portelli/Grid
1
0
Fork 0
mirror of https://github.com/paboyle/Grid.git synced 2025-04-09 21:50:45 +01:00
Code Releases Activity

Mark up a table if a given site only receives from itself; including MPI3 splitting info.

Browse Source
This commit is contained in:
paboyle 2017-04-22 19:28:37 +01:00
parent b722889234
commit 3703b718aa
1 changed files with 28 additions and 15 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

43
lib/stencil/Stencil.h
View File

@ -94,7 +94,8 @@ void Gather_plane_exchange_table(std::vector<std::pair<int,int> >& table,const L
uint64_t _byte_offset; uint64_t _byte_offset;
uint16_t _is_local; uint16_t _is_local;
uint16_t _permute; uint16_t _permute;
uint32_t _around_the_world; //256 bits, 32 bytes, 1/2 cacheline uint16_t _around_the_world; //256 bits, 32 bytes, 1/2 cacheline
uint16_t _node_local;
}; };
//////////////////////////////////////// ////////////////////////////////////////
@ -170,15 +171,6 @@ class CartesianStencil { // Stencil runs along coordinate axes only; NO diagonal
int u_comm_offset; int u_comm_offset;
int _unified_buffer_size; int _unified_buffer_size;
/*
std::vector<int> compute2sites;
const std::vector<int> &Compute2Sites(void) { return compute2sites; }
void CalculateCompute2Sites(void) {
for(int i=0;i< _entries.size();i++){
}
}
*/
cobj *CommBuf(void) { return u_recv_buf_p; } cobj *CommBuf(void) { return u_recv_buf_p; }
///////////////////////////////////////// /////////////////////////////////////////
@ -198,7 +190,9 @@ class CartesianStencil { // Stencil runs along coordinate axes only; NO diagonal
//////////////////////////////////////// ////////////////////////////////////////
// Stencil query // Stencil query
//////////////////////////////////////// ////////////////////////////////////////
inline int GetNodeLocal(int osite) {
return _entries[_npoints*osite]._node_local;
}
inline StencilEntry * GetEntry(int &ptype,int point,int osite) { inline StencilEntry * GetEntry(int &ptype,int point,int osite) {
ptype = _permute_type[point]; return & _entries[point+_npoints*osite]; ptype = _permute_type[point]; return & _entries[point+_npoints*osite];
} }
@ -246,7 +240,6 @@ class CartesianStencil { // Stencil runs along coordinate axes only; NO diagonal
for(int i=0;i<Packets.size();i++){ for(int i=0;i<Packets.size();i++){
_grid->StencilSendToRecvFromComplete(reqs[i]); _grid->StencilSendToRecvFromComplete(reqs[i]);
} }
_grid->StencilBarrier();// Synch shared memory on a single nodes
commtime+=usecond(); commtime+=usecond();
} }
@ -373,8 +366,13 @@ class CartesianStencil { // Stencil runs along coordinate axes only; NO diagonal
m.buffer_size = buffer_size; m.buffer_size = buffer_size;
mv.push_back(m); mv.push_back(m);
} }
template<class decompressor> void CommsMerge(decompressor decompress) { CommsMerge(decompress,Mergers,Decompressions); } template<class decompressor> void CommsMerge(decompressor decompress) {
template<class decompressor> void CommsMergeSHM(decompressor decompress) { CommsMerge(decompress,MergersSHM,DecompressionsSHM);} CommsMerge(decompress,Mergers,Decompressions);
}
template<class decompressor> void CommsMergeSHM(decompressor decompress) {
_grid->StencilBarrier();// Synch shared memory on a single nodes
CommsMerge(decompress,MergersSHM,DecompressionsSHM);
}
template<class decompressor> template<class decompressor>
void CommsMerge(decompressor decompress,std::vector<Merge> &mm,std::vector<Decompress> &dd) { void CommsMerge(decompressor decompress,std::vector<Merge> &mm,std::vector<Decompress> &dd) {
@ -410,6 +408,18 @@ class CartesianStencil { // Stencil runs along coordinate axes only; NO diagonal
_entries[i]._byte_offset = _entries[i]._offset*sizeof(cobj); _entries[i]._byte_offset = _entries[i]._offset*sizeof(cobj);
} }
} }
int osites = _grid->oSites();
for(int o=0;o<osites;o++){
int node_local=1;
for(int p=0;p<_npoints;p++){
if (!_entries[p+o*_npoints]._is_local ){
node_local = 0;
}
}
for(int p=0;p<_npoints;p++){
_entries[p+o*_npoints]._node_local = node_local;
}
}
}; };
CartesianStencil(GridBase *grid, CartesianStencil(GridBase *grid,
@ -937,7 +947,10 @@ class CartesianStencil { // Stencil runs along coordinate axes only; NO diagonal
cobj *shm = (cobj *) _grid->ShmBufferTranslate(recv_from_rank,sp); cobj *shm = (cobj *) _grid->ShmBufferTranslate(recv_from_rank,sp);
if (shm==NULL) { if (shm==NULL) {
shm = rp; shm = rp;
shm_receive_only = 0; // we found a packet that comes from MPI and contributes to this // we found a packet that comes from MPI and contributes to this shift.
// same_node is only used in the WilsonStencil, and gets set for this point in the stencil.
// Kernel will add the exterior_terms except if same_node.
shm_receive_only = 0;
// leg of stencil // leg of stencil
} }
// if Direct, StencilSendToRecvFrom will suppress copy to a peer on node // if Direct, StencilSendToRecvFrom will suppress copy to a peer on node