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

Coalesced write

Browse Source
This commit is contained in:
Peter Boyle 2019-07-02 17:25:13 +01:00
parent 4c3225412b
commit d6ffadb33b
1 changed files with 10 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

10
Grid/tensors/Tensor_SIMT.h
View File

@ -69,6 +69,11 @@ void coalescedWrite(vobj & __restrict__ vec,const vobj & __restrict__ extracted,
// vstream(vec, extracted); // vstream(vec, extracted);
vec = extracted; vec = extracted;
} }
template<class vobj> accelerator_inline
void coalescedWriteNonTemporal(vobj & __restrict__ vec,const vobj & __restrict__ extracted,int lane=0)
{
vstream(vec, extracted);
}
#else #else
accelerator_inline int SIMTlane(int Nsimd) { return threadIdx.y; } // CUDA specific accelerator_inline int SIMTlane(int Nsimd) { return threadIdx.y; } // CUDA specific
@ -92,6 +97,11 @@ void coalescedWrite(vobj & __restrict__ vec,const typename vobj::scalar_object &
{ {
insertLane(lane,vec,extracted); insertLane(lane,vec,extracted);
} }
template<class vobj> accelerator_inline
void coalescedWriteNonTemporal(vobj & __restrict__ vec,const vobj & __restrict__ extracted,int lane=0)
{
insertLane(lane,vec,extracted);
}
#endif #endif