[Torch Dialect] emit aten.nonzero op, recompose aten.index.Tensor with aten.nonzero_numpy to aten.masked_select

[Vremold]

In detail, this PR includes:

1. emit aten.nonzero and aten.nonzero_numpy op;
2. add missing shape infer function for aten.masked_select op
3. add recomposition rule for aten.index.Tensor op. When it's indices come from aten.nonzero_numpy op and the shapes of associated tensor are exactly same, it's equivalent to a aten.masked_select op. But for now, no backend supports aten.masked_select or aten.nonzero_numpy op, this PR only adds a MLIR test case.

PS: This PR relies on canonicalizer of prim.ListConstruct op which is introduced in PR #2306 .