Implement name inference for t(), transpose(...)

[zou3519]

Stack from ghstack:

• Implement name inference for torch.dot #24474 Implement name inference for torch.dot
• Implement name inference for addmv, addmv_, mv #24471 Implement name inference for addmv, addmv_, mv
• Implement name inference for expand #24469 Implement name inference for expand
• Implement name inference for mm, addmm, addmm_ #24306 Implement name inference for mm, addmm, addmm_
• Fix binary op name inference between unnamed and named tensors. #24921 Fix binary op name inference between unnamed and named tensors.
• Add helper function Tensor::names() #24914 Add helper function Tensor::names()
• Rename Tensor::names() to Tensor::opt_names() #24907 Rename Tensor::names() to Tensor::opt_names()
• Fix FIXME_default_names by storing static list of 64 none names #24885 Fix FIXME_default_names by storing static list of 64 none names
• Add thread-local-state NamesMode and NoNamesGuard #24942 Add thread-local-state NamesMode and NoNamesGuard
• Implement name inference for t(), transpose(...) #24941 Implement name inference for t(), transpose(...)

Test Plan

• [namedtensor ci]

Differential Revision: D16930707

[zou3519]

@nairbv this is a resubmit of the reverted #24203. The reason for the revert was a merge conflict / race in editing a checked in autogenerated file (see #24931 for details).

[facebook-github-bot]

@zou3519 merged this pull request in 3a59a9b.