[clang] [CIR][CUDA] Support built-in CUDA surface type (PR #196079)

[Akimasa Watanuki via cfe-commits]

Men-cotton wrote:

These APIs are already present on current main. See #189345 and #195382.
It seems like the original branch point is older than them, so you might need to update your local environment.

Could you update your local branch and rebuild ClangIR? If the errors still reproduce after that, please share the log.

https://github.com/llvm/llvm-project/pull/196079