Replace remaining uses of linearize_access_indexes for broadcasting with BroadcastIndexesRange

BroadcastIndexesRange is more efficient than linearize_access_indexes-based loops. We have several remaining places that use linearize_access_indexes, and what they have in common is that they're not looping over an existing Tensor's .sizes() and .strides(). To make these work nicely, we need a way for BroadcastIndexesRange to play nicely with .sizes() and .strides() ArrayRefs that aren't attached to a Tensor.

Specific usage sites that can be improved:
https://github.com/pytorch/executorch/blob/95f779ae2120d94e20bb95ae6af45da76ce3ff52/kernels/portable/cpu/op_cdist_forward.cpp#L79

https://github.com/pytorch/executorch/blob/95f779ae2120d94e20bb95ae6af45da76ce3ff52/kernels/portable/cpu/op_index_put.cpp#L145.(I think?)

https://github.com/pytorch/executorch/blob/95f779ae2120d94e20bb95ae6af45da76ce3ff52/kernels/portable/cpu/op_split_with_sizes_copy.cpp#L129

https://github.com/pytorch/executorch/blob/95f779ae2120d94e20bb95ae6af45da76ce3ff52/kernels/portable/cpu/op_masked_select.cpp#L125 (might be straightforwardly fixable right away with no extra work)

cc @larryliu0820 @manuelcandales

Start by reading BroadcastIndexesRange and the cited usage sites: kernels/portable/cpu/op_cdist_forward.cpp, op_index_put.cpp, op_split_with_sizes_copy.cpp, and op_masked_select.cpp. Determine how its iteration can accept standalone sizes and strides ArrayRefs, then replace the remaining linearize_access_indexes-based broadcasting loops and validate the affected kernels.