git-svn-id: svn://scm.gforge.inria.fr/svnroot/paradiseo@2083 331e1502-861f-0410-8da2-ba01fb791d7f
This commit is contained in:
parent
79a0a85863
commit
0ab74e3bb2
1 changed files with 12 additions and 9 deletions
|
|
@ -38,7 +38,7 @@
|
||||||
#include <eval/moCudakernelEval.h>
|
#include <eval/moCudakernelEval.h>
|
||||||
|
|
||||||
/**
|
/**
|
||||||
* class for the cuda evaluation
|
* class for the K-swap neighborhood evaluation
|
||||||
*/
|
*/
|
||||||
|
|
||||||
template<class Neighbor, class IncrementEval>
|
template<class Neighbor, class IncrementEval>
|
||||||
|
|
@ -75,7 +75,6 @@ public:
|
||||||
moCudaEval<Neighbor> (_neighborhoodSize), incrEval(_incrEval) {
|
moCudaEval<Neighbor> (_neighborhoodSize), incrEval(_incrEval) {
|
||||||
mutex = false;
|
mutex = false;
|
||||||
mutex_kswap=false;
|
mutex_kswap=false;
|
||||||
compt=0;
|
|
||||||
}
|
}
|
||||||
/**
|
/**
|
||||||
* Destructor
|
* Destructor
|
||||||
|
|
@ -99,7 +98,7 @@ public:
|
||||||
/**
|
/**
|
||||||
* Compute fitness for all solution neighbors in device
|
* Compute fitness for all solution neighbors in device
|
||||||
* @param _sol the solution which generate the neighborhood
|
* @param _sol the solution which generate the neighborhood
|
||||||
* @param _mapping the array of indices mapping
|
* @param _mapping the array of mapping indexes for K-swap neighborhood
|
||||||
* @param _Kswap the number of swap
|
* @param _Kswap the number of swap
|
||||||
*/
|
*/
|
||||||
|
|
||||||
|
|
@ -110,6 +109,7 @@ public:
|
||||||
|
|
||||||
// Get Current solution fitness
|
// Get Current solution fitness
|
||||||
Fitness fitness = _sol.fitness();
|
Fitness fitness = _sol.fitness();
|
||||||
|
|
||||||
//Case of Permutation
|
//Case of Permutation
|
||||||
if (_Kswap == 1) {
|
if (_Kswap == 1) {
|
||||||
if (!mutex) {
|
if (!mutex) {
|
||||||
|
|
@ -117,6 +117,7 @@ public:
|
||||||
cudaMalloc((void**) &device_solution.vect, _size * sizeof(T));
|
cudaMalloc((void**) &device_solution.vect, _size * sizeof(T));
|
||||||
mutex = true;
|
mutex = true;
|
||||||
}
|
}
|
||||||
|
|
||||||
//Copy the solution vector from the host to device
|
//Copy the solution vector from the host to device
|
||||||
cudaMemcpy(device_solution.vect, _sol.vect, _size * sizeof(T),
|
cudaMemcpy(device_solution.vect, _sol.vect, _size * sizeof(T),
|
||||||
cudaMemcpyHostToDevice);
|
cudaMemcpyHostToDevice);
|
||||||
|
|
@ -130,6 +131,7 @@ public:
|
||||||
//Case Kswap
|
//Case Kswap
|
||||||
else if (_Kswap > 1) {
|
else if (_Kswap > 1) {
|
||||||
if (!mutex_kswap) {
|
if (!mutex_kswap) {
|
||||||
|
|
||||||
vect = new T[neighborhoodSize * _size];
|
vect = new T[neighborhoodSize * _size];
|
||||||
//Allocate the space for set of solution in the device global memory
|
//Allocate the space for set of solution in the device global memory
|
||||||
cudaMalloc((void**) &device_setSolution.vect, neighborhoodSize
|
cudaMalloc((void**) &device_setSolution.vect, neighborhoodSize
|
||||||
|
|
@ -138,6 +140,7 @@ public:
|
||||||
cudaMalloc((void**) &device_tmp.vect, neighborhoodSize
|
cudaMalloc((void**) &device_tmp.vect, neighborhoodSize
|
||||||
* sizeof(T));
|
* sizeof(T));
|
||||||
mutex_kswap = true;
|
mutex_kswap = true;
|
||||||
|
|
||||||
}
|
}
|
||||||
|
|
||||||
for (int i = 0; i < neighborhoodSize; i++) {
|
for (int i = 0; i < neighborhoodSize; i++) {
|
||||||
|
|
@ -145,19 +148,17 @@ public:
|
||||||
vect[j + i * _size] = _sol.vect[j];
|
vect[j + i * _size] = _sol.vect[j];
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
//Copy the set of solution from the host to device
|
//Copy the set of solution from the host to device
|
||||||
cudaMemcpy(device_setSolution.vect, vect, neighborhoodSize * _size
|
cudaMemcpy(device_setSolution.vect, vect, neighborhoodSize * _size
|
||||||
* sizeof(T), cudaMemcpyHostToDevice);
|
* sizeof(T), cudaMemcpyHostToDevice);
|
||||||
|
|
||||||
//Launch the Kernel to compute all Kswap neighbors fitness
|
//Launch the Kernel to compute all Kswap neighbors fitness
|
||||||
kernelKswap<EOT,Fitness,Neighbor,IncrementEval><<<kernel_Dim,BLOCK_SIZE >>>(incrEval,device_setSolution,device_tmp,device_FitnessArray,fitness,neighborhoodSize,_mapping,_Kswap,_size);
|
kernelKswap<EOT,Fitness,Neighbor,IncrementEval><<<kernel_Dim,BLOCK_SIZE >>>(incrEval,device_setSolution,device_tmp,device_FitnessArray,fitness,neighborhoodSize,_mapping,_Kswap,_size);
|
||||||
|
|
||||||
//Copy the result from device to host
|
//Copy the result from device to host
|
||||||
cudaMemcpy(host_FitnessArray, device_FitnessArray, neighborhoodSize
|
cudaMemcpy(host_FitnessArray, device_FitnessArray, neighborhoodSize
|
||||||
* sizeof(Fitness), cudaMemcpyDeviceToHost);
|
* sizeof(Fitness), cudaMemcpyDeviceToHost);
|
||||||
std::cout<<" Iteration number "<<compt<<" "<<std::endl;
|
|
||||||
/* for(unsigned int i=0;i<neighborhoodSize;i++)
|
|
||||||
std::cout<<" "<<host_FitnessArray[i]<<" ";//<<std::endl;*/
|
|
||||||
compt++;
|
|
||||||
}
|
}
|
||||||
|
|
||||||
}
|
}
|
||||||
|
|
@ -165,7 +166,7 @@ public:
|
||||||
/**
|
/**
|
||||||
* Compute fitness for all solution neighbors(K-flip of binary solution) in device
|
* Compute fitness for all solution neighbors(K-flip of binary solution) in device
|
||||||
* @param _sol the solution which generate the neighborhood
|
* @param _sol the solution which generate the neighborhood
|
||||||
* @param _mapping the array of indices mapping
|
* @param _mapping the array of mapping indexes for k-flip neighborhood
|
||||||
* @param _Kflip the number of flip to do
|
* @param _Kflip the number of flip to do
|
||||||
*/
|
*/
|
||||||
|
|
||||||
|
|
@ -176,17 +177,20 @@ public:
|
||||||
|
|
||||||
// Get Current solution fitness
|
// Get Current solution fitness
|
||||||
Fitness fitness = _sol.fitness();
|
Fitness fitness = _sol.fitness();
|
||||||
|
|
||||||
if (!mutex) {
|
if (!mutex) {
|
||||||
//Allocate the space for solution in the device global memory
|
//Allocate the space for solution in the device global memory
|
||||||
cudaMalloc((void**) &device_solution.vect, _size * sizeof(T));
|
cudaMalloc((void**) &device_solution.vect, _size * sizeof(T));
|
||||||
mutex = true;
|
mutex = true;
|
||||||
}
|
}
|
||||||
|
|
||||||
//Copy the solution vector from the host to device
|
//Copy the solution vector from the host to device
|
||||||
cudaMemcpy(device_solution.vect, _sol.vect, _size * sizeof(T),
|
cudaMemcpy(device_solution.vect, _sol.vect, _size * sizeof(T),
|
||||||
cudaMemcpyHostToDevice);
|
cudaMemcpyHostToDevice);
|
||||||
|
|
||||||
//Launch the Kernel to compute all flip neighbors fitness
|
//Launch the Kernel to compute all flip neighbors fitness
|
||||||
kernelKflip<EOT,Fitness,Neighbor,IncrementEval><<<kernel_Dim,BLOCK_SIZE >>>(incrEval,device_solution,device_FitnessArray,fitness,neighborhoodSize,_mapping,_Kflip);
|
kernelKflip<EOT,Fitness,Neighbor,IncrementEval><<<kernel_Dim,BLOCK_SIZE >>>(incrEval,device_solution,device_FitnessArray,fitness,neighborhoodSize,_mapping,_Kflip);
|
||||||
|
|
||||||
//Copy the result from device to host
|
//Copy the result from device to host
|
||||||
cudaMemcpy(host_FitnessArray, device_FitnessArray, neighborhoodSize
|
cudaMemcpy(host_FitnessArray, device_FitnessArray, neighborhoodSize
|
||||||
* sizeof(Fitness), cudaMemcpyDeviceToHost);
|
* sizeof(Fitness), cudaMemcpyDeviceToHost);
|
||||||
|
|
@ -203,7 +207,6 @@ protected:
|
||||||
T * vect;
|
T * vect;
|
||||||
bool mutex_kswap;
|
bool mutex_kswap;
|
||||||
bool mutex;
|
bool mutex;
|
||||||
unsigned int compt;
|
|
||||||
};
|
};
|
||||||
|
|
||||||
#endif
|
#endif
|
||||||
|
|
|
||||||
Loading…
Add table
Add a link
Reference in a new issue