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

[Orphic Dreamer via cfe-commits]

Siya-05 wrote:

Hi @andykaylor , just following up on this PR.

I addressed the review comments and the latest CI is passing. I wasn't sure whether the earlier MemorySanitizer failures were related to my changes, so I wanted to check if there's anything else you'd like me to address before this can move forward.

Thanks!

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