Support kernel splitting in code generation

doc/ref_transform.rst

doc/tutorial.rst

examples/python/global_barrier_removal.py

examples/python/ispc-stream-harness.py

loopy/__init__.py

loopy/auto_test.py

loopy/check.py

loopy/codegen/__init__.py

loopy/codegen/bounds.py

loopy/codegen/control.py

loopy/codegen/device_mapping.py

loopy/codegen/instruction.py

loopy/codegen/loop.py

loopy/codegen/result.py

loopy/compiled.py

loopy/expression.py

loopy/kernel/__init__.py

loopy/kernel/array.py

loopy/kernel/data.py

loopy/kernel/tools.py