1
0
mirror of https://github.com/paboyle/Grid.git synced 2024-11-10 15:55:37 +00:00
Grid/Grid_stencil.h
2015-04-10 05:24:01 +02:00

364 lines
10 KiB
C++

#ifndef GRID_STENCIL_H
#define GRID_STENCIL_H
//////////////////////////////////////////////////////////////////////////////////////////
// Must not lose sight that goal is to be able to construct really efficient
// gather to a point stencil code. CSHIFT is not the best way, so probably need
// additional stencil support.
//
// Stencil based code could pre-exchange haloes and use a table lookup for neighbours
//
// Lattice <foo> could also allocate haloes which get used for stencil code.
//
// Grid could create a neighbour index table for a given stencil.
//
// Could also implement CovariantCshift, to fuse the loops and enhance performance.
//
//
// General stencil computation:
//
// Generic services
// 0) Prebuild neighbour tables
// 1) Compute sizes of all haloes/comms buffers; allocate them.
//
// 2) Gather all faces, and communicate.
// 3) Loop over result sites, giving nbr index/offnode info for each
//
// Could take a
// SpinProjectFaces
// start comms
// complete comms
// Reconstruct Umu
//
// Approach.
//
//////////////////////////////////////////////////////////////////////////////////////////
namespace Grid {
class Stencil {
public:
Stencil(GridBase *grid,
int npoints,
int checkerboard,
std::vector<int> directions,
std::vector<int> distances);
void Stencil_local (int dimension,int shift,int cbmask);
void Stencil_comms (int dimension,int shift,int cbmask);
void Stencil_comms_simd(int dimension,int shift,int cbmask);
// Will need to implement actions for
Copy_plane;
Copy_plane_permute;
Gather_plane;
// The offsets to all neibours in stencil in each direction
int _checkerboard;
int _npoints; // Move to template param?
GridBase * _grid;
// Store these as SIMD Integer needed
//
// std::vector< iVector<Integer, Npoint> > _offsets;
// std::vector< iVector<Integer, Npoint> > _local;
// std::vector< iVector<Integer, Npoint> > _comm_buf_size;
// std::vector< iVector<Integer, Npoint> > _permute;
std::vector<std::vector<int> > _offsets;
std::vector<std::vector<int> > _local;
std::vector<int> _comm_buf_size;
std::vector<int> _permute;
};
Stencil::Stencil(GridBase *grid,
int npoints,
int checkerboard,
std::vector<int> directions,
std::vector<int> distances){
_npoints = npoints;
_grid = grid;
for(int i=0;i<npoints;i++){
int dimension = directions[i];
int displacement = distances[i];
int fd = _grid->_fdimensions[dimension];
int rd = _grid->_rdimensions[dimension];
_checkerboard = checkerboard;
// the permute type
int simd_layout = _grid->_simd_layout[dimension];
int comm_dim = _grid->_processors[dimension] >1 ;
int splice_dim = _grid->_simd_layout[dimension]>1 && (comm_dim);
int sshift[2];
if ( !comm_dim ) {
sshift[0] = _grid->CheckerBoardShift(_checkerboard,dimension,shift,0);
sshift[1] = _grid->CheckerBoardShift(_checkerboard,dimension,shift,1);
if ( sshift[0] == sshift[1] ) {
Stencil_local(dimension,shift,0x3);
} else {
Stencil_local(dimension,shift,0x1);// if checkerboard is unfavourable take two passes
Stencil_local(dimension,shift,0x2);// both with block stride loop iteration
}
} else if ( splice_dim ) {
sshift[0] = _grid->CheckerBoardShift(_checkerboard,dimension,shift,0);
sshift[1] = _grid->CheckerBoardShift(_checkerboard,dimension,shift,1);
if ( sshift[0] == sshift[1] ) {
Stencil_comms_simd(dimension,shift,0x3);
} else {
Stencil_comms_simd(dimension,shift,0x1);// if checkerboard is unfavourable take two passes
Stencil_comms_simd(dimension,shift,0x2);// both with block stride loop iteration
}
} else {
// Cshift_comms(ret,rhs,dimension,shift);
sshift[0] = _grid->CheckerBoardShift(_checkerboard,dimension,shift,0);
sshift[1] = _grid->CheckerBoardShift(_checkerboard,dimension,shift,1);
if ( sshift[0] == sshift[1] ) {
Stencil_comms(dimension,shift,0x3);
} else {
Stencil_comms(dimension,shift,0x1);// if checkerboard is unfavourable take two passes
Stencil_comms(dimension,shift,0x2);// both with block stride loop iteration
}
}
}
}
void Stencil::Stencil_local (int dimension,int shift,int cbmask)
{
int fd = _grid->_fdimensions[dimension];
int rd = _grid->_rdimensions[dimension];
int ld = _grid->_ldimensions[dimension];
int gd = _grid->_gdimensions[dimension];
// Map to always positive shift modulo global full dimension.
shift = (shift+fd)%fd;
// the permute type
int permute_dim =_grid->PermuteDim(dimension);
int permute_type=_grid->PermuteType(dimension);
for(int x=0;x<rd;x++){
int o = 0;
int bo = x * _grid->_ostride[dimension];
int cb= (cbmask==0x2)? 1 : 0;
int sshift = _grid->CheckerBoardShift(_checkerboard,dimension,shift,cb);
int sx = (x+sshift)%rd;
int permute_slice=0;
if(permute_dim){
int wrap = sshift/rd;
int num = sshift%rd;
if ( x< rd-num ) permute_slice=wrap;
else permute_slice = 1-wrap;
}
if ( permute_slice ) Copy_plane_permute(dimension,x,sx,cbmask,permute_type);
else Copy_plane (dimension,x,sx,cbmask);
}
}
void Stencil::Stencil_comms (int dimension,int shift,int cbmask)
{
typedef typename vobj::vector_type vector_type;
typedef typename vobj::scalar_type scalar_type;
GridBase *grid=_grid;
int fd = _grid->_fdimensions[dimension];
int rd = _grid->_rdimensions[dimension];
int simd_layout = _grid->_simd_layout[dimension];
int comm_dim = _grid->_processors[dimension] >1 ;
assert(simd_layout==1);
assert(comm_dim==1);
assert(shift>=0);
assert(shift<fd);
int buffer_size = _grid->_slice_nblock[dimension]*rhs._grid->_slice_block[dimension];
// FIXME: Do something with buffer_size??
int cb= (cbmask==0x2)? 1 : 0;
int sshift= _grid->CheckerBoardShift(_checkerboard,dimension,shift,cb);
for(int x=0;x<rd;x++){
int offnode = ( x+sshift >= rd );
int sx = (x+sshift)%rd;
int comm_proc = (x+sshift)/rd;
if (!offnode) {
Copy_plane(dimension,x,sx,cbmask);
} else {
int words = send_buf.size();
if (cbmask != 0x3) words=words>>1;
int bytes = words * sizeof(vobj);
Gather_plane_simple (dimension,sx,cbmask);
int rank = grid->_processor;
int recv_from_rank;
int xmit_to_rank;
grid->ShiftedRanks(dimension,comm_proc,xmit_to_rank,recv_from_rank);
/*
grid->SendToRecvFrom((void *)&send_buf[0],
xmit_to_rank,
(void *)&recv_buf[0],
recv_from_rank,
bytes);
*/
Scatter_plane_simple (dimension,x,cbmask);
}
}
}
void Stencil::Stencil_comms_simd(int dimension,int shift,int cbmask)
{
GridBase *grid=_grid;
const int Nsimd = _grid->Nsimd();
typedef typename vobj::vector_type vector_type;
typedef typename vobj::scalar_type scalar_type;
int fd = _grid->_fdimensions[dimension];
int rd = _grid->_rdimensions[dimension];
int ld = _grid->_ldimensions[dimension];
int simd_layout = _grid->_simd_layout[dimension];
int comm_dim = _grid->_processors[dimension] >1 ;
assert(comm_dim==1);
assert(simd_layout==2);
assert(shift>=0);
assert(shift<fd);
int permute_type=_grid->PermuteType(dimension);
///////////////////////////////////////////////
// Simd direction uses an extract/merge pair
///////////////////////////////////////////////
int buffer_size = _grid->_slice_nblock[dimension]*grid->_slice_block[dimension];
// FIXME do something with buffer size
std::vector<scalar_type *> pointers(Nsimd); //
std::vector<scalar_type *> rpointers(Nsimd); // received pointers
///////////////////////////////////////////
// Work out what to send where
///////////////////////////////////////////
int cb = (cbmask==0x2)? 1 : 0;
int sshift= _grid->CheckerBoardShift(_checkerboard,dimension,shift,cb);
std::vector<int> comm_offnode(simd_layout);
std::vector<int> comm_proc (simd_layout); //relative processor coord in dim=dimension
std::vector<int> icoor(grid->Nd());
for(int x=0;x<rd;x++){
int comm_any = 0;
for(int s=0;s<simd_layout;s++) {
int shifted_x = x+s*rd+sshift;
comm_offnode[s] = shifted_x >= ld;
comm_any = comm_any | comm_offnode[s];
comm_proc[s] = shifted_x/ld;
}
int o = 0;
int bo = x*grid->_ostride[dimension];
int sx = (x+sshift)%rd;
if ( comm_any ) {
for(int i=0;i<Nsimd;i++){
pointers[i] = (scalar_type *)&send_buf_extract[i][0];
}
Gather_plane_extract(rhs,pointers,dimension,sx,cbmask);
for(int i=0;i<Nsimd;i++){
int s;
grid->iCoorFromIindex(icoor,i);
s = icoor[dimension];
if(comm_offnode[s]){
int rank = grid->_processor;
int recv_from_rank;
int xmit_to_rank;
grid->ShiftedRanks(dimension,comm_proc[s],xmit_to_rank,recv_from_rank);
/*
grid->SendToRecvFrom((void *)&send_buf_extract[i][0],
xmit_to_rank,
(void *)&recv_buf_extract[i][0],
recv_from_rank,
bytes);
*/
rpointers[i] = (scalar_type *)&recv_buf_extract[i][0];
} else {
rpointers[i] = (scalar_type *)&send_buf_extract[i][0];
}
}
// Permute by swizzling pointers in merge
int permute_slice=0;
int lshift=sshift%ld;
int wrap =lshift/rd;
int num =lshift%rd;
if ( x< rd-num ) permute_slice=wrap;
else permute_slice = 1-wrap;
int toggle_bit = (Nsimd>>(permute_type+1));
int PermuteMap;
for(int i=0;i<Nsimd;i++){
if ( permute_slice ) {
PermuteMap=i^toggle_bit;
pointers[i] = rpointers[PermuteMap];
} else {
pointers[i] = rpointers[i];
}
}
Scatter_plane_merge(pointers,dimension,x,cbmask);
} else {
int permute_slice=0;
int wrap = sshift/rd;
int num = sshift%rd;
if ( x< rd-num ) permute_slice=wrap;
else permute_slice = 1-wrap;
if ( permute_slice ) Copy_plane_permute(ret,rhs,dimension,x,sx,cbmask,permute_type);
else Copy_plane(ret,rhs,dimension,x,sx,cbmask);
}
}
}
};
#endif