* Add make_einsum

* einsum: allow customizing arg names, specify name

* Fix double use of variable i in make_einsum

Co-authored-by: Kaushik Kulkarni <15399010+kaushikcfd@users.noreply.github.com>

* Use dictionary unpacking for knl_creation_kwargs

* Add some einsum unit tests

* Fix up einsum tests

* Update comment: 'diagonalization' -> 'extract diagonal'

* Add documentation for "make_einsum"

* Fix doc build errors

* Check that the number of arg names matches the number of arg specs

* Add error handling test

* Update make_einsum docstring

* Compactify kernel creation

* Ensure output subscripts contain unique indices

* Add einsum sanity check in unit tests

* Use set comprehension in make_einsum

* Add note about lack of complexity reduction in make_einsum

* Remove whitespace in docstring (flake8)

* Fix doc syntax in make_einsum

Co-authored-by: Kaushik Kulkarni <15399010+kaushikcfd@users.noreply.github.com>
Co-authored-by: Thomas Gibson <gibsonthomas1120@hotmail.com>

Add a benchmark

See merge request !397

Do not fail fast on downstream matrix build

Add LoopKernel.default_entrypoint for kernel callables forward compat…

[Minor]: Clarify a conditional expression

Do not map args, tvs in split_iname

Co-authored-by: Kaushik Kulkarni <15399010+kaushikcfd@users.noreply.github.com>

Make guess_var_shape work on multiple var_names simultaneously

Don't change the kernel in preprocess_kernel if there's no change

Make discard_dep_reqs_in_order faster

Add wrap_in_typecast_lazy