Moving to mgpu scan

src/Vector/cuda/vector_dist_comm_util_funcs.cuh

@@ -11,11 +11,10 @@
 #define SKIP_LABELLING 512
 #define KEEP_PROPERTIES 512
 
-template<unsigned int dim, typename St, typename prop, typename Memory, template<typename> class layout_base, typename Decomposition, typename scan_type, bool is_ok_cuda>
+template<unsigned int dim, typename St, typename prop, typename Memory, template<typename> class layout_base, typename Decomposition, bool is_ok_cuda>
 struct labelParticlesGhost_impl
 {
 	static void run(CudaMemory & mem,
-					scan_type & sc,
 					Decomposition & dec,
 					openfpm::vector<aggregate<unsigned int,unsigned long int>,
 							CudaMemory,
@@ -44,11 +43,10 @@ struct labelParticlesGhost_impl
 
 
 
-template<unsigned int dim, typename St, typename prop, typename Memory, template<typename> class layout_base, typename Decomposition, typename scan_type>
-struct labelParticlesGhost_impl<dim,St,prop,Memory,layout_base,Decomposition,scan_type,true>
+template<unsigned int dim, typename St, typename prop, typename Memory, template<typename> class layout_base, typename Decomposition>
+struct labelParticlesGhost_impl<dim,St,prop,Memory,layout_base,Decomposition,true>
 {
 	static void run(CudaMemory & mem,
-					scan_type & sc,
 					Decomposition & dec,
 					openfpm::vector<aggregate<unsigned int,unsigned long int>,
 							CudaMemory,
@@ -92,8 +90,9 @@ struct labelParticlesGhost_impl<dim,St,prop,Memory,layout_base,Decomposition,sca
 			dec.toKernel(),v_pos.toKernel(),proc_id_out.toKernel());
 
 			// scan
-			sc.scan_(proc_id_out,starts);
+			//sc.scan_(proc_id_out,starts);
 			starts.resize(proc_id_out.size());
+			mgpu::scan((unsigned int *)proc_id_out.template getDeviceBuffer<0>(), proc_id_out.size(), (unsigned int *)starts.template getDeviceBuffer<0>() , v_cl.getmgpuContext());
 			starts.template deviceToHost<0>(starts.size()-1,starts.size()-1);
 			size_t sz = starts.template get<0>(starts.size()-1);
 

src/Vector/vector_dist_comm.hpp

@@ -16,7 +16,6 @@
 #include "util/cuda/moderngpu/kernel_scan.hxx"
 #endif
 
-#include "util/cuda/scan_cuda.cuh"
 #include "Vector/util/vector_dist_funcs.hpp"
 #include "cuda/vector_dist_comm_util_funcs.cuh"
 
@@ -87,9 +86,6 @@ class vector_dist_comm
 	//! It map the processor id with the communication request into map procedure
 	openfpm::vector<size_t> p_map_req;
 
-	//! scan functionality required for gpu
-	scan<unsigned int,unsigned int> sc;
-
 	//! For each near processor, outgoing particle id
 	//! \warning opart is assumed to be an ordered list
 	//! first id particle id
@@ -1240,8 +1236,8 @@ class vector_dist_comm
 		if (opt & RUN_ON_DEVICE)
 		{
 			labelParticlesGhost_impl<dim,St,prop,Memory,layout_base,
-			                         Decomposition,scan<unsigned int,unsigned int>,std::is_same<Memory,CudaMemory>::value>
-			::run(mem,sc,dec,g_opart_device,proc_id_out,starts,v_cl,v_pos,v_prp,prc,prc_sz,prc_offset,g_m,opt);
+			                         Decomposition,std::is_same<Memory,CudaMemory>::value>
+			::run(mem,dec,g_opart_device,proc_id_out,starts,v_cl,v_pos,v_prp,prc,prc_sz,prc_offset,g_m,opt);
 		}
 		else
 		{