generateA.cu

Go to the documentation of this file.

#include "generateA.h"


namespace kernels
{

__global__
void generateA(int *ARows, int *ACols, real *AVals,
               real *MVals,
               int *LRows, int *LCols, real *LVals,
               int ASize, real alpha)
{
    for (int I=threadIdx.x + blockIdx.x*blockDim.x; I<ASize; I += blockDim.x*gridDim.x)
    {
        ARows[I] = LRows[I];
        ACols[I] = LCols[I];
        AVals[I] = -alpha*LVals[I] + (LRows[I]==LCols[I])*MVals[LRows[I]];
    }
} // generateA


__global__
void generateADirectForcing(int *ARows, int *ACols, real *AVals,
                            real *MVals,
                            int *LRows, int *LCols, real *LVals,
                            int ASize, real alpha, int *tags)
{
    for(int I=threadIdx.x + blockIdx.x*blockDim.x; I<ASize; I += blockDim.x*gridDim.x)
    {
        ARows[I] = LRows[I];
        ACols[I] = LCols[I];
        AVals[I] =   (tags[LRows[I]] == -1)*(-alpha*LVals[I]) // if the current location is untagged, add -alpha*L
                   + (tags[LRows[I]] != -1)*(-LVals[I]) // if the current location is tagged, add -L
                   + (LRows[I]==LCols[I])*MVals[LRows[I]]; // if it is a diagonal, add M
    }
} // generateADirectForcing
    
} // End of namespace kernels