[!305] Performance transformation: Loop reordering in sumfact kernel

Merge branch 'feature/sumfact-loop-reordering' into 'master'

ref:extensions/dune-codegen Performance transformation through loop nest
reordering. There are two ways to reorder loops in a tensor contraction:

1.  Directly accumulate in output variable after setting to zero
2.  Accumulating in a large enough temporary

This merge request implements these ways of loop reordering and the
possibility to create an autotune target directly from the loopy kernel.

See merge request [extensions/dune-codegen!305]

  [extensions/dune-codegen!305]: gitlab.dune-project.org/extensions/dune-codegen/merge_requests/305