints1d()
64 template <int NUM_COMP, int NUM_POINTS, int P_1D, int Q_1D>
70 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in InterpTransposeAtPoints1d()
89 template <int NUM_COMP, int NUM_POINTS, int P_1D, int Q_1D>
95 for (CeedInt i = 0; i < NUM_COMP; i++) r_V[i] = 0.0; in GradAtPoints1d()
96 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in GradAtPoints1d()
111 template <int NUM_COMP, int NUM_POINTS, int P_1D, int Q_1D>
117 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in GradTransposeAtPoints1d()
[all …]
H A Dcuda-ref-operator-assemble-diagonal.h82 for (IndexType comp_out = 0; comp_out < NUM_COMP; comp_out++) { in __launch_bounds__()
85 for (IndexType comp_in = 0; comp_in < NUM_COMP; comp_in++) { in __launch_bounds__()
91 …assembled_qf_array[((((e_in * NUM_COMP + comp_in) * NUM_EVAL_MODES_OUT + e_out) * NUM_COMP + comp_… in __launch_bounds__()
96 … elem_diag_array[((comp_out * NUM_COMP + comp_in) * num_elem + e) * NUM_NODES + tid] += e_value; in __launch_bounds__()
105 …assembled_qf_array[((((e_in * NUM_COMP + comp_out) * NUM_EVAL_MODES_OUT + e_out) * NUM_COMP + comp… in __launch_bounds__()
H A Dcuda-ref-basis-nontensor-templates.h15 template <int NUM_COMP, int Q_COMP, int P, int Q>
24 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in Contract()
42 template <int NUM_COMP, int Q_COMP, int P, int Q>
51 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in ContractTranspose()
/libCEED/include/ceed/jit-source/hip/
H A Dhip-shared-basis-tensor-templates.h19 template <int NUM_COMP, int P_1D, int Q_1D>
35 template <int NUM_COMP, int P_1D, int Q_1D>
51 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
53 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in Interp1d()
54 ContractX1d<NUM_COMP, P_1D, Q_1D>(data, &r_U[comp], c_B, &r_V[comp]); in Interp1d()
61 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
64 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in InterpTranspose1d()
65 ContractTransposeX1d<NUM_COMP, P_1D, Q_1D>(data, &r_U[comp], c_B, &r_V[comp]); in InterpTranspose1d()
72 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
75 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in InterpCollocatedNodes1d()
[all …]
H A Dhip-shared-basis-tensor-flattened-templates.h19 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
36 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
53 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
70 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
87 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
103 template <int NUM_COMP, int Q_1D, int T_1D>
107 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in QPack2d()
115 template <int NUM_COMP, int Q_1D, int T_1D>
119 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in QUnpack2d()
130 template <int NUM_COMP, int P_1D, int Q_1D, int T_1D>
[all …]
H A Dhip-shared-basis-nontensor-templates.h15 template <int NUM_COMP, int P_1D, int Q_1D>
31 template <int NUM_COMP, int P_1D, int Q_1D>
46 template <int NUM_COMP, int P, int Q, int T_1D>
49 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in InterpNonTensor()
50 Contract1d<NUM_COMP, P, Q>(data, &r_U[comp], c_B, &r_V[comp]); in InterpNonTensor()
57 template <int NUM_COMP, int P, int Q, int T_1D>
60 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in InterpTransposeNonTensor()
62 ContractTranspose1d<NUM_COMP, P, Q>(data, &r_U[comp], c_B, &r_V[comp]); in InterpTransposeNonTensor()
69 template <int NUM_COMP, int DIM, int P, int Q, int T_1D>
72 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in GradNonTensor()
[all …]
H A Dhip-gen-templates.h27 template <int NUM_COMP, int COMP_STRIDE, int NUM_PTS>
32 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in ReadPoint()
40 template <int NUM_COMP, int COMP_STRIDE, int NUM_PTS>
46 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in WritePoint()
59 template <int NUM_COMP, int P_1D>
72 template <int NUM_COMP, int COMP_STRIDE, int P_1D>
79 for (CeedInt comp = 0; comp < NUM_COMP; comp++) r_u[comp] = d_u[ind + COMP_STRIDE * comp]; in ReadLVecStandard1d()
86 template <int NUM_COMP, int P_1D, int STRIDES_NODE, int STRIDES_COMP, int STRIDES_ELEM>
92 for (CeedInt comp = 0; comp < NUM_COMP; comp++) r_u[comp] = d_u[ind + comp * STRIDES_COMP]; in ReadLVecStrided1d()
99 template <int NUM_COMP, int COMP_STRIDE, int P_1D>
[all …]
H A Dhip-shared-basis-read-write-templates.h27 template <int NUM_COMP, int P_1D>
34 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in ReadElementStrided1d()
43 template <int NUM_COMP, int P_1D>
50 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in WriteElementStrided1d()
56 template <int NUM_COMP, int P_1D>
63 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in SumElementStrided1d()
76 template <int NUM_COMP, int P_1D>
83 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in ReadElementStrided2d()
92 template <int NUM_COMP, int P_1D>
99 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in WriteElementStrided2d()
[all …]
H A Dhip-shared-basis-tensor-at-points-templates.h43 template <int NUM_COMP, int NUM_POINTS, int P_1D, int Q_1D>
48 for (CeedInt i = 0; i < NUM_COMP; i++) r_V[i] = 0.0; in InterpAtPoints1d()
50 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in InterpAtPoints1d()
65 template <int NUM_COMP, int NUM_POINTS, int P_1D, int Q_1D>
71 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in InterpTransposeAtPoints1d()
90 template <int NUM_COMP, int NUM_POINTS, int P_1D, int Q_1D>
96 for (CeedInt i = 0; i < NUM_COMP; i++) r_V[i] = 0.0; in GradAtPoints1d()
97 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in GradAtPoints1d()
112 template <int NUM_COMP, int NUM_POINTS, int P_1D, int Q_1D>
118 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in GradTransposeAtPoints1d()
[all …]
H A Dhip-ref-operator-assemble-diagonal.h82 for (IndexType comp_out = 0; comp_out < NUM_COMP; comp_out++) { in __launch_bounds__()
85 for (IndexType comp_in = 0; comp_in < NUM_COMP; comp_in++) { in __launch_bounds__()
91 …assembled_qf_array[((((e_in * NUM_COMP + comp_in) * NUM_EVAL_MODES_OUT + e_out) * NUM_COMP + comp_… in __launch_bounds__()
96 … elem_diag_array[((comp_out * NUM_COMP + comp_in) * num_elem + e) * NUM_NODES + tid] += e_value; in __launch_bounds__()
105 …assembled_qf_array[((((e_in * NUM_COMP + comp_out) * NUM_EVAL_MODES_OUT + e_out) * NUM_COMP + comp… in __launch_bounds__()
H A Dhip-ref-basis-nontensor-templates.h15 template <int NUM_COMP, int Q_COMP, int P, int Q>
24 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in Contract()
42 template <int NUM_COMP, int Q_COMP, int P, int Q>
51 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in ContractTranspose()
/libCEED/include/ceed/jit-source/magma/
H A Dmagma-common-tensor.h18 template <typename T, int LENGTH, int NUM_COMP>
19 …_inline__ void read_1d(const T *devptr, const int compstride, T *sBuffer[NUM_COMP], const int tx) { in read_1d() argument
21 for (int comp = 0; comp < NUM_COMP; comp++) { in read_1d()
30 template <typename T, int LENGTH, int NUM_COMP>
31 static __device__ __inline__ void write_1d(T *sBuffer[NUM_COMP], T *devptr, const int compstride, c… in write_1d() argument
33 for (int comp = 0; comp < NUM_COMP; comp++) { in write_1d()
42 template <typename T, int LENGTH, int NUM_COMP>
43 static __device__ __inline__ void sum_1d(T *sBuffer[NUM_COMP], T *devptr, const int compstride, con… in sum_1d() argument
45 for (int comp = 0; comp < NUM_COMP; comp++) { in sum_1d()
58 template <typename T, int P, int DIM_U, int NUM_COMP, int rU_SIZE, int i_DIM>
[all …]
H A Dmagma-basis-weight-2d.h14 template <typename T, int DIM, int NUM_COMP, int Q, int i_DIM, int i_COMP>
15 static __device__ __inline__ void magma_weight_2d_device(const T *sTweight, T rV[DIM][NUM_COMP][Q],… in magma_weight_2d_device()
H A Dmagma-basis-weight-3d.h14 template <typename T, int DIM, int NUM_COMP, int Q, int i_DIM, int i_COMP>
15 static __device__ __inline__ void magma_weight_3d_device(const T *sTweight, T rV[DIM][NUM_COMP][Q],… in magma_weight_3d_device()
H A Dmagma-basis-grad-1d.h17 template <typename T, int DIM, int NUM_COMP, int P, int Q>
18 static __device__ __inline__ void magma_grad_1d_device(const T *sT, T *sU[NUM_COMP], T *sV[NUM_COMP… in magma_grad_1d_device() argument
29 for (int comp = 0; comp < NUM_COMP; comp++) { in magma_grad_1d_device()
H A Dmagma-basis-interp-1d.h17 template <typename T, int DIM, int NUM_COMP, int P, int Q>
18 …ice__ __inline__ void magma_interp_1d_device(const T *sT, T *sU[NUM_COMP], T *sV[NUM_COMP], const … in magma_interp_1d_device() argument
29 for (int comp = 0; comp < NUM_COMP; comp++) { in magma_interp_1d_device()
H A Dmagma-basis-interp-2d.h18 template <typename T, int DIM_U, int DIM_V, int NUM_COMP, int P, int Q, int rU_SIZE, int rV_SIZE>
19 …__ void magma_interp_2d_device(const T *sT, T rU[DIM_U][NUM_COMP][rU_SIZE], T rV[DIM_V][NUM_COMP][… in magma_interp_2d_device()
31 for (int comp = 0; comp < NUM_COMP; comp++) { in magma_interp_2d_device()
H A Dmagma-basis-interp-3d.h18 template <typename T, int DIM_U, int DIM_V, int NUM_COMP, int P, int Q, int rU_SIZE, int rV_SIZE>
19 …__ void magma_interp_3d_device(const T *sT, T rU[DIM_U][NUM_COMP][rU_SIZE], T rV[DIM_V][NUM_COMP][… in magma_interp_3d_device()
32 for (int comp = 0; comp < NUM_COMP; comp++) { in magma_interp_3d_device()
/libCEED/include/ceed/jit-source/sycl/
H A Dsycl-shared-basis-read-write-templates.h28 inline void ReadElementStrided1d(const CeedInt NUM_COMP, const CeedInt P_1D, const CeedInt num_elem… in ReadElementStrided1d() argument
37 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in ReadElementStrided1d()
46 inline void WriteElementStrided1d(const CeedInt NUM_COMP, const CeedInt P_1D, const CeedInt num_ele… in WriteElementStrided1d() argument
55 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in WriteElementStrided1d()
68 inline void ReadElementStrided2d(const CeedInt NUM_COMP, const CeedInt P_1D, const CeedInt num_elem… in ReadElementStrided2d() argument
78 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in ReadElementStrided2d()
87 inline void WriteElementStrided2d(const CeedInt NUM_COMP, const CeedInt P_1D, const CeedInt num_ele… in WriteElementStrided2d() argument
97 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in WriteElementStrided2d()
110 inline void ReadElementStrided3d(const CeedInt NUM_COMP, const CeedInt P_1D, const CeedInt num_elem… in ReadElementStrided3d() argument
121 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in ReadElementStrided3d()
[all …]
H A Dsycl-shared-basis-tensor-templates.h57 inline void Interp1d(const CeedInt NUM_COMP, const CeedInt P_1D, const CeedInt Q_1D, private const … in Interp1d() argument
59 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in Interp1d()
67 inline void InterpTranspose1d(const CeedInt NUM_COMP, const CeedInt P_1D, const CeedInt Q_1D, priva… in InterpTranspose1d() argument
69 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in InterpTranspose1d()
77 inline void Grad1d(const CeedInt NUM_COMP, const CeedInt P_1D, const CeedInt Q_1D, private const Ce… in Grad1d() argument
79 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in Grad1d()
87 inline void GradTranspose1d(const CeedInt NUM_COMP, const CeedInt P_1D, const CeedInt Q_1D, private… in GradTranspose1d() argument
89 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in GradTranspose1d()
208 inline void InterpTensor2d(const CeedInt NUM_COMP, const CeedInt P_1D, const CeedInt Q_1D, private … in InterpTensor2d() argument
212 for (CeedInt comp = 0; comp < NUM_COMP; comp++) { in InterpTensor2d()
[all …]

12