portelli/Grid
1
0
Fork 0
mirror of https://github.com/paboyle/Grid.git synced 2025-06-16 06:47:06 +01:00
Code Releases Activity

Threading support rework.

Browse Source 
Placed parallel pragmas as macros; implemented deterministic thread reduction in style of
BFM.
This commit is contained in:
Peter Boyle
2015-05-12 07:51:41 +01:00
parent b1d2c60d07
commit 6103c29ee3
26 changed files with 276 additions and 184 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
lib/lattice/Grid_lattice_local.h
View File

@ -16,7 +16,7 @@ namespace Grid {
inline auto localNorm2 (const Lattice<vobj> &rhs)-> Lattice<typename vobj::tensor_reduced>
{
Lattice<typename vobj::tensor_reduced> ret(rhs._grid);
#pragma omp parallel for
PARALLEL_FOR_LOOP
for(int ss=0;ss<rhs._grid->oSites(); ss++){
ret._odata[ss]=innerProduct(rhs._odata[ss],rhs._odata[ss]);
}
@ -29,7 +29,7 @@ namespace Grid {
-> Lattice<typename vobj::tensor_reduced>
{
Lattice<typename vobj::tensor_reduced> ret(rhs._grid);
#pragma omp parallel for
PARALLEL_FOR_LOOP
for(int ss=0;ss<rhs._grid->oSites(); ss++){
ret._odata[ss]=innerProduct(lhs._odata[ss],rhs._odata[ss]);
}
@ -42,7 +42,7 @@ namespace Grid {
inline auto outerProduct (const Lattice<ll> &lhs,const Lattice<rr> &rhs) -> Lattice<decltype(outerProduct(lhs._odata[0],rhs._odata[0]))>
{
Lattice<decltype(outerProduct(lhs._odata[0],rhs._odata[0]))> ret(rhs._grid);
#pragma omp parallel for
PARALLEL_FOR_LOOP
for(int ss=0;ss<rhs._grid->oSites(); ss++){
ret._odata[ss]=outerProduct(lhs._odata[ss],rhs._odata[ss]);
}