Api is almost finished, new test added in order to test the user kernel interface

Addons/CKernelApi/Src/CScalfmmApi.h

@@ -362,7 +362,7 @@ typedef void (*Callback_L2L)(int level, void* parentCell, int childPosition, voi
  * @param particleIndexes indexes of particles currently computed
  * @param userData datas specific to the user's kernel
  */
-typedef void (*Callback_L2P)(void* leafCell, int nbParticles, int* particleIndexes, void* userData);
+typedef void (*Callback_L2P)(void* leafCell, int nbParticles,const int* particleIndexes, void* userData);
 
 /**
  * @brief Function to be filled by user's P2P
@@ -380,7 +380,7 @@ typedef void (*Callback_P2P)(int nbParticles, const int* particleIndexes, int nb
  * @param particleIndexes indexes of particles currently computed
  * @param userData datas specific to the user's kernel
  */
-typedef void (*Callback_P2PInner)(int nbParticles, int* particleIndexes, void* userData);
+typedef void (*Callback_P2PInner)(int nbParticles, const int* particleIndexes, void* userData);
 
 
 
@@ -434,6 +434,10 @@ typedef void* (*Callback_init_cell)(int level, long long morton_index, int* tree
 typedef void (*Callback_free_cell)(void*);
 
 
+void scalfmm_init_cell(scalfmm_handle Handle, Callback_init_cell user_cell_initializer);
+
+void scalfmm_free_cell(scalfmm_handle Handle, Callback_free_cell user_cell_deallocator);
+
 
 ///////////////// Common kernel part : /////////////////
 

Addons/CKernelApi/Src/FInterEngine.hpp

@@ -67,11 +67,12 @@ public:
      * @param BoxCenter double[3] coordinate of the center of the
      * simulation box
      */
-    FInterEngine(int TreeHeight, double BoxWidth , double * BoxCenter) : kernel(nullptr), matrix(nullptr), octree(nullptr),arranger(nullptr){
+    FInterEngine(int TreeHeight, double BoxWidth , double * BoxCenter,scalfmm_kernel_type KernelType) :
+        kernel(nullptr), matrix(nullptr), octree(nullptr),arranger(nullptr){
         octree = new OctreeClass(TreeHeight,FMath::Min(3,TreeHeight-1),BoxWidth,FPoint(BoxCenter));
         this->matrix = new MatrixKernelClass();
         this->kernel = new InterKernel(TreeHeight,BoxWidth,FPoint(BoxCenter),matrix);
-        kernelType = chebyshev;
+        kernelType = KernelType;
         arranger = nullptr;
     }
 

Addons/CKernelApi/Src/FScalFMMEngine.hpp

@@ -56,7 +56,6 @@ protected:
 
 public:
     FScalFMMEngine() : Algorithm(multi_thread), progress(), nbPart(0){
-        //Default behavior in case of out of the box particles is exiting
     }
 
     //First function displayed there are common function for every
@@ -197,6 +196,14 @@ public:
         FAssertLF(0,"No user kernel defined, exiting ...\n");
     }
 
+    virtual void init_cell(Callback_init_cell user_cell_initializer){
+        FAssertLF(0,"No user kernel defined, exiting ...\n");
+    }
+
+    virtual void free_cell(Callback_free_cell user_cell_initializer){
+        FAssertLF(0,"No user kernel defined, exiting ...\n");
+    }
+
     virtual void execute_fmm(){
         FAssertLF(0,"No kernel set, cannot execute anything, exiting ...\n");
     }
@@ -371,6 +378,13 @@ extern "C" void scalfmm_execute_fmm(scalfmm_handle Handle){
     ((ScalFmmCoreHandle * ) Handle)->engine->execute_fmm();
 }
 
+extern "C" void scalfmm_init_cell(scalfmm_handle Handle, Callback_init_cell user_cell_initializer){
+    ((ScalFmmCoreHandle * ) Handle)->engine->init_cell(user_cell_initializer);
+}
+
+extern "C" void scalfmm_free_cell(scalfmm_handle Handle, Callback_free_cell user_cell_deallocator){
+    ((ScalFmmCoreHandle * ) Handle)->engine->free_cell(user_cell_deallocator);
+}
 
 
 #endif

Addons/CKernelApi/Src/FScalfmmApiInit.cpp

@@ -8,6 +8,7 @@ extern "C" {
 }
 
 #include "FInterEngine.hpp"
+#include "FUserKernelEngine.hpp"
 
 extern "C" scalfmm_handle scalfmm_init(int TreeHeight,double BoxWidth,double* BoxCenter, scalfmm_kernel_type KernelType){
     ScalFmmCoreHandle * handle = new ScalFmmCoreHandle();
@@ -26,7 +27,7 @@ extern "C" scalfmm_handle scalfmm_init(int TreeHeight,double BoxWidth,double* Bo
         typedef FInterpMatrixKernelR                                        MatrixKernelClass;
         typedef FChebSymKernel<ChebCell,ContainerClass,MatrixKernelClass,7>        ChebKernel;
 
-        handle->engine = new FInterEngine<ChebCell,ChebKernel>(TreeHeight,BoxWidth,BoxCenter);
+        handle->engine = new FInterEngine<ChebCell,ChebKernel>(TreeHeight,BoxWidth,BoxCenter, KernelType);
         break;
     case 2:
         //TODO typedefs
@@ -36,7 +37,7 @@ extern "C" scalfmm_handle scalfmm_init(int TreeHeight,double BoxWidth,double* Bo
         typedef FInterpMatrixKernelR                                        MatrixKernelClass;
         typedef FUnifKernel<UnifCell,ContainerClass,MatrixKernelClass,7>           UnifKernel;
 
-        handle->engine = new FInterEngine<UnifCell,UnifKernel>(TreeHeight,BoxWidth,BoxCenter);
+        handle->engine = new FInterEngine<UnifCell,UnifKernel>(TreeHeight,BoxWidth,BoxCenter,KernelType);
         break;
 
     default:

Addons/CKernelApi/Src/FUserKernelEngine.hpp

+// ===================================================================================
+// Copyright ScalFmm 2014 I
+// This software is a computer program whose purpose is to compute the FMM.
+//
+// This software is governed by the CeCILL-C and LGPL licenses and
+// abiding by the rules of distribution of free software.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU General Public and CeCILL-C Licenses for more details.
+// "http://www.cecill.info".
+// "http://www.gnu.org/licenses".
+// ===================================================================================
+
+
+/**
+ * @file This file contains a class that inherits from FScalFMMEngine,
+ * and will implement the API functions for a user defined kernel.
+ */
+#ifndef FUSERKERNELENGINE_HPP
+#define FUSERKERNELENGINE_HPP
+
+#include "FScalFMMEngine.hpp"
+
+
+/**
+ * @brief CoreCell : Cell used to store User datas
+ */
+class CoreCell : public FBasicCell {
+    // Mutable in order to work with the API
+    mutable void* userData;
+
+public:
+    CoreCell() : userData(nullptr) {
+    }
+
+    ~CoreCell(){
+    }
+    /**
+     * @brief setContainer store the ptr to the user data inside our
+     * struct
+     */
+    void setContainer(void* inContainer) const {
+        userData = inContainer;
+    }
+
+    /**
+     * @brief getContainer : return the user datas (in order to give
+     * it back to the user defined kernel function)
+     */
+    void* getContainer() const {
+        return userData;
+    }
+};
+
+
+/**
+ * This class simply call the function pointers from Scalfmm_Kernel_Descriptor.
+ * If not pointer is set the calls are skipped.
+ * The userData is given at any calls.
+ */
+template< class CellClass, class ContainerClass>
+class CoreKernel : public FAbstractKernels<CellClass,ContainerClass> {
+    Scalfmm_Kernel_Descriptor kernel;
+    void* userData;
+
+public:
+    CoreKernel(Scalfmm_Kernel_Descriptor inKernel, void* inUserData) : kernel(inKernel) , userData(inUserData){
+    }
+
+    /** Default destructor */
+    virtual ~CoreKernel(){
+    }
+
+    /** Do nothing */
+    virtual void P2M(CellClass* const cell, const ContainerClass* const container) {
+        if(kernel.p2m) kernel.p2m(cell->getContainer(), container->getNbParticles(), container->getIndexes().data(), userData);
+    }
+
+    /** Do nothing */
+    virtual void M2M(CellClass* const FRestrict cell, const CellClass*const FRestrict *const FRestrict children, const int level) {
+        if(kernel.m2m){
+            for(int idx = 0 ; idx < 8 ; ++idx){
+                if( children[idx] ){
+                    kernel.m2m(level, cell->getContainer(), idx, children[idx]->getContainer(), userData);
+                }
+            }
+        }
+    }
+
+    /** Do nothing */
+    virtual void M2L(CellClass* const FRestrict cell, const CellClass* interactions[], const int , const int level) {
+        if(kernel.m2l){
+            for(int idx = 0 ; idx < 343 ; ++idx){
+                if( interactions[idx] ){
+                    kernel.m2l(level, cell->getContainer(), idx, interactions[idx]->getContainer(), userData);
+                }
+            }
+        }
+    }
+
+    /** Do nothing */
+    virtual void L2L(const CellClass* const FRestrict cell, CellClass* FRestrict *const FRestrict children, const int level) {
+        if(kernel.l2l){
+            for(int idx = 0 ; idx < 8 ; ++idx){
+                if( children[idx] ){
+                    kernel.l2l(level, cell->getContainer(), idx, children[idx]->getContainer(), userData);
+                }
+            }
+        }
+    }
+
+    /** Do nothing */
+    virtual void L2P(const CellClass* const cell, ContainerClass* const container){
+        if(kernel.l2p) kernel.l2p(cell->getContainer(), container->getNbParticles(), container->getIndexes().data(), userData);
+    }
+
+
+    /** Do nothing */
+    virtual void P2P(const FTreeCoordinate& ,
+                     ContainerClass* const FRestrict targets, const ContainerClass* const FRestrict /*sources*/,
+                     ContainerClass* const neighbors[27], const int ){
+        if(kernel.p2pinner) kernel.p2pinner(targets->getNbParticles(), targets->getIndexes().data(), userData);
+
+        if(kernel.p2p){
+            for(int idx = 0 ; idx < 27 ; ++idx){
+                if( neighbors[idx] ){
+                    kernel.p2p(targets->getNbParticles(), targets->getIndexes().data(),
+                                    neighbors[idx]->getNbParticles(), neighbors[idx]->getIndexes().data(), userData);
+                }
+            }
+        }
+    }
+
+    /** Do nothing */
+    virtual void P2PRemote(const FTreeCoordinate& ,
+                     ContainerClass* const FRestrict , const ContainerClass* const FRestrict ,
+                     ContainerClass* const [27], const int ){
+    }
+
+};
+
+
+
+
+class FUserKernelEngine : public FScalFMMEngine{
+
+private:
+
+    //Typedefs :
+    typedef FP2PParticleContainerIndexed<>           ContainerClass;
+    typedef FSimpleLeaf<ContainerClass>                   LeafClass;
+    typedef FOctree<CoreCell,ContainerClass,LeafClass>  OctreeClass;
+    typedef CoreKernel<CoreCell,ContainerClass>     CoreKernelClass;
+    typedef FP2PLeafInterface<OctreeClass>            LeafInterface;
+    //For arranger classes
+    typedef FOctreeArranger<OctreeClass,ContainerClass,LeafInterface>           ArrangerClass;
+    typedef FArrangerPeriodic<OctreeClass,ContainerClass,LeafInterface> ArrangerClassPeriodic;
+
+
+    //Attributes
+    OctreeClass * octree;
+    CoreKernelClass * kernel;
+    ArrangerClass * arranger;
+
+public:
+    FUserKernelEngine(int TreeHeight, double BoxWidth , double * BoxCenter, scalfmm_kernel_type KernelType) :
+        octree(nullptr), kernel(nullptr), arranger(nullptr){
+        octree = new OctreeClass(TreeHeight,FMath::Min(3,TreeHeight-1),BoxWidth,FPoint(BoxCenter));
+        kernelType = KernelType;
+        //Kernel is not set now because the user must provide a
+        //Scalfmm_Kernel_descriptor
+        //kernel = new CoreKernelClass();
+    }
+
+
+    void user_kernel_config( Scalfmm_Kernel_Descriptor userKernel, void * userDatas){
+        kernel = new CoreKernelClass(userKernel,userDatas);
+    }
+
+
+    void tree_insert_particles( int NbPositions, double * arrayX, double * arrayY, double * arrayZ){
+        for(int idPart = 0; idPart<NbPositions ; ++idPart){
+            octree->insert(FPoint(arrayX[idPart],arrayY[idPart],arrayZ[idPart]),idPart);
+        }
+        nbPart += NbPositions;
+    }
+
+    void tree_insert_particles_xyz( int NbPositions, double * XYZ){
+        for(int idPart = 0; idPart<NbPositions ; ++idPart){
+            octree->insert(FPoint(&XYZ[3*idPart]),idPart);
+        }
+        nbPart += NbPositions;
+    }
+
+    /*
+     * Call the user allocator on userDatas member field of each cell
+     */
+    void init_cell(Callback_init_cell user_cell_initializer){
+        if(user_cell_initializer){
+            double boxwidth = octree->getBoxWidth();
+            FPoint BoxCenter = octree->getBoxCenter();
+            double boxCorner[3];
+            boxCorner[0] = BoxCenter.getX() - boxwidth/2.0;
+            boxCorner[1] = BoxCenter.getY() - boxwidth/2.0;
+            boxCorner[2] = BoxCenter.getZ() - boxwidth/2.0;
+            //apply user function on each cell
+            octree->forEachCellWithLevel([&](CoreCell * currCell,const int currLevel){
+                    FTreeCoordinate currCoord = currCell->getCoordinate();
+                    int arrayCoord[3] = {currCoord.getX(),currCoord.getY(),currCoord.getZ()};
+                    MortonIndex    currMorton = currCoord.getMortonIndex(currLevel);
+                    double position[3];
+                    position[0] = boxCorner[0] + currCoord.getX()*boxwidth/double(1<<currLevel);
+                    position[1] = boxCorner[1] + currCoord.getY()*boxwidth/double(1<<currLevel);
+                    position[2] = boxCorner[2] + currCoord.getZ()*boxwidth/double(1<<currLevel);
+                    currCell->setContainer(user_cell_initializer(currLevel,currMorton,arrayCoord,position));
+                });
+        }
+    }
+
+    void free_cell(Callback_free_cell user_cell_deallocator){
+        octree->forEachCell([&](CoreCell * currCell){
+                user_cell_deallocator(currCell->getContainer());
+            });
+    }
+
+    void execute_fmm(){
+        //FAssertLF(kernel,"No kernel set, please use scalfmm_user_kernel_config before calling the execute routine ... Exiting \n");
+        switch(Algorithm){
+        case 0:
+            {
+                typedef FFmmAlgorithm<OctreeClass,CoreCell,ContainerClass,CoreKernelClass,LeafInterface> AlgoClassSeq;
+                AlgoClassSeq algoSeq(octree,kernel);
+                algoSeq.execute();
+                break;
+            }
+        case 1:
+            {
+                typedef FFmmAlgorithmThread<OctreeClass,CoreCell,ContainerClass,CoreKernelClass,LeafClass> AlgoClassThread;
+                AlgoClassThread algoThread(octree,kernel);
+                algoThread.execute();
+                break;
+            }
+        case 2:
+            {
+                typedef FFmmAlgorithmPeriodic<OctreeClass,CoreCell,ContainerClass,CoreKernelClass,LeafClass> AlgoClassPeriodic;
+                AlgoClassPeriodic algoPeriod(octree,2);
+                algoPeriod.setKernel(kernel);
+                algoPeriod.execute();
+                break;
+            }
+        default :
+            std::cout<< "No algorithm found (probably for strange reasons) : "<< Algorithm <<" exiting" << std::endl;
+        }
+
+    }
+
+};
+
+
+#endif //FUSERKERNELENGINE_HPP

Addons/CKernelApi/Tests/testUserDefinedKernelApi.c

+// ===================================================================================
+// Logiciel initial: ScalFmm Version 0.5
+// Co-auteurs : Olivier Coulaud, Bérenger Bramas.
+// Propriétaires : INRIA.
+// Copyright © 2011-2012, diffusé sous les termes et conditions d’une licence propriétaire.
+// Initial software: ScalFmm Version 0.5
+// Co-authors: Olivier Coulaud, Bérenger Bramas.
+// Owners: INRIA.
+// Copyright © 2011-2012, spread under the terms and conditions of a proprietary license.
+// ===================================================================================
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+
+/**
+  * In this file we implement a example kernel which is simply printing information
+  * about the cells and particles.
+  * It is recommanded to compile and execute this code in order to understand the behavior
+  * of the application.
+  * We mark some part with "JUST-PUT-HERE:" to give instruction to user to create its own kernel.
+  **/
+
+// Include the FMM API (should be in a different folder for you)
+#include "../Src/CScalfmmApi.h"
+
+// Uncomment the next line to avoid the print in the kernel and verbosity
+// #define NOT_TOO_MUCH_VERBOSE
+#ifdef NOT_TOO_MUCH_VERBOSE
+#define VerbosePrint(X...)
+#else
+#define VerbosePrint(X...) printf(X)
+#endif
+
+///////////////////////////////////////////////////////////////////////////
+/// Cell struct and functions
+///////////////////////////////////////////////////////////////////////////
+
+// This represent a cell
+struct MyCellDescriptor{
+    int level;
+    long long mortonIndex;
+    int coord[3];
+    double position[3];
+    // JUST-PUT-HERE:
+    // You local and multipole arrays
+};
+
+// This is our function that init a cell (struct MyCellDescriptor)
+void* my_Callback_init_cell(int level, long long mortonIndex, int* coord, double* position){
+    VerbosePrint("\tAllocating cell for level %d, morton index %lld, coord %d/%d/%d\n", level, mortonIndex, coord[0], coord[1], coord[2]);
+    struct MyCellDescriptor* cellData = (struct MyCellDescriptor*)malloc(sizeof(struct MyCellDescriptor));
+    memset(cellData, 0, sizeof(struct MyCellDescriptor));
+    cellData->level       = level;
+    cellData->mortonIndex = mortonIndex;
+    memcpy(cellData->coord, coord, sizeof(int)*3);
+    memcpy(cellData->position, position, sizeof(double)*3);
+    // JUST-PUT-HERE:
+    // Fill your structure
+    return cellData;
+}
+
+// To dealloc a cell
+void my_Callback_free_cell(void* cellData){
+    free(cellData);
+}
+
+
+///////////////////////////////////////////////////////////////////////////
+/// Kernel struct and functions
+///////////////////////////////////////////////////////////////////////////
+
+// This is the data passed to our kernel
+struct MyData {
+    // We simply strore the number of call the and particle indexes
+    double* insertedPositions;
+    int countP2M;
+    int countM2M;
+    int countM2L;
+    int countL2L;
+    int countL2P;
+    int countP2PInner;
+    int countP2P;
+    // JUST-PUT-HERE:
+    // everything your kernel will need for its computation
+    // pre-computed matrices, etc....
+};
+
+
+// Our P2M
+void my_Callback_P2M(void* cellData, int nbParticlesInLeaf, const int* particleIndexes, void* userData){
+    struct MyData* my_data = (struct MyData*)userData;
+    my_data->countP2M += 1;
+
+    struct MyCellDescriptor* my_cell = (struct MyCellDescriptor*) cellData;
+    VerbosePrint("Cell morton %lld is doing P2M with %d particles :\n", my_cell->mortonIndex, nbParticlesInLeaf);
+    int idxPart;
+    for(idxPart = 0 ; idxPart < nbParticlesInLeaf ; ++idxPart){
+        double* position = &my_data->insertedPositions[particleIndexes[idxPart]*3];
+        VerbosePrint("\t particle idx %d position %e/%e%e\n", particleIndexes[idxPart],
+               position[0], position[1], position[2]);
+        // JUST-PUT-HERE:
+        // Your real P2M computation
+    }
+}
+
+void my_Callback_M2M(int level, void* cellData, int childPosition, void* childData, void* userData){
+    struct MyData* my_data = (struct MyData*)userData;
+    my_data->countM2M += 1;
+
+    struct MyCellDescriptor* my_cell = (struct MyCellDescriptor*) cellData;
+    struct MyCellDescriptor* my_child = (struct MyCellDescriptor*) childData;
+
+    int childFullPosition[3];
+    Scalfmm_utils_parentChildPosition(childPosition, childFullPosition);
+
+    VerbosePrint("Doing a M2M at level %d, between cells %lld and %lld (position %d/%d/%d)\n",
+           level, my_cell->mortonIndex, my_child->mortonIndex,
+           childFullPosition[0], childFullPosition[1], childFullPosition[2]);
+    // JUST-PUT-HERE: your real M2M computation
+}
+
+void my_Callback_M2L(int level, void* cellData, int srcPosition, void* srcData, void* userData){
+    struct MyData* my_data = (struct MyData*)userData;
+    my_data->countM2L += 1;
+
+    struct MyCellDescriptor* my_cell = (struct MyCellDescriptor*) cellData;
+    struct MyCellDescriptor* my_src_cell = (struct MyCellDescriptor*) srcData;
+
+    int interactionFullPosition[3];
+    Scalfmm_utils_interactionPosition(srcPosition, interactionFullPosition);
+
+    VerbosePrint("Doing a M2L at level %d, between cells %lld and %lld (position %d/%d/%d)\n",
+           level, my_cell->mortonIndex, my_src_cell->mortonIndex,
+           interactionFullPosition[0], interactionFullPosition[1], interactionFullPosition[2]);
+    // JUST-PUT-HERE: Your M2L
+}
+
+void my_Callback_L2L(int level, void* cellData, int childPosition, void* childData, void* userData){
+    struct MyData* my_data = (struct MyData*)userData;
+    my_data->countL2L += 1;
+
+    struct MyCellDescriptor* my_cell = (struct MyCellDescriptor*) cellData;
+    struct MyCellDescriptor* my_child = (struct MyCellDescriptor*) childData;
+
+    int childFullPosition[3];
+    Scalfmm_utils_parentChildPosition(childPosition, childFullPosition);
+
+    VerbosePrint("Doing a L2L at level %d, between cells %lld and %lld (position %d/%d/%d)\n",
+           level, my_cell->mortonIndex, my_child->mortonIndex,
+           childFullPosition[0], childFullPosition[1], childFullPosition[2]);
+    // JUST-PUT-HERE: Your L2L
+}
+
+void my_Callback_L2P(void* cellData, int nbParticlesInLeaf, const int* particleIndexes, void* userData){
+    struct MyData* my_data = (struct MyData*)userData;
+    my_data->countL2P += 1;
+
+    struct MyCellDescriptor* my_cell = (struct MyCellDescriptor*) cellData;
+    VerbosePrint("Cell morton %lld is doing L2P with %d particles :\n", my_cell->mortonIndex, nbParticlesInLeaf);
+    int idxPart;
+    for(idxPart = 0 ; idxPart < nbParticlesInLeaf ; ++idxPart){
+        double* position = &my_data->insertedPositions[particleIndexes[idxPart]*3];
+        VerbosePrint("\t particle idx %d position %e/%e%e\n", particleIndexes[idxPart],
+               position[0], position[1], position[2]);
+        // JUST-PUT-HERE: Your L2P
+    }
+}
+
+void my_Callback_P2P(int nbParticlesInLeaf, const int* particleIndexes, int nbParticlesInSrc, const int* particleIndexesSrc, void* userData){
+    struct MyData* my_data = (struct MyData*)userData;
+    my_data->countP2P += 1;
+
+    VerbosePrint("Doing P2P between two leaves of %d particles and %d particles :\n", nbParticlesInLeaf, nbParticlesInSrc);
+    int idxPart;
+    for(idxPart = 0 ; idxPart < nbParticlesInLeaf ; ++idxPart){
+        double* position = &my_data->insertedPositions[particleIndexes[idxPart]*3];
+        VerbosePrint("\t Target >> particle idx %d position %e/%e%e\n", particleIndexes[idxPart],
+               position[0], position[1], position[2]);
+    }
+    for(idxPart = 0 ; idxPart < nbParticlesInSrc ; ++idxPart){
+        double* position = &my_data->insertedPositions[particleIndexesSrc[idxPart]*3];
+        VerbosePrint("\t Target >> Src idx %d position %e/%e%e\n", particleIndexesSrc[idxPart],
+               position[0], position[1], position[2]);
+    }
+
+    // JUST-PUT-HERE:
+    // Put one loop into the other to make all particles from source
+    // interacting with the target particles
+}
+
+void my_Callback_P2PInner(int nbParticlesInLeaf, const int* particleIndexes, void* userData){
+    struct MyData* my_data = (struct MyData*)userData;
+    my_data->countP2PInner += 1;
+
+    VerbosePrint("Doing P2P inside a leaf of %d particles :\n", nbParticlesInLeaf);
+    int idxPart;
+    for(idxPart = 0 ; idxPart < nbParticlesInLeaf ; ++idxPart){
+        double* position = &my_data->insertedPositions[particleIndexes[idxPart]*3];
+        VerbosePrint("\t particle idx %d position %e/%e%e\n", particleIndexes[idxPart],
+               position[0], position[1], position[2]);
+        // JUST-PUT-HERE: another loop to have all particles interacting with
+        // the other from the same leaf
+    }
+}
+
+///////////////////////////////////////////////////////////////////////////
+/// Main
+///////////////////////////////////////////////////////////////////////////
+
+// Simply create particles and try the kernels
+int main(int argc, char ** argv){
+    // The properties of our tree
+    int treeHeight = 4;
+    double boxWidth = 1.0;
+    double boxCenter[3];
+    boxCenter[0] = boxCenter[1] = boxCenter[2] = 0.0;
+
+    // Create random particles
+    int nbParticles = 10;
+    int particleIndexes[nbParticles];
+    double particleXYZ[nbParticles*3];
+    {
+        printf("Creating Particles:\n");
+        int idxPart;
+        for(idxPart = 0 ; idxPart < nbParticles ; ++idxPart){
+            particl