CL_INVALID_BINARY (-42) while trying to use double on Tesla

Accelerated Computing CUDA CUDA Programming and Performance
1

I get CL_INVALID_BINARY (-42) in Program::build() while trying to use double on Tesla C1060, here is the code:

#pragma OPENCL EXTENSION cl_khr_fp64 : enable

__kernel void compute(__global double *vd0,
__global double *vd1,
__global double *vd2)
{
uint index = get_global_id(0);
(vd2[index]=2);
(vd0[index]=(2+pown(vd2[index],3)));
(vd1[index]=index);
}

what is wrong?

thanks

2

Solution: I had 2 cards and only one of them had double-precision support, so while creating context one needs only to mention one:

devices.push_back(context.getInfo<CL_CONTEXT_DEVICES>()[0]);