- Convert GpuKernelBase to use contexts.
- Start using explicit context for all the ops.

This may break some stuff outside of basic_ops.py since sometimes a
context will be required, but should be ok otherwise.

Still in transitory state, this should work with ops that rely on type
context or default context (as long as they are equal).  Will
gradually move to type context only.

Otherwise it is possible to get into a situation where we use the dnn flags, but the cuda ones aren't defined and stuff breaks.

Also starts adressing issue #3468 a bit.

Speedup scan grad() method

renaming tensor4d to tensor4 in stack examples

Use make_new_inplace mechanism to fix optimization with cast

Fix stack(tensors, axis=0) documentation