TEST device(cuda)

aprod.c

@@ -336,6 +336,7 @@ void aprod(int mode, long int m, long int n, double *vVect, double *knownTerms,
                                
                 for(int nt=0; nt < ntasks; nt++ )
                 {
+		    #pragma omp target device (cuda) copy_deps	
                     #pragma omp task label(vVect_nAstroPSolved) //out(vVect[jstartAstro : jstartAstro + nAstroPSolved-1]) 
                     {
                         lset = mapForThread[nt][0] * nparam;