Home
last modified time | relevance | path

Searched refs:r_u (Results 1 – 10 of 10) sorted by relevance

/libCEED/include/ceed/jit-source/sycl/
H A Dsycl-gen-templates.h34 …edInt *restrict indices, const global CeedScalar *restrict d_u, private CeedScalar *restrict r_u) { in readDofsOffset1d() argument
42 r_u[comp] = d_u[ind + strides_comp * comp]; in readDofsOffset1d()
52 private CeedScalar *restrict r_u) { in readDofsStrided1d() argument
60 r_u[comp] = d_u[ind + comp * strides_comp]; in readDofsStrided1d()
107 …edInt *restrict indices, const global CeedScalar *restrict d_u, private CeedScalar *restrict r_u) { in readDofsOffset2d() argument
115 for (CeedInt comp = 0; comp < num_comp; ++comp) r_u[comp] = d_u[ind + strides_comp * comp]; in readDofsOffset2d()
124 private CeedScalar *restrict r_u) { in readDofsStrided2d() argument
132 for (CeedInt comp = 0; comp < num_comp; ++comp) r_u[comp] = d_u[ind + comp * strides_comp]; in readDofsStrided2d()
178 …edInt *restrict indices, const global CeedScalar *restrict d_u, private CeedScalar *restrict r_u) { in readDofsOffset3d() argument
187 …for (CeedInt comp = 0; comp < num_comp; ++comp) r_u[z + comp * P_1D] = d_u[ind + strides_comp * co… in readDofsOffset3d()
[all …]
H A Dsycl-shared-basis-read-write-templates.h30 private CeedScalar *restrict r_u) { in ReadElementStrided1d() argument
38 r_u[comp] = d_u[ind + comp * strides_comp]; in ReadElementStrided1d()
70 private CeedScalar *restrict r_u) { in ReadElementStrided2d() argument
79 r_u[comp] = d_u[ind + comp * strides_comp]; in ReadElementStrided2d()
112 private CeedScalar *restrict r_u) { in ReadElementStrided3d() argument
122 r_u[z + comp * P_1D] = d_u[ind + comp * strides_comp]; in ReadElementStrided3d()
/libCEED/include/ceed/jit-source/cuda/
H A Dcuda-shared-basis-read-write-templates.h29 … const CeedInt strides_elem, const CeedScalar *__restrict__ d_u, CeedScalar *r_u) { in ReadElementStrided1d() argument
35 r_u[comp] = d_u[ind + comp * strides_comp]; in ReadElementStrided1d()
78 … const CeedInt strides_elem, const CeedScalar *__restrict__ d_u, CeedScalar *r_u) { in ReadElementStrided2d() argument
84 r_u[comp] = d_u[ind + comp * strides_comp]; in ReadElementStrided2d()
127 … const CeedInt strides_elem, const CeedScalar *__restrict__ d_u, CeedScalar *r_u) { in ReadElementStrided3d() argument
134 r_u[z + comp * P_1D] = d_u[ind + comp * strides_comp]; in ReadElementStrided3d()
183 const CeedScalar *__restrict__ d_u, CeedScalar *r_u) { in ReadPoint() argument
188 r_u[comp] = d_u[ind + comp * strides_comp]; in ReadPoint()
192 r_u[comp] = 0.0; in ReadPoint()
H A Dcuda-gen-templates.h29 … const CeedInt *__restrict__ indices, const CeedScalar *__restrict__ d_u, CeedScalar *r_u) { in ReadPoint() argument
33 r_u[comp] = d_u[ind + comp * COMP_STRIDE]; in ReadPoint()
42 … const CeedInt *__restrict__ indices, const CeedScalar *__restrict__ r_u, CeedScalar *d_u) { in WritePoint() argument
47 d_u[ind + comp * COMP_STRIDE] += r_u[comp]; in WritePoint()
74 … const CeedScalar *__restrict__ d_u, CeedScalar *__restrict__ r_u) { in ReadLVecStandard1d() argument
79 for (CeedInt comp = 0; comp < NUM_COMP; comp++) r_u[comp] = d_u[ind + COMP_STRIDE * comp]; in ReadLVecStandard1d()
88 CeedScalar *__restrict__ r_u) { in ReadLVecStrided1d() argument
93 for (CeedInt comp = 0; comp < NUM_COMP; comp++) r_u[comp] = d_u[ind + comp * STRIDES_COMP]; in ReadLVecStrided1d()
196 … const CeedScalar *__restrict__ d_u, CeedScalar *__restrict__ r_u) { in ReadLVecStandard2d() argument
201 for (CeedInt comp = 0; comp < NUM_COMP; comp++) r_u[comp] = d_u[ind + COMP_STRIDE * comp]; in ReadLVecStandard2d()
[all …]
H A Dcuda-ref-qfunction.h16 …oid readQuads(const CeedInt quad, const CeedInt num_qpts, const CeedScalar *d_u, CeedScalar *r_u) { in readQuads() argument
18 r_u[comp] = d_u[quad + num_qpts * comp]; in readQuads()
H A Dcuda-shared-basis-tensor-at-points-templates.h183 const CeedScalar r_u = p < NUM_POINTS ? r_U[comp] : 0.0; in InterpTransposeAtPoints2d() local
186 buffer[i] = chebyshev_x[i] * r_u; in InterpTransposeAtPoints2d()
258 const CeedScalar r_u = p < NUM_POINTS ? r_U[comp + dim * NUM_COMP] : 0.0; in GradTransposeAtPoints2d() local
261 buffer[i] = chebyshev_x[i] * r_u; in GradTransposeAtPoints2d()
344 const CeedScalar r_u = p < NUM_POINTS ? r_U[comp] : 0.0; in InterpTransposeAtPoints3d() local
347 buffer[i] = chebyshev_x[i] * r_u * z; in InterpTransposeAtPoints3d()
442 const CeedScalar r_u = (p < NUM_POINTS) ? r_U[comp + dim * NUM_COMP] : 0.0; in GradTransposeAtPoints3d() local
445 buffer[i] = chebyshev_x[i] * r_u * zz; in GradTransposeAtPoints3d()
/libCEED/include/ceed/jit-source/hip/
H A Dhip-shared-basis-read-write-templates.h29 … const CeedInt strides_elem, const CeedScalar *__restrict__ d_u, CeedScalar *r_u) { in ReadElementStrided1d() argument
35 r_u[comp] = d_u[ind + comp * strides_comp]; in ReadElementStrided1d()
78 … const CeedInt strides_elem, const CeedScalar *__restrict__ d_u, CeedScalar *r_u) { in ReadElementStrided2d() argument
84 r_u[comp] = d_u[ind + comp * strides_comp]; in ReadElementStrided2d()
127 … const CeedInt strides_elem, const CeedScalar *__restrict__ d_u, CeedScalar *r_u) { in ReadElementStrided3d() argument
134 r_u[z + comp * P_1D] = d_u[ind + comp * strides_comp]; in ReadElementStrided3d()
182 …nt strides_comp, const CeedInt strides_elem, const CeedScalar *__restrict__ d_u, CeedScalar *r_u) { in ReadPoint() argument
187 r_u[comp] = d_u[ind + comp * strides_comp]; in ReadPoint()
191 r_u[comp] = 0.0; in ReadPoint()
H A Dhip-gen-templates.h29 … const CeedInt *__restrict__ indices, const CeedScalar *__restrict__ d_u, CeedScalar *r_u) { in ReadPoint() argument
33 r_u[comp] = d_u[ind + comp * COMP_STRIDE]; in ReadPoint()
42 … const CeedInt *__restrict__ indices, const CeedScalar *__restrict__ r_u, CeedScalar *d_u) { in WritePoint() argument
47 d_u[ind + comp * COMP_STRIDE] += r_u[comp]; in WritePoint()
74 … const CeedScalar *__restrict__ d_u, CeedScalar *__restrict__ r_u) { in ReadLVecStandard1d() argument
79 for (CeedInt comp = 0; comp < NUM_COMP; comp++) r_u[comp] = d_u[ind + COMP_STRIDE * comp]; in ReadLVecStandard1d()
87 …_Hip &data, const CeedInt elem, const CeedScalar *__restrict__ d_u, CeedScalar *__restrict__ r_u) { in ReadLVecStrided1d() argument
92 for (CeedInt comp = 0; comp < NUM_COMP; comp++) r_u[comp] = d_u[ind + comp * STRIDES_COMP]; in ReadLVecStrided1d()
195 … const CeedScalar *__restrict__ d_u, CeedScalar *__restrict__ r_u) { in ReadLVecStandard2d() argument
200 for (CeedInt comp = 0; comp < NUM_COMP; comp++) r_u[comp] = d_u[ind + COMP_STRIDE * comp]; in ReadLVecStandard2d()
[all …]
H A Dhip-ref-qfunction.h16 …oid readQuads(const CeedInt quad, const CeedInt num_qpts, const CeedScalar *d_u, CeedScalar *r_u) { in readQuads() argument
18 r_u[comp] = d_u[quad + num_qpts * comp]; in readQuads()
H A Dhip-shared-basis-tensor-at-points-templates.h184 const CeedScalar r_u = p < NUM_POINTS ? r_U[comp] : 0.0; in InterpTransposeAtPoints2d() local
187 buffer[i] = chebyshev_x[i] * r_u; in InterpTransposeAtPoints2d()
259 const CeedScalar r_u = p < NUM_POINTS ? r_U[comp + dim * NUM_COMP] : 0.0; in GradTransposeAtPoints2d() local
262 buffer[i] = chebyshev_x[i] * r_u; in GradTransposeAtPoints2d()
345 const CeedScalar r_u = p < NUM_POINTS ? r_U[comp] : 0.0; in InterpTransposeAtPoints3d() local
348 buffer[i] = chebyshev_x[i] * r_u * z; in InterpTransposeAtPoints3d()
443 const CeedScalar r_u = p < NUM_POINTS ? r_U[comp + dim * NUM_COMP] : 0.0; in GradTransposeAtPoints3d() local
446 buffer[i] = chebyshev_x[i] * r_u * zz; in GradTransposeAtPoints3d()