Refactor thrust::reduce to use cub::DeviceReduce

[elstehle]

This is a sub-task of Thrust/CUB kernel consolidation #26

Prepare cub::DeviceReduce for feature parity needed by thrust::reduce:

• Introduce vsmem utility to cub::DeviceReduce
• Add tests to CUB that check that cub::DeviceReduce correctly uses the fallback policy (see fdf565e)
• Add tests to CUB that check that cub::DeviceReduce correctly uses virtual shared memory (see fdf565e)

Refactor thrust::reduce to use cub::DeviceReduce:

• Make thrust::reduce use cub::DeviceReduce (see 948817e)
• Add dynamic 32/64-bit offset type-dispatch to thrust::reduce (see 948817e L210-216)
• Add sanity tests for large number of items for thrust::reduce (see 01f32dd)
• Add more elaborate testing for cub::DeviceReduce (see Adds tests for large number of items for cub::DeviceSelect #1612)