arrays of pointers

hi all,

is there a special way i should be sending arrays of pointers to functions in cuda, and then using them? this is how i do it now:

__device__ inline void foo(float3 **v, int numValues) {

   for (int i=0; i<numValues; i++) {

        v[i]....                                       // this crashes

   }

  v[0], v[1]...                   // this works

}

any ideas?

Does it crash in emulation mode, too? I use

for (int i=0; i<numValues; i++) {

       v[i]....                                       // this crashes

  }

in my code (except with float4’s) without any problems. I would suspect your crash is caused somewhere else and somehow the code generated by the for loop triggers it.