Problem with cub::BlockLoad

Accelerated Computing GPU-Accelerated Libraries
1

I have this problem with in my code:

no instance of overloaded function “cub::BlockLoad<InputT, BLOCK_DIM_X, ITEMS_PER_THREAD, ALGORITHM, BLOCK_DIM_Y, BLOCK_DIM_Z, PTX_ARCH>::Load [with InputT=__nv_bool *, BLOCK_DIM_X=240, ITEMS_PER_THREAD=1, ALGORITHM=cub::BLOCK_LOAD_DIRECT, BLOCK_DIM_Y=1, BLOCK_DIM_Z=1, PTX_ARCH=600]” matches the argument list

template<int blockthreads>
__global__ static void cudaSOSInRegion (
		unsigned int * message,
		const unsigned int * region,
		bool * connexionMatrix, unsigned int * result,
		const int totalNbFanals,
		const unsigned int regionLength,
		const unsigned int memoryEffect)
{
	unsigned long idx = blockIdx.x * blockDim.x
			+ threadIdx.x;
	if (idx < totalNbFanals)
	{
		// Specialize BlockStore for a 1D block of 256 threads owning 1 boolean items each
		typedef cub::BlockLoad<bool*, blockthreads, 1> BlockLoad;
		// Allocate shared memory for BlockStore
		__shared__ typename BlockLoad::TempStorage temp_storage;

		// Obtain a segment of consecutive items that are blocked across threads
		bool thread_data[1];
		unsigned long messageId = message[blockIdx.y];
		unsigned long partie1 = (messageId
				* totalNbFanals)
				+ (blockIdx.x * blockDim.x);

		BlockLoad(temp_storage).Load(
				connexionMatrix + partie1,
				thread_data);
		if (thread_data[0] > 0)
		{
			atomicAdd(result + idx,
					1
							+ (messageId == idx)
									* memoryEffect);
		}
	}
}

These are the parameters according to CUB: cub::BlockLoad< InputT, BLOCK_DIM_X, ITEMS_PER_THREAD, ALGORITHM, BLOCK_DIM_Y, BLOCK_DIM_Z, PTX_ARCH > Class Template Reference

InputT The data type to read into (which must be convertible from the input iterator’s value type).
BLOCK_DIM_X The thread block length in threads along the X dimension
ITEMS_PER_THREAD The number of consecutive items partitioned onto each thread.

Do you know what could be the problem ?

Thanks you